WO2009114928A1

WO2009114928A1 - Method and apparatus for assessing characteristics of liquids

Info

Publication number: WO2009114928A1
Application number: PCT/CA2008/001591
Authority: WO
Inventors: Dan Gudmundson; A Doyle; Vinh Phuc Pham
Original assignee: Optosecurity, Inc.
Priority date: 2008-03-17
Filing date: 2008-09-05
Publication date: 2009-09-24

Abstract

A method to perform qualification of liquids by using an electromagnetic radiation scan, such as X-rays. The method includes scanning a product that contains a liquid to be qualified and using the scan results to search a database of candidates for a potential match. If a match is found the liquid is considered equivalent to the candidate. The invention can be used to perform quality control or arranging commercial liquid materials into classes or categories.

Description

TITLE: Method and apparatus for assessing characteristics of liquids

FIELD OF THEINVENTION

The present invention relates to technologies for assessing properties of liquids, in particular for qualifying liquids. The invention has numerous applications, in particular it can be used for performing quality control or arranging commercial liquid materials into classes or categories.

BACKGROUND OF THEINVENTION

Mass production of liquids for industrial use or human or animal consumption requires some sort of quality control step to ensure that the liquid material meets quality standards. In the case of industrial liquid materials the quality standards are meant to insure a proper product performance when the product is used. Another characteristic that a quality control step ascertains is the safety of the liquid material. Many industrial liquid materials present inherent safety risks and if the production process has not been conducted properly, those safety risks can be increased.

The quality control performed on liquids for human or animal consumption (by human or animal consumption is mean materials that are intended to be ingested by a human or animal, such as food products or medicines or products that are intended to be in intimate contact with the body of the human or animal, such as preparations to clean or beatify or soothe the body by direct application) is even more important. In many instances, liquid materials for human or animal consumption are of perishable nature and easily contaminated by bacteria or other ailments. The quality control process is intended to insure that when a liquid material is made available for public consumption the liquid material is predominantly safe and will not cause any harm. Also the quality control process ensures that the liquid material, in particular liquid material for human consumption will have the expected flavour and aroma characteristics.

In the specific case of medicines, the quality control is event more stringent since manor deviations from a preset chemical formulation can cause serious harm to the patient. Such deviations can occur as a result of equipment malfunction, human mistakes or tampering.

Quality control processes as currently implemented are costly and provide uneven results, particularly in the case of large production runs where it is not practical or even possible to test a large number of samples. In such instances, a very small sample of product population is subjected to testing which usually is destructive testing, meaning that the tested sample can no longer be used and must be discarded after the test is performed. Also, the results of the testing process may become available only after some time, such as hours or days, which means that if problems are discovered many defective articles would have been already produced, packaged and maybe even sold to consumers.

Against this background it clearly appears that a need exists in the industry to develop improved methods and devices to perform qualification of liquid materials, such as quality control operations.

SUMMARY OF THE INVENTION

As embodied and broadly described herein the invention provides a method to perform quality inspection of a liquid material held in a container, the method including: a. subjecting the liquid material to X-ray radiation to generate X-ray image data representing an X-ray image of the container holding the liquid; b. comparing information derived from the X-ray image data to a plurality of database records, each database record including reference information corresponding to a particular level of quality, among several possible levels or quality of the liquid material; c. releasing information indicative of the level of quality of the liquid material, the level of quality conveyed by the released information corresponding to the database record matched to the X-ray image data by the comparing.

As embodied and broadly described herein the invention also provides an apparatus for performing quality inspection of a liquid material held in a container, the apparatus including: a. an X-ray inspection station for subjecting the liquid material to X- ray radiation to generate X-ray image data representing an X-ray image of the container holding the liquid; b. a computer based processing unit for comparing information derived from the X-ray image data to a plurality database records, each database record including reference information corresponding to a particular level of quality, among several possible levels or quality of the liquid material; c. an output for releasing information indicative of the level of quality of the liquid material, the level of quality conveyed by the released information corresponding to the database record matched to the X-ray image data by the comparing.

As embodied and broadly described herein the invention also provides a method to determine if a liquid material held in a container is safe for human consumption, the method including: a. subjecting the liquid material to X-ray radiation to generate X-ray image data representing an X-ray image of the container holding the liquid; b. comparing information derived from the X-ray image data to a plurality database records, the plurality of database records, including: i. a first database record including reference information corresponding to a liquid material that is safe for human consumption; ii. a second database record including reference information corresponding to a liquid material that is not safe for human consumption; c. releasing information on the safeness of the liquid product for human consumption on the basis of the results of the comparing.

As embodied and broadly described herein the invention also provides an apparatus to determine if a liquid material held in a container is safe for human consumption, the apparatus including: a. an inspection station for subjecting the liquid material to X-ray radiation to generate X-ray image data representing an X-ray image of the container holding the liquid; b. a computer based processing unit comparing information derived from the X-ray image data to a plurality database records, the plurality of database records, including: i. a first database record including reference information corresponding to a liquid material that is safe for human consumption; ii. a second database record including reference information corresponding to a liquid material that is not safe for human consumption; c. an output for releasing information on the safeness of the liquid product for human consumption on the basis of the results of the comparing.

As embodied and broadly described herein the invention also provides a method to determine a level of sugar content of a liquid material held in a container, the method including: a. subjecting the liquid material to X-ray radiation to generate X-ray image data representing an X-ray image of the container holding the liquid; b. comparing information derived from the X-ray image data to a plurality database records, each database record including reference information corresponding to a particular level of sugar, among several possible levels or sugar in the liquid material;

c. releasing information indicative of the level of sugar content in the liquid material, the level of sugar content conveyed by the released information corresponding to the database record matched to the X-ray image data by the comparing.

As embodied and broadly described herein the invention also provides an apparatus to determine a level of sugar content of a liquid material held in a container, the method including: a. an inspection station for subjecting the liquid material to X-ray radiation to generate X-ray image data representing an X-ray image of the container holding the liquid; b. a processing unit for comparing information derived from the X-ray image data to a plurality database records, each database record including reference information corresponding to a particular level of sugar, among several possible levels or sugar in the liquid material; c. an output for releasing information indicative of the level sugar content in the liquid material, the level of sugar content conveyed by the released information corresponding to the database record matched to the X-ray image data by the comparing.

As embodied and broadly described herein the invention also provides a method to determine a level of alcohol content of a liquid material held in a container, the method including: a. subjecting the liquid material to X-ray radiation to generate X-ray image data representing an X-ray image of the container holding the liquid; b. comparing information derived from the X-ray image data to a plurality database records, each database record including reference information corresponding to a particular level of alcohol, among several possible levels or alcohol of the liquid material; c. releasing information indicative of the level of alcohol of the liquid material, the level of alcohol conveyed by the released information corresponding to the database record matched to the X-ray image data by the comparing.

As embodied and broadly described herein the invention also provides an apparatus to determine a level of alcohol of a liquid material held in a container, the method including: a. an inspection station for subjecting the liquid material to X-ray radiation to generate X-ray image data representing an X-ray image of the container holding the liquid; b. a processing unit for comparing information derived from the X-ray image data to a plurality database records, each database record including reference information corresponding to a particular level of alcohol, among several possible levels or alcohol of the liquid material; c. an output for releasing information indicative of the alcohol content in the liquid material, the level of alcohol conveyed by the released information corresponding to the database record matched to the X-ray image data by the comparing. As embodied and broadly described herein the invention also provides a method to determine the taste of a liquid material held in a container, the method including: a. subjecting the liquid material to X-ray radiation to generate X-ray image data representing an X-ray image of the container holding the liquid; b. comparing information derived from the X-ray image data to a plurality database records, each database record including reference information corresponding to a different taste, among several possible tastes; c. releasing information indicative of the taste of the liquid material, the taste conveyed by the released information corresponding to the database record matched to the X-ray image data by the comparing.

As embodied and broadly described herein the invention also provides an apparatus to determine the taste of a liquid material held in a container, the method including: a. subjecting the liquid material to X-ray radiation to generate X-ray image data representing an X-ray image of the container holding the liquid; b. comparing information derived from the X-ray image data to a plurality database records, each database record including reference information corresponding to a different taste, among several possible tastes; c. releasing information indicative of the taste of the liquid material, the taste conveyed by the released information corresponding to the database record matched to the X-ray image data by the comparing. As embodied and broadly described herein the invention also provides a method to determine the mouth felt texture of a liquid material held in a container, the method including: a. subjecting the liquid material to X-ray radiation to generate X-ray image data representing an X-ray image of the container holding the liquid; b. comparing information derived from the X-ray image data to a plurality database records, each database record including reference information corresponding to a different mouth felt texture, among several possible mouth felt textures; c. releasing information indicative of the mouth felt texture of the liquid material, the mouth felt texture conveyed by the released information corresponding to the database record matched to the X-ray image data by the comparing.

As embodied and broadly described herein the invention also provides an apparatus to determine the mouth felt texture of a liquid material held in a container, the method including: a. subjecting the liquid material to X-ray radiation to generate X-ray image data representing an X-ray image of the container holding the liquid; b. comparing information derived from the X-ray image data to a plurality database records, each database record including reference information corresponding to a different mouth felt texture, among several possible mouth felt textures; c. releasing information indicative of the mouth felt texture of the liquid material, the mouth felt texture conveyed by the released information corresponding to the database record matched to the X-ray image data by the comparing.

As embodied and broadly described herein the invention also provides a method to method to determine if a container bearing a liquid product identification holds a liquid matching that product identification, the method including: a. subjecting the liquid material to X-ray radiation to derive a response of the liquid to X-ray radiation; b. inspecting the container to derive product identification information; c. providing a knowledge bank containing responses of different liquids to X-rays, each response being mapped to a corresponding product identification information; d. searching the knowledge bank to determine if an entry exists in the knowledge bank matching the product identification information acquired by the inspecting and also matching the response of the liquid to X-ray radiation acquired by the subjecting; e. using the results of the searching to determine if the liquid in the container matches the product identification on the container.

BRIEF DESCRIPTION OF THE DRAWINGS

A detailed description of examples of implementation of the present invention is provided hereinbelow with reference to the following drawings, in which:

Figure 1a is a block diagram of an apparatus using X-rays to perform qualification of liquids, according to a non-limiting example of implementation of the invention;

Figure 1 b is a more detailed illustration of the X-ray apparatus of Figure 1a;

Figure 2 is a more detailed block diagram of the processing module of the apparatus shown in Figure 1 b;

Figure 3 is a generalized block diagram of the process implemented by the apparatus at Figure 1 ; Figure 4 is graph illustrating the total X-ray attenuation in H₂O due to various X- ray matter interactions;

Figure 5 is a generalized illustration of the photoelectric X-ray absorption process;

Figure 6 is a generalized illustration of the Compton scattering effect;

Figure 7 is detailed block diagram of a first non-limiting example of the process shown in Figure 3;

Figure 8 illustrates an experimental set-up for implementing the method shown in Figure 7;

Figure 9 is an illustration showing the image information derived from the set-up of Figure 7;

Figure 10 is an X-ray image of a container holding a liquid, showing a line-like Region Of Interest (ROI) along which grey level values are calculated;

Figure 11 is a graph illustrating the grey level profile along the ROI of Figure 10;

Figure 12 is a graph illustrating the grey level profile of a high energy (hi-E) X- ray image of a container holding a liquid, showing that the grey level profile matches the cross-sectional shape of the container;

Figure 13 is a graph illustrating the grey level profile of the low energy (low-E) X- ray image of the container shown in Figure 12, also showing that the low -E grey level profile matches the cross-sectional shape of the container;

Figures 14 to 18 are graphs illustrating the grey level profiles of hi-E X-ray images of different liquid containers and the corresponding container shapes; Figure 19a is a detailed block diagram of a second non-limiting example of implementation of the process shown in Figure 3;

Figure 19b is a table-like representation of a knowledge bank storing information about liquid products;

Figure 20 is a set-up for implementing the method shown in Figure 19a;

Figure 21 is a graph showing the variation of the diffraction/scattering signature with molecular density;

Figure 22 is a flow chart of a process for performing X- ray image processing to remove the contribution in the image of the belt of the X-ray imaging system;

Figure 23 is a flow chart of a process for performing X-ray image processing to determine in the X-ray image the location and orientation of a tray;

Figure 24 is a flow chart of a process for performing X-ray image processing to remove the contribution in the image of the tray detected in Figure 23;

Figure 25 is a flowchart of a process for performing a calculation of the density and the effective atomic number of a liquid in an X-ray image;

Figure 26 is a flowchart of a process for performing X-ray image processing to remove the contribution in the image of the wall of a container that appears in the image;

Figure 27 is a flowchart of a process for performing qualification of a liquid;

Figure 28 is a diagram of an X-ray image scanner illustrating a method to compute the path length of the X-ray beams through a body of liquid held inside a container; Figure 29 is a simulated x-ray image of two overlapping containers;

Figure 30 is a flowchart of a process for allowing the operator to specify on the image at Figure 29 the container to be analyzed;

Figure 31 is a simulated X-ray image illustrating the mapping between image portions and individual detectors of the X-ray imaging system.

In the drawings, embodiments of the invention are illustrated by way of example. It is to be expressly understood that the description and drawings are only for purposes of illustration and as an aid to understanding, and are not intended to be a definition of the limits of the invention.

DETAILED DESCRIPTION

With reference to Fig. 1a, there is shown a specific non-limiting example of a system 10 for use in qualifying liquid materials, in accordance with a non-limiting embodiment of the present invention. The system 10 comprises an x-ray apparatus 100 that applies an x-ray screening process to a liquid 104 (Note that for the purpose of this specification "liquid" refers to a state of matter that is neither gas nor solid and that generally takes the shape of the container in which it is put. This definition would, therefore encompass substances that are pastes or gels, in addition to substances having a characteristic readiness to flow. For instance, toothpaste, and other materials having the consistency of toothpaste would be considered to fall in the definition of "liquid". ) contained in a container 102 that is located within a screening area of the x-ray apparatus 100.

The processing module 200 may be co-located with the x-ray apparatus 100 or it may be remote from the x-ray apparatus 100 and connected thereto by a communication link, which may be wireless, wired, optical, etc. The processing module 200 receives the image signal 1 16 and executes a method (to be described later on) to produce information 1 18 that qualifies the liquid. The processing module 200 has access to a database 400 which constitutes a knowledge bank via a communication link 120 that may be local to the processing module 200 (e.g., on a common printed circuit board, or connected as a peripheral device thereto by cable or Bluetooth), or which can be remote from the processing module 200 (e.g., connected via a wireline, wireless or optical link that may traverse a data network). The processing module 200 may be implemented using software, hardware, control logic or a combination thereof.

The information that qualifies the liquid is provided to a console 350, where the information 118 can be conveyed to an operator 130 or other security personnel. The console 350 can be embodied as a piece of equipment that is in proximity to the x-ray apparatus 100, or can be embodied as a piece of equipment that is remote from the x-ray apparatus 100.

The console 350 may comprise suitable software and/or hardware and/or control logic to implement a graphical user interface (GUI) for permitting interaction with the operator 130. Consequently, the console 350 may provide a control link 122 to the x-ray apparatus 100, thereby allowing the operator 130 to control motion (e.g., forward/backward and speed) of the conveyor 1 14 and, as a result, to control the position of the container 102 within the screening area of the x-ray apparatus 100.

In accordance with a specific non-limiting embodiment, and with reference to Fig. 1 b, the x-ray apparatus 100 is a dual-energy x-ray apparatus 100A. However, persons skilled in the art will appreciate that the present invention is not limited to such an embodiment. Continuing with the description of the dual- energy x-ray apparatus 100A, an x-ray source 202 emits x-rays 206 at two distinct photon energy levels, either simultaneously or in sequence. Example energy levels include 50 keV (50 thousand electron-volts) and 150 keV, although persons skilled in the art will appreciate that other energy levels are possible.

Generally speaking, x-rays are typically defined as electromagnetic radiation having wavelengths that lie within a range of 0.001 to 10 nm (nanometers) corresponding to photon energies of 120 eV to 1.2 MeV. Although the electromagnetic radiation referred to primarily throughout this description are x- rays, those skilled in the art will appreciate that the present invention is also applicable to electromagnetic radiation having wavelengths (and corresponding photon energies) outside this range.

A detector 218 located generally along an extension of the path of the x-rays 206 receives photons emanating from the combination of the liquid 104 and the container 102 in which it is located. Some of the incoming photons (X-rays 206) will go straight through the container/liquid 104 combination while some will interact with the container/liquid 104 combination. There are a number of interactions possible, such as:

• The Rayleigh scattering (coherent scattering)

• The photoelectric absorption (incoherent scattering)

• The Compton scattering (incoherent scattering)

• The pair production;

• Diffraction

The total attenuation of the contribution of the various X-rays - matter interactions is shown in figure 4. In this example the matter is H₂O but the attenuation profile for other materials is generally similar. For today's state-of- the-art security screening systems, the energy levels commonly utilized lie between 50 keV and 150 keV. The photoelectric absorption (Figure 5) of X-rays occurs when the X-ray photon is absorbed, resulting in the ejection of electrons from the shells of the atom, and hence the ionization of the atom. Subsequently, the ionized atom returns to the neutral state with the emission of whether an Auger electron or an X-ray characteristic of the atom. This subsequent X-ray emission of lower energy photons is however generally absorbed and does not contribute to (or hinder) the image making process. This type of X-ray interaction is dependent on the effective atomic number of the material or atom and is dominant for atoms of high atomic numbers. Photoelectron absorption is the dominant process for X- ray absorption up to energies of about 25 keV. Nevertheless, in the energy range of interest for security applications, the photoelectric effect plays a smaller role with respect to the Compton scattering, which becomes dominant.

Compton scattering (Figure 6) occurs when the incident X-ray photon is deflected from its original path by an interaction with an electron. The electron gains energy and is ejected from its orbital position. The X-ray photon looses energy due to the interaction but continues to travel through the material along an altered path. Since the scattered X-ray photon has less energy, consequently it has a longer wavelength than the incident photon. The event is also known as incoherent scattering, because the photon energy change resulting from an interaction is not always orderly and consistent. The energy shift depends on the angle of scattering and not on the nature of the scattering medium. Compton scattering is proportional to material density and the probability of it occurring increases as the incident photon energy increases.

The diffraction phenomenon of the x-rays by a material with which they interact is related to the scattering effect described earlier. When the x-rays are scattered by the individual atoms of the material, the scattered x-rays may then interact and produce diffraction patterns that depend upon the internal structure of the material that is being examined. The photons received by the detector 218 include photons that have gone straight through the liquid 104 and the container 102; these photons have not interacted in any significant matter with the liquid 104. Others of the received photons have interacted with the liquid 104 or the container.

In accordance with a specific non-limiting embodiment of the present invention, the detector 218 may comprise a low-energy scintillator 208 and a high-energy scintillator 210, which can be made of different materials. The low-energy scintillator 208 amplifies the intensity of the received photons such that a first photodiode array 212 can produce a low-energy image 220. Similarly, the high- energy scintillator 210 amplifies the intensity of the received photons such that a second photodiode array 214 can produce a high-energy image 222. The low- energy image 220 and the high-energy image 222 may be produced simultaneously or in sequence. Together, the low-energy image 220 and the high-energy image 222 form the aforesaid image signal 116.

Referring back to Figure 1a, the processing module 200 receives the image signal 1 16 and processes the signal in conjunction with data contained in a knowledge bank 400 to qualify the liquid material. The qualification can include an explicit determination as to whether the liquid material meets a predetermined quality standard. Alternatively, the qualification of the liquid material can be effected without explicitly saying whether the liquid meets a particular quality standard. For example, the processing module can determine that the liquid is "high grade", "intermediate grade" or "low grade" hence the operator 130 would be able to conclude what type of further treatment or conditioning the material has be subjected on the basis of its qualification. A "high grade material may attract a higher price than a lower grade material. In a different example, a higher grade material may attract a different shelf life, for instance longer shelf life than a lower grade material. In a yet different example a higher grade material may require different packaging or labeling than a lower grade material. The results of the qualification are conveyed in the information 118 which is communicated to the console 350. Figure 2 is a high level block diagram of the processing module 200. The processing module 200 has a Central Processing Unit (CPU) 300 that communicates with a memory 302 over a data bus 304. The memory 302 stores the software that is executed by the CPU 300 and which defines the functionality of the processing module 200. The CPU 300 exchanges data with external devices through an Input/Output (I/O) interface 306. Specifically, the image signal 116 is received at the I/O interface 306 and the data contained in the signal is processed by the CPU 300. The qualification signal 118 that is generated by the CPU 300 is output to the console 350 via the I/O interface 306. Also, communications between the knowledge bank 400 and the processing module 200 are made via the I/O interface 306.

Figure 3 is a high level block diagram that illustrates the functions performed by the processing module 200 in qualifying the liquid material. This block diagram applies to the example of implementation shown at Figure 1a and also to other examples of implementation that will be described later. The first step of the process, illustrated at 401 is to perform a characterization of the product that is being screened. By "product" is meant the combination container and liquid inside. The characterization step returns information conveying distinctive features of the product that allows distinguishing the product from other products. The characterization step is performed on the container but it may also include the liquid inside. For instance the characterization step 401 may return information such as the general shape of the container, its height, cross- sectional profile and width among many other parameters. Characterization of the liquid is optional and may provide information such as the color of the liquid (assuming of course the container is transparent).

The characterization step 401 can be performed by using different types of equipment capable to capture the distinctive features of the product. One example is an apparatus using penetrating radiation such as the X-ray imaging system 100 of Figure 1a. This is convenient since the same apparatus can be used to characterize the product and also obtain the response of the liquid in the container to X-rays. Yet another example is to use a device that will obtain an image of the product and perform the characterization based on that image. The image may be two dimensional or three dimensional. Yet another possibility is to use equipment to read machine readable labels or tags on the container. The reading can be done optically or via radio frequency (RF) information capture.

Note that the characterization step 401 is optional. The characterization step can be useful in instances where the apparatus 100 qualifies products that are not necessarily the same. For example the apparatus 100 receives different product samples and performs the qualification of the different product samples. The samples may differ in terms of the composition of the liquid or differ in terms of packaging. In the former case the apparatus 100 would process products that have different liquid materials (water, wine oil, industrial chemicals, medicines, etc) . In the latter case the apparatus 100 would be processing products that have the same liquid material but the packaging (for example orange juice in different size containers).

In instances where the apparatus 100 qualifies the samples of the same product the characterization step 401 may be dispensed with.

The characterization step of the product is followed by a determination of the response of the liquid in the container to X-rays, as shown at step 402. The response represents the interaction of the liquid with the X-rays as discussed above. The response can be expressed in terms parameters characterizing the liquid. Examples of parameters include:

• The density of the liquid;

• The effective atomic number of the liquid (Z_eff);

• The diffraction/scattering signature • The viscosity of the liquid

At step 404, a knowledge bank is searched on the basis of the product characterization performed at step 400. The knowledge bank contains characterization data for a number of products that the apparatus 100 is designed to qualify. For instance the apparatus 100 is designed to test a range of soft drink products, namely 100ml soft drink bottle, 200 ml soft drink bottle and a 300 ml soft drink bottle. The knowledge bank would, therefore contain characterization data for each of the products. Also, the knowledge bank contains the associated responses to X-rays of reference liquids in the containers. So₁ step 404 searches the knowledge bank to locate one or more entries that match the product characterization derived at step 400. If one or more entries are found that match the product characterization, the corresponding responses to X-rays are extracted from the knowledge bank and compared to the response obtained at step 402. If the response extracted from the knowledge bank 400 matches the response obtained at step 402 then the process concludes that the product that is being screened is equal to the reference, in other words the liquid inside is of the same quality or class as the reference.

Note that the characterization data may be associated with responses to X-rays of several reference liquids, where each reference may be associated with a quality level. For example, there may be an X-ray response associated to liquid that is spoiled and therefore not suitable for human consumption (first reference liquid), and X-ray response associated with a liquid that while still suitable for human consumption is not of premium quality (second reference liquid) and an X-ray response associated to a liquid that is of premium quality (third reference liquid).

The process determines at step 406 a qualification on the basis of the knowledge bank search. The qualification conveys qualification information about the product. For example, the qualification information may indicate that the product is not suitable for human consumption if a match has been found with the first reference liquid. Accordingly that product should not be sold and should be discarded. Alternatively, the qualification information may indicate that the product is suitable for human consumption but is not of premium quality, if a match has been found with the second reference liquid, therefore it may need to be sold at a lower price or processed or packaged differently. Also, the qualification information may indicate that the product is of premium quality if a match has been found with the third reference liquid, therefore it may be sold at a premium price or processed or packaged in a way to reflect that qualification.

Figure 7 is a more detailed flowchart of the process for performing a qualification process on a container holding a liquid, according to a first non- limiting example of implementation. The process uses X-ray scanning to perform the characterization of the product (container + liquid) and also to determine the response of the liquid to X-rays. In other words a single X-ray scan is used to extract both pieces of information. One example of an X-ray imaging system that can be used for this purpose is the equipment manufactured by Gilardoni in Italy, model number FEP ME 640 DETEX. This machine is a dual energy device that produces X-rays at high and low energy values that are HI (high) = 74.298 keV and Lo (low) = 55.398 keV, respectively.

Figure 8 illustrates the general configuration of the X-ray imaging system. The machine 800 has a conveyor belt 802 on which items to be scanned are placed. The X-ray source 804 is located below the conveyor belt 802. Detector arrays 806, 808 are placed on the vertical and the horizontal walls of the casing. For clarity, when the conveyor belt 802 advances the container through the x-ray machine 800, the direction of movement would follow an imaginary line that would be perpendicular to the sheet of the drawing.

A container that is being scanned is shown at 810. In this example, the container is a 1.3 mm thick polypropylene bottle filled with liquid.

Referring back to the flowchart of Figure 7, the process starts at step 702 where the container is placed in a tray (not shown in Figure 8 for clarity) and then placed on the conveyor belt 802. Note that the tray is optional and may very well be dispensed with. The X-ray scan is then performed. At step 702 the processing module 200 (Figure 1a) acquires the image information 116. In this particular example, the image information 116 is the raw data file output by the X-ray imaging system. The raw data file is then converted at step 706 into distinct image files. This is best shown at Figure 9. The raw data file exported from the X-ray imaging system is converted into three separate image files, namely HI, Lo and class data. The HI file represents the X-ray attenuation at the HI energy level. The Lo file represents the X-ray attenuation at the Lo level. Finally, the class data file is the material classification image that uses colors to illustrate the materials from which the objects shown in the image are made. Class data files are generated by the X-ray imaging system directly and they are normally displayed on the monitor of the X-ray imaging system. In this particular example the class data information is not being used, however one can certainly envisage integrating the class data information to the processing to further refine the results of the security assessment.

The HI and the Lo files are grey level image files showing X-ray energies quantified in a number of different levels. The number of grey levels used can vary depending upon the desired resolution; usually the higher the number of grey levels used the better the precision will be. Tests conducted with images encoded at 256 grey levels (each pixel is represented by an 8 bit value) have demonstrated that the process works, however the error resulting from the loss of information due to the fairly coarse encoding is not negligible. Therefore, grey levels in excess of 256 would be preferred. However, images encoded at less than 256 grey levels can still be uses for some specific applications that require a lesser degree of detection detail.

Referring back to Figure 7, the image files HI and Lo are then subjected to two parallel processing threads, 710 and 712 that determine respectively, the density and effective atomic number. Note that these threads are not independent. The results of the processing thread 712 are supplied to the processing thread 710, such that the density and effective number computations can take into account the X-rays attenuation resulting from the presence of the container. The processing thread 712 starts at step 714 where an edge detection of the container is performed. The purpose is to derive from the information in the HI, Lo image files the location and characteristics of the container. Figures 10 and 11 illustrate the general principle of the edge detection process. Consider in Figure 10 the X-ray image of the container 1000 (Lo image information). Figure 11 shows the grey level profile in the image taken along the imaginary line 1002 drawn across the container 1000. The areas 1004 and 1006 in Figure 11 correspond to areas along the line 1002 that are outside the container 100. The zone 1008 corresponds to the location of the container. It can be observed that the shape of the grey level profile curve matches quite precisely the cross- sectional shape of the container 1000. Figures 12 to 18 provide additional examples. Figure 12 is the HI image of a container and the associated grey level profile curve. Figure 13 shows the grey level curve of the corresponding HI image. In both cases, the curves match the generally rectangular cross- sectional profile of the container. Specifically, the inflection points 1202 and 1204 correspond to the container edges 1206 and 1208, respectively. The flat region 1210 between the inflection points 1202 and 1204 corresponds to the flat top surface 1212 of the container.

Figures 14, 15, 16 and 17 show examples of grey level profiles of containers having rounded features. Figures 14, 16 and 17 clearly show that the grey scale profile matches the rounded cross-sectional contour of the bottle.

Figure 18 is the grey level profile along the container (from top to bottom). Again the profile shows characteristic features of the container. In particular, the area 1802 of the curve corresponds to the bottom portion of the container, the area 1804 shows the top of the container, the area 1806 reveals the notch below the cap and the depression 1808 corresponds to the waist in the middle of the container.

Referring back to Figure 7, the edge detection process 714, therefore performs an analysis of the HI and the Lo image data to detect the edges of the container. Assume for the sake of this example that the container lies horizontally in the tray as it is being scanned by the x-ray machine. Accordingly, the grey level image produced by the x-ray machine will resemble a plan view of the container. The software executed by the processing module 200 which performs the edge detection process applies the following logic:

1. The first step is to locate a portion of the edge. The software searches for detectable grey level transition that occurs in the image as a result of the container wall. Specifically, due to the structure/material of the container wall a well defined grey level transition will show in the image. To facilitate the edge detection process it is possible to provide the operator console 350 with user interface tools that will allow the operator to designate in the X-ray image the general area where the container is located. In this fashion, the software will start the image analysis in an area of the image that is known to contain the image of a container.

Specifically, the user interface on the console 350 is designed such as to display to the operator 130 the X-ray image obtained as a result of the scanning operation. The X-ray image displayed may be derived from the HI image data, the Lo image data or a combination thereof. Once the image is shown to the operator

130, he or she uses a tool to indicate where a container lies. Figure 29 shows an example of such x-ray image where several containers appear at once. Specifically this image shows two containers 3100 and 3102 that are partially on top of each other. This may arise when they have been placed in the tray hastily.

The operator 130 first identifies the container to be processed. Assume that this is container 3100. The operator 3100 then uses a user interface tool to designate the container 3100 to the software. The tool may be any suitable user interface tool such as pointer device such as a mouse or a touch sensitive feature allowing the operator 130 to touch the screen at the area of interest. When the pointer device is activated at the location 3104, which by convention is deemed to correspond generally to the center of the container 3100, the activation will produce location data. The location data identifies an area in the image where the container 3100 resides. The software uses the location data to select the portion of the image data to which the location data points to and starts the image analysis in that area. The selected area corresponds to the location 3104. The software operates with the assumption that the container features that will be identified should have some degree of symmetry about that location. The software scans the image data by moving further away from the location 3104 until a sharp grey level gradient is located that corresponds to a container edge. In principle since the location 3104 is in the center of the container then a container edge should be detected in the image at two places equally spaced from the location 3104.

Another possibility is for the operator to designate with the pointing device specifically the edge of the container that is to be analysed. For instance the operator 130 "clicks" the mouse or touches the screen with his/her finger at the location 3106 that corresponds to the edge of the container 3100.

Yet another possibility is for the operator to perform the designation by "drawing" on the image a zone curtailing the area where the container 3100 is located. For instance the operator

130 can use the pointing device to draw the line 3108 around the container 3100.

With any one of the methods described earlier, the edge detection software receives operator guidance to perform an image analysis and extract from the image one or more characterizing features of the container 3100. Figure 30 provides a flowchart that that summarizes the above process. At step 3200 the image of the one or more containers is shown on the console 350 of the operator. At step 3202 the operator uses a suitable user interface tool to designate the container to be analyzed. As indicated earlier, the user interface tool may be a pointing device, among others. At step 3204 information about the location in the image where the container is located is communicated to the processing module 200 such that the container analysis can be performed.

2. Referring back to figure 29, the next step of the process is to track the outline of the container 3100. As the software has identified a portion of the container's edge, the software logic then starts tracking that edge. The tracking logic tracks the sharp grey level gradient in the image to follow the container edge. In doing so, the tracking logic uses a set of assumptions; otherwise it may stray, in particular at areas where two or more container edges meet. This is shown in the area 3110 where the edges of the two different containers 3100 and 3102 intersect each other. If the tracking software is moving along the edge 3112 (in the direction shown by the arrow) it will eventually encounter the location 3114 where the edges of the two containers 3100 and 3102 cross each other. At that location, the edge tracking software has at least three different edges that it can track, namely edge portion 3116, 3118 and 3120, while only one solution (edge 3120 is valid). To avoid straying along the non-valid solutions (edges 31 18 and 3116) one of the assumptions is that the edge of the container has no sharp edges or turns. A sharp edge or turn is defined by a radius value, which is a parameter that can be permanently set or made adjustable. Accordingly, when the tracking software reaches the location 3114 the solutions that correspond to edge portions 3118 and 3116 are rejected because they involve a sharp departure from the existing course (edge portion 3112). Then only solution 3120 remains as valid.

Other assumptions can also be used. One is the container symmetry attribute. Most of the containers are symmetrical about one or more axes. When one side of the container wall has been tracked the other side should in principle be a mirror image of the first side, accordingly only solutions that correspond to that mirror image path would be retained. Another assumption is the maximal or minimal dimension of the container or of its constituent parts. For instance, it is known that containers typically have dimensions that do not exceed a certain limit that is considered to be a maximal value. Accordingly if an edge length extends beyond those limits the detection process may be considered invalid. Similarly, minimal dimensions can also be taken into consideration. If an edge length is below a value that is considered to be a minimum for a container height or width, the detection process may be considered invalid.

When the tracking logic has completed the identification of the container edge, then the software performs a validation on the basis of the overall container shape defined. Specifically the software will compute certain geometric features or properties of the container and determine if they fall into acceptable acceptance windows. Examples of such geometric features include:

The height of the container. Usually, most containers would have a height that would fall in a certain range, say from 3 inches up to 18 inches. Any container height dimension outside that range should be suspect.

The width of the container. As in the case with the container height, the container width usually falls in a certain range, for instance between 1 inch and 6 inches. Containers having a width outside that range would also be suspect.

The ratio height/width which is considered to be valid only if the value computed falls in a predetermined range.

A volume prediction of the container. On the basis of the container outline one can predict what the internal volume could be. While to perform an accurate volume computation the actual thickness (3^rd dimension) of the container is required, that dimension can be assumed in order to provide volume estimation. The container thickness would normally be in the range of 1 inch to 6 inches. This allows providing a volume estimation that defines a window allowing rejecting solutions associated with volume values that are outside the window.

4. When the container validation process has been completed, the outline of the container can be emphasised to the operator 130, as a final "sanity check". This step is identified at block 716 of Figure

7. Specifically, the processing module 200 issues commands to the display such that the display visually enhances a portion of the image where the container is located. This makes the container more visible with relation to other objects in the x-ray image. Examples image enhancements include:

a. Colouring or otherwise highlighting the areas of the image that correspond to the portions where the edge has been identified; b. Colouring or otherwise highlighting the container in its entirety; c. De-emphasising the image except the areas where the container lies. This technique does not change the pixels of the x-ray image in the region of the container but changes all the pixels that surround the container image such as to make the container more visible.

The highlighting process uses the edge detection data obtained by the edge detection software as a result of the x-ray image analysis. The edge detection data defines in the x-ray image the areas where an edge has been identified. The highlighting process then uses this information to manipulate the x-ray image pixels such that the container stands out with relation to its surroundings.

If the edge identification has been done correctly the operator 130 would see the container 3100 highlighted. The operator 130 can then apply human judgment on the results. If the edge tracking operation is correct then the results can be accepted and the processing allowed continuing. Otherwise, if the operator 130 sees on the screen a highlighted shape that does not correspond to a container then he/she aborts the operation.

At step 718 the edge detection data obtained by the edge detection software is processed to extract one or preferably more that one characteristics of the container. Examples of characteristics include:

• The height of the container

• The maximal transverse dimension of the container;

• Wall thickness

• Generalized geometric shapes that are found in the container.

^■ The geometric shape identification is a software processing of the container image to try to identify in that image geometric features or shapes that can be used to characterize the container. For example, the software may look at the main body of the container (disregard the neck portion) to determine if the container falls in any one of a set of predefined geometric shapes. Examples of geometric shapes include:

- rectangular container; - square container;

- upwardly tapered container;

- downwardly tapered container.

At step 720 the knowledge bank 400 is searched on the basis of the characteristics of the container identified previously. The knowledge bank 400 is designed as a database that has a number of entries, each entry being associated with a product containing a liquid on which the qualification process is being run. Each entry includes two different classes of information. The first class is characterization information about the product. The characterization information includes one or more features of the container in which the liquid is stored. Examples of features include:

• Container height;

• Wall thickness; • The transverse dimension of the container;

• Geometric shapes found in the container or the set of predefined geometric shapes to which the container belongs;

• Generic container templates;

• Physical parameters of the container; • Chemical parameters of the container such as the material from which the container is made;

• Height off belt;

• Path length calculation parameters (see description later for path length calculation); • Contour details. In addition the characterization information may also include information about the liquid (other than its response to X-rays), such as the color of the liquid, aroma or visual texture, among others. Under the current example, the characterization information includes solely information about the container.

The second class includes the responses of liquids in the containers having the characteristics stored in the knowledge bank 400. In the specific example of implementation discussed here, the penetration radiation used to obtain a response from the liquid is X-rays, however, other types of electromagnetic radiation can be used without departing from the spirit of the invention. The information stored in the knowledge bank 400 that characterizes the response to the liquid to X-rays includes density and effective atomic numbers for each liquid. This is useful for applications where the X-ray imaging system only provides an image output obtained on the basis of photons that have passed straight through the sample. For X-ray imaging systems where the image output also takes into account scattering/diffraction then the knowledge bank 400 can also include the diffraction/scattering signature of the liquid.

Figure 21 shows a graph of the diffraction/scattering signature for a number of different materials, in particular propanol, acetone, methanol and hydrogen peroxide. The visible texture of the scattering/diffraction signature changes with the density of the materials and constitutes a feature that can be used to qualify the different materials.

Accordingly, the knowledge bank can be augmented by storing in association with each entry the diffraction/scattering signature of the liquid. The diffraction/scattering signature can be in the form of an image file or under any other suitable representation that would allow a comparison to be made with the diffraction/signature of a material that is being scanned such as to determine if both signatures match.

The diffraction/scattering signature can be used alone to determine if a liquid matches an entry in the knowledge bank, but preferably it can be used in conjunction with the other elements of information that define the response of the liquid to X-rays, such as density and effective atomic number.

Typically, a neural network would be used to determine if the observed diffraction/scattering signature of a liquid matches anyone of the signatures stored in the knowledge bank 400.

Assuming now that the knowledge bank search is successful and a unique and unmistakable match is found on the basis of the product characterization information provided, then the search will extract the nominal container height (step 722) and the nominal wall thickness (724) of the container from the knowledge bank 400. The read container height and wall thickness are communicated to a processing block 726 which computes the X-ray path length of the container that is being scanned by the X-ray apparatus. This processing block will be discussed in greater detail later.

On the other hand, if no match is found in the knowledge bank 400, then the processing continues at step 740 where height estimation is performed for the container. In this case, the container height data generated during the container characterization step 718 is read and that information is used as container height information. Similarly, at step 738 an estimate of the container wall thickness is produced from the edge detection data obtained at the edge detection process. Both the estimated edge thickness and container height are then supplied to the block 726 which performs the X-ray path length computation. The x-ray path length analysis will be described in greater detail later.

The processing thread 710 that runs in parallel with the processing thread 712 performs image processing in order to identify the response of the liquid in the container that is being scanned to X-rays. The first step of the process (step 728) is to locate in the HI and Lo images the tray in which the container is placed for the scanning process. Since the tray signature is known, known image processing techniques can be used to identify the location of the tray in the images and its orientation. The tray signature resides in the memory 302 of the processing module 200. As indicated earlier, the use of a tray is optional and can be dispensed with. In such case the product would be placed directly on the conveyor belt and scanned with X-rays as such.

The flowchart of Figure 23 shows in greater detail the process for identifying the location and the orientation of the tray in the HI and Lo images. To make the identification of the tray simpler, the tray is provided with a marker that is highly visible to X-rays. This may be a piece of metal that will highly attenuate X-rays, which is located at a known position in the tray. Therefore, the detection of the tray position in the image starts by determining where in the HI and the Lo images that marker can be found. For easier identification, the maker can be of an easily recognizable shape unlikely to be confused with other objects placed in the tray during the X-ray scanning process.

At step 2400 the process receives the HI and the Lo image information. The HI image is scanned at 2402 to locate the marker. The image is searched using any well known image scanning techniques on the basis of the marker signature at the HI energy level extracted from the memory 302 of the processing module 200. If the marker is found, its coordinates are recorded. The same process is repeated at step 2404, this time on the Lo image. The coordinates of the marker are also generated.

At step 2406 the tray position and contour is determined by processing both sets of marker coordinates. Since the position of the marker in the tray is known and the shape of the tray is also known, then step 2406 will determine the location of the tray in the HI and Lo images, its contour and its orientation. The process outputs at step 2408 data that defines the location of the tray, its contour and its orientation in both images. The location, contour and orientation should be such as to allow identifying in each image the pixels "overlaid" by the tray, in other words the pixels whose grey levels include the contribution of the tray to the overall X-ray attenuation. A somewhat similar operation is performed at step 730 on the HI and Lo images to remove the contribution from the belt 802 (Figure 8). The belt 802 attenuates to a known degree the X-ray radiation and step 730 compensates the images accordingly. This is done by modifying the grey levels of the pixels in the HI and the Lo images to produce a compensated image that will show a lesser degree of attenuation. The detailed process for removing the contribution of the belt 802 is shown by the flowchart of Figure 22. Step 2300 receives the HI and the Lo images information. At step 2302 the signature of the belt 802 for the HI energy level is read from the memory 302 of the processing module 200. A search is made in the image such as align or "overlay" the read signature with the signature appearing in the image. A similar operation is performed at step 2304 for the Lo image. Steps 2306 and 2308 compensate the HI and Lo images such as to remove the effect of the belt 802. The compensation is done only in the areas of the HI and Lo images that are encompassed within the belt signature, hence the areas where the gray levels convey attenuation information due to the belt 802 presence (the attenuation due to the belt 802 is stored in the memory 302). The compensation is done by changing the grey levels to remove the attenuation due to the belt. Since the belt 802 is a relatively uniform structure, the compensation that is made on the HI and the Lo images consists of reducing the grey level intensity in each pixel by a value that corresponds to the attenuation caused by the belt 802. Accordingly, steps 2306 and 2308 produce synthetic HI and Lo images in which the effect of the belt 802 is removed.

The HI and Lo synthetic images are processed at step 732 (Figure 7) to remove the contribution of the tray. The details of the tray removal are shown in the flowchart of Figure 24. The HI and the Lo synthetic images as well as the data that defines the location of the tray (obtained from the process at Figure 23) are received by the process at step 2500. Step 2502 processes the data that defines the location of the tray for the HI and the Lo levels in conjunction with the tray X-ray signature at the HI and Lo levels. The X-ray signature for the HI and the Lo levels is extracted from the memory 302. The processing at step 2502 modifies the signature extracted from the memory 302 such as to shift it to the current tray location. In other words, the X-ray signature of the tray that is stored in the memory 302 corresponds to a certain reference tray location. To be able to use this signature in cases where the tray is in a position other than the reference position, then the signature must be manipulated such as to displace the grey level features that define the signature to the positions where the tray is actually located. Step 2502 performs this operation by using any suitable image processing techniques that translate and/ or rotate the pixels that convey the X-ray attenuation caused by the tray in the actual tray position that was previously determined. This produces a real tray signature, for both HI and Lo energy levels that can be used subsequently to compensate the HI and the Lo images for the presence of the tray.

Step 2504 performs the tray removal operation. The process at step 2504 receives the synthetic HI and Lo images (compensated for the belt) and also the real tray signature generated earlier. The real tray signature for each energy level is "subtracted" from the corresponding synthetic image such as to remove from the synthetic image the X-ray attenuation information resulting from the tray.

Step 2506 outputs the HI and Lo synthetic images that have been cleaned to remove the effects of the belt and the tray.

Referring back to Figure 7, step 734 further modifies the HI and the Lo images received from the process at step 732 to remove from the image information the attenuation due to the container wall. The material from which the container is made will determine the extent to which the container wall removal is critical.

For glass materials it is necessary to remove their contribution since glass materials tend to attenuate X-ray significantly as in practice they are quite thick.

On the other hand, when the container is made of plastic that attenuates X-rays to a much lesser degree, the compensation of the image is not absolutely required. The same also holds true for thin walled metallic containers, such as aluminum beverage cans. The step 734 receives the HI and the Lo images compensated for the presence of the belt and of the tray, information that approximates the wall thickness of the container (the approximation will be described later), real wall thickness information and material of container extracted form the knowledge bank 400 as output at step 724 (if a match in the knowledge bank 400 has been found) and the coordinates of the container contour from the edge detection process 714. If the product (container + liquid) that is being scanned has been accurately recognized at step 720 (a match exists in the knowledge bank 400), then the approximation of the wall thickness is not required. The wall thickness approximation is used only if the product recognition process at step 720 is uncertain or has failed. The flowchart at Figure 26 illustrates in greater detail the process for compensating the HI and the Lo images for the attenuation resulting from the container walls.

Step 2800 is the start of the process. That step receives the following information:

1. HI and Lo images compensated for the attenuation by the belt and tray; 2. Coordinates of the container contour. This information is received from the processing at step 714 (edge detection). This information specifies the outline of the container and defines the area of the HI and Lo images that will need to be compensated to remove the effect of the container wall. 3. The estimated wall thickness;

4. The real wall thickness and the material from which the container is made (information obtained from step 724, if available).

If only a wall thickness estimation is available (no real wall thickness information found) then the process proceeds at step 2802 that computes the attenuation brought by the container. Since at that point no knowledge exists about the material from which the container is made, the process at step 2802 assumes that the material is glass, which in most practical cases would be the worst case scenario (the greatest degree of attenuation). The step 2802, therefore computes the attenuation that the glass material of the estimated thickness will create such that the HI and Lo images can be compensated accordingly. The process performed at step 2502 is a computational step that uses the following algorithm for HI energy level image:

Bottle _Contr. Hi = MAX_GS [[I - e ^s'°" ^{L r} Jx 100%j

and the following algorithm for the Lo energy level image

Bottle _Contr. Lo

|_x 1_OO%

Where:

• Bottle Contr. Hi is the container wall attenuation at the Hi energy level expressed in percentage;

• Bottle _Contr. Lo is the container wall attenuation at the Lo energy level expressed in percentage;

• MAX_GS is the Maximum Gray Scale (actual value of the background or input energy) • p_glass = 2.469 g/ cm'

• Zeff__glass = 12- 12

• am, bm, a_Lo and b_Lo are constants that are dependent on the particular X-ray imaging system used for the scanning. The values of those constants are obtained during the calibration phase of the machine and they are stored in the memory 302 of the processing module 200.

The glass density {p_glass) and effective atomic number (Z_{eff glass}) are stored in the memory 302 of the processing module 200. Alternatively, the glass density and effective atomic number could be stored in the knowledge bank 400, as a parameter of container. In this fashion, it could be possible to provide for each glass container specific density and effective atomic numbers that match well the specific container material. This could be useful if it is expected to find in use different containers made of different glass compositions such that the density and the effective atomic numbers are not all the same across the glass containers population.

Therefore, the step 2802 outputs the attenuation in the X-ray images at the HI and at the Lo energy levels that the glass container produces. The output is supplied to step 2804 that uses this information to compensate the HI and the Lo images accordingly. Step 2804 will be described in greater detail later.

Assuming now that instead of estimated wall thickness information, real wall thickness information is available, then the step 2802 is performed only if the material from which the container is made is glass. Specifically, at decision step 2808 the material from which the container is made is verified. The material from which the container is made is stored in the knowledge bank 400. If the material is glass then step 2802 described above is performed. On the other hand, if the material is plastic then the processing goes directly to the output 2810. In other words, if the container is made of plastic, no image compensation is performed. The reason for bypassing the image compensation is that a plastic introduces a negligible degree of X-ray attenuation, therefore the HI and the Lo images do not need to be compensated.

Step 2804 receives the X-ray attenuation introduced by the glass container for the HI and the Lo energy levels. Also, step 2804 receives the HI and the Lo images compensated for the belt and the tray and the container contour information. Step 2804 performs image processing to remove the attenuation introduced by the container in the area defined by the container contour information. The pixels in the area defined by the container contour information are modified such that their values no longer reflect the contribution of the attenuation introduced by the glass material. Step 2804 therefore outputs at step 2808 HI and Lo images that have been compensated for the influence on the X-rays of the belt, the tray and the container wall. Therefore, the HI and the Lo images now provide attenuation information of the liquid and allow computing parameters of the liquid.

For clarity, it should be mentioned that the compensation for the container wall has essentially the effect to "remove" the container wall in the x-ray image within the contour of the container. In other words the portion of the wall that is generally parallel to the x-ray image plane is being erased. The wall portions of the container that are generally perpendicular to the x-ray image plane and which would define its contour still remain in the image.

Since the HI and Lo X-ray images are two dimensional, the path length calculation, in one non-limiting example of implementation, is an indirect mathematical operation based on a combination of trigonometry operations and shape recognition algorithms. Knowing the exact physical characteristics of the X-ray imaging system, it is possible to calculate the height of the liquid container, and therefore the path lengths followed by the X-ray beams, by using the position of the container on the conveyor belt 802 with respect to the fixed reference points of the X-ray scanner itself. As these reference points remain identical from one scan to the next, the path length calculation is not affected by the random position of the containers in the plastic tray. Should there be bubbles in the liquid under test, their presence can be filtered out by either appropriate filtering algorithms or by considering the bubble physical characteristics in order to remove their contribution from the liquid.

Figure 28 illustrates the path length determination process. Figure 28 is a cross-section of the X-ray imaging system 3000 showing the belt 802 on which the container 3002 is placed. For clarity, the belt 802 moves the container 3002 through the x-ray imaging system 3000 in a direction that is perpendicular to the sheet. This X-ray imaging system 3000 has a radiation source 3004 that is located below the belt 802 and also an L-shaped set of detectors that has a vertical array 3006 and a horizontal array 3008. The array 3006 is shown arbitrarily as having 12 detectors, (3006i 3006i₂) and the array 3008 has 12 detectors (3008i 3008i₂) as well. Note that in practice, X-ray imaging systems have a much higher numbers of detectors in order to provide a suitable image resolution.

The position of the source 3004 is well known and fixed. In addition, the geometry of the detector arrays 3006 and 3008 is such that it is possible to map portions of the x-ray image (Lo and Hi) to individual detectors of the arrays 3006 and 3008. In other words, it is possible to tell for a certain portion of the image, which ones of the detectors produced that portion of the image. Figure 31 provides more details in this regard. Figure 31 shows a simulated x-ray image of a body of liquid 3300, shaped in the form of a container. The image is obtained as a result of a movement of the container 3300 by the belt 802 with relation to the detector arrays 3006 and 3008. Therefore, individual detectors of the arrays 3006, 3008 produce individual bands in the image. The image bands are shown in Figure 31 and for clarity numbered with the corresponding detector reference numerals.

Referring back to Figure 28 assume for the sake of this example that the X-ray source 3004 is turned on and generates X-ray beams that are directed through the container 3002. While there are many beams passing through the container 3002, consider only two of them, namely the beam 3010 and the beam 3012 that intersect the top and bottom edges of the container 3002. The beam 3010 will reach the detector 3008₂ while the beam 3012 will reach the detector 3008₇. By analyzing the image it is possible to determine which detectors of the arrays 3006, 3008 received the beams 3010 and 3012. Specifically, the features of the container 3002 through which the beams 3010 and 3012 pass are first located in the image and their respective positions in the image noted. In particular the processing module processed the x-ray image information to locate the top and the bottom edges of the container 3002 and once those features have been identified, their position in the image is recorded. Since the image positions are mapped to corresponding detectors of the arrays 3006 and 3008, it is possible to derive which ones of the detectors in the arrays 3006, 3008 received the beams 3010 and 3012. On the basis of the position of those features in the image, the detectors are identified. Once the identity of the detectors has bee found, both lengths L1 and L2 can be trigonometrical Iy calculated using angles alpha and beta. Finally, the path length H can be simply derived by the formula H = (Z₁ - L₂) tan α . In this example, Η would be the height of the body of liquid held in the container.

The above process works well for containers that are generally rectangular in shape. For containers that are rounded, such as cylindrical shapes for instance, the following cylinder parametric equation can be used:

$(z , θ ) = u (z , θ

zk

Where u(z,θ) will be adjusted according to every individual shape of container.

This equation is a known ray casting formula that this is used that is used to calculate object interceptions in 3d space.

Once the path length through the liquid has been computed at step 726, the process continues at step 738 where the density and the effective atomic number of the liquid are computed. The process will be described in greater detail in conjunction with the flowchart on Figure 25. The process starts at step 2700. The information that is used to perform the various computations includes:

1. The HI and the Lo images as output from the processing at step 734 (the contribution of the belt, the tray and the container wall have been removed).

2. HI, Lo (bgnd) which are the images compensated for the presence of the belt. 3. Coordinates in the HI and the Lo images that are within the boundary of the liquid body in the container, where the density and the effective atomic number will be assessed. Typically, to obtain a better accuracy the density and the effective atomic number will be assessed at more than one location. 4. The path length (height of the liquid body) at the coordinates specified at 3. Both the coordinates and the path length values are obtained from the processing at step 726.

Step 2702 receives the HI and the Lo image information as well as the coordinates where the density and effective atomic numbers will be assessed. The processing at step 2702 will essentially extract from the HI and the Lo images the grey level values at each of the coordinates. If each coordinate is larger than a single pixel, say it encompasses several pixels in the HI and the Lo images, then the grey level extraction could include averaging the grey level values encompassed within each coordinate area. Therefore, the processing at step 2702 outputs two sets of grey level values, the first set extracted from the HI image and the second extracted from the Lo image.

The two sets of grey level values are handled by the process at step 2704. That step computes the X-ray attenuation coefficients for each of the coordinates. So, in addition to the grey level values sets, the process at step 2704 also receives the path length values from step 2700, where each path length value is associated to a given coordinate. As mentioned above, a given path length value is essentially the thickness of the body of fluid through which the X-rays pass. Note that the path length is not necessarily the same for all the coordinates.

The processing at step 2704 applies the following algorithm for computing the attenuation coefficient for the various coordinates at the HI energy level:

Where: 1. μ_Hι _ __ is the attenuation coefficient at the HI energy level for the coordinates 1 n; 2. XPL_{1 n} is the path length at coordinates 1..n for the HI energy level;

3. I_mbgnd) at coordinates 1 n for the HI energy level;

4. I_mflnal)ι are the grey level values at coordinates 1 n for the HI energy level.

A similar equation is used to compute the attenuation coefficients at the various coordinates at the Lo energy level.

Where:

1. μ_Lθj is the attenuation coefficient at the HI energy level for the coordinates 1 n;

2. XPL_{1 n} is the path length at coordinates 1..n for the HI energy level;

3. I_Lo{bgnd) at coordinates 1 n for the HI energy level; 4. I_Lo(flnal)i are the grey level values at coordinates 1 n for the HI energy level.

The processing continues at steps 2706 and 2708 that compute the density of the liquid and the effective atomic number of the liquid at the respective coordinates. The density computation at step 2706 receives as input the X-ray attenuation coefficients, and machine calibration constants. Specifically, the density computation is effected by using the following algorithm:

(^aH_i ^{X b}L_o ) - (^a _Lo * b_Hi )

Where: 1. p_{x n} is the density of the liquid at the coordinates 1 n. Note that the density computation uses grey level information from both the HI and the Lo X-ray images;

2. a_Hl,a_Lo,b_Hl,b_Lo are X-ray imaging system constants. These constants are stored in the memory 302 of the processing module

200;

3. μ_Lθι is the attenuation coefficient at the HI energy level for the coordinates 1 n;

4. μ_Hh is the attenuation coefficient at the HI energy level for the coordinates 1 n.

Step 2708 computes the effective atomic number at the coordinates 1 n. This computation also makes use of the attenuation coefficients computed earlier for the HI and Lo energy levels and also uses the X-ray imaging system constants. Specifically, the following algorithm can be used to perform the computation:

J l \- b Hrri X ^ ULjo₁ / ) + W \ Lro X r U^*H „iχ - / )

Where: 1. Z_efA is the effective atomic number of the liquid measured at the coordinates 1 n; 2. a_Hl,a_Lo,b_Hl,b_Lo are X-ray imaging system constants. These constants are stored in the memory 302 of the processing module 200; 3. μ_Lθi is the attenuation coefficient at the HI energy level for the coordinates 1 n; 4. μ_Hh is the attenuation coefficient at the HI energy level for the coordinates 1 n. Finally, step 2710 outputs the density and the effective atomic number for each or the 1 n coordinates.

Referring back to the flowchart of Figure 7, the computation of the density and the effective atomic number at step 738 leads to step 741 where the liquid is qualified. This determination will be described in greater detail in connection with the flowchart on Figure 27. The process starts at step 2900. The processing at step 2900 receives the following information:

1. Z_effι is the effective atomic number of the liquid measured at the coordinates 1 n, as computed at step 738.

2. P_{1 n} is the density of the liquid at the coordinates 1 n, also as computed at step 738.

3. (Ap,A_Zeff )_{s s} which is the system error or standard error generated by the system itself.

Step 2902 computes an average density value for the liquid and also the standard deviation. Specifically, the average density is determined by:

1 "

— > r average / i r i n ,₌₁

Where:

1. p_avemge is the average density of the liquid.

Step 2902 also computes the standard deviation Ap of ^^{1 ■■■«} with relation

fofi '^average _ y_he standard deviation is expressed by Ap = σ(p₁ p₂,p₃,....p_n).

Similarly, step 2904 computes the average effective atomic number along with the standard deviation. Specifically, the average effective atomic number is determined by: 1 « z eff -average =-Y Z_ιz eff, n _j=l

Where:

1. Z_eff__avemge is the average effective atomic number of the liquid.

Step also computes the standard deviation AZ_eff of Z_{efh n} with relation to Z_eff__averaσe . The standard deviation is expressed by

Steps 2902 and 2904 output to step 2906, which is the next step in the processing thread, _Pavemge , Ap , Z_eff__avemge , AZ_eJf , A_Psys and AZ_eff__sys .

Step 2906 generates density and effective atomic number lookup values to query the knowledge bank 400. More specifically, the processing at step 2906 computes an effective atomic number lookup window to select potential matching candidates in the knowledge bank 400. This lookup window is mathematically defined as:

[Ztf-LU ] = ^Z eff -average ± Σ (^ eff + ^eff-sys )

The lookup window is defined by a low effective atomic number value Z_eff__LU__Low and by a high effective atomic number value Z_ef__w__Hi .

The density lookup window is mathematically defined as:

^P_Sys) • ^Tne lookup window is defined by a low effective density value ρ_w__!ow and by a high effective atomic number value p_Lυ__hiσh ■

The knowledge bank 400 is queried on the basis of the density and effective atomic number lookup windows. The selection process is such that a product in the knowledge bank 400 for which an effective atomic number and a density value fall in the respective lookup windows are retained as potential candidates. The list of candidates is then processed at step 2910. More specifically, the processing at step 2910 tries to determine to what degree anyone of the candidates matches the characteristics of the product scanned by the X-ray imaging system.

A "candidate" is essentially an entry in the knowledge bank 400. Most of those entries are associated a liquid and may represent a particular condition of the liquid corresponding to the degree of quality, performance state, potency, taste, mouth felt texture or aroma, among others. As discussed earlier, each candidate that is selected at step 2908 is defined by certain characterizing information, such as density, effective atomic number and container characterization among others. This characterizing information is affected by the particular condition of the liquid. In particular when the degree of quality, performance state, potency, taste, mouth felt texture or aroma changes, so does the characterizing information.

This characterizing information is then compared with the product characterization effected as a result of the X-ray scan to determine if a match can be found. If a match exists, this means that in all likelihood the liquid in the container that was scanned by the X-ray imaging system has the same degree of quality, performance state, potency, taste, mouth felt texture or aroma than the candidate in the knowledge bank 400.

The process for determining if the product characterization matches any one of the candidates involves comparing the product characterization with the information that characterizes each candidate. In a specific and non-limiting example of implementation, a first comparison is made between the density (as computed from the X-ray images) of the scanned product and the density information for each one of the candidates. The candidate that matches best the density of the scanned product is retained. Next, the effective atomic number (as computed from the X-ray images) of the product is compared to the effective atomic number of the candidate that was retained. If a match is found then the final step of the assessment includes comparing the container features identified from the X-ray images with the container features stored for that candidate in the knowledge bank 400. If a match is found then the system concludes that the product that was scanned by the X-ray imaging system corresponds to the candidate.

After the processing that tries to match the scanned product with the candidates in the knowledge bank 400 has been completed the results are shown to the operator. If a match exists, this means that the product that is being scanned corresponds to the state of the product reflected by the candidate. Specific examples of products along with the type of determination that the X-ray scanning can do are discussed below:

1. Liquid medicinal product, such as cough syrup, eye drops, ear drops, nasal spray and a wide range of liquid pharmaceutical products for aural ingestion or syringe injection. Candidates can be provided in the knowledge bankbank that correspond to different: a. Tastes of products for aural ingestion; b. Aroma of products; c. Concentration of active ingredients. d. A particular quality standard e. Products contaminated with certain substances, such as bacteria, or chemical contaminants

Since those attributes are affected by the composition of the liquid, the X- ray image of the liquid will also be different and can be used to distinguish between them. Accordingly, the X-ray inspection, in this fashion can be used to determine for a given product if the product is matches a candidate that corresponds to a predetermined quality standard (detect tempering for example), or determine the smell, taste or concentration of active ingredients. 2. Personal care products for direct application to the body, such as shampoo, hair treatment preparations (hair gel, hair spray, conditioner, hair straightener, anti-fizz, hair dye, etc), UV protection preparations, skin cream and cosmetics, among others. Candidates can be provided in the knowledge bank that correspond to different: a. Concentration of active ingredients. b. Potency c. A particular quality standard d. Products contaminated with certain substances, such as bacteria, or chemical contaminants

Since those attributes are affected by the composition of the liquid, the X- ray image of the liquid will also be different and can be used to distinguish between them. Accordingly, the X-ray inspection, in this fashion can be used to determine for a given product if the product is matches a candidate that corresponds to a predetermined quality standard, or determine the smell, concentration or potency of active ingredients.

3. Industrial products such as, solvents, degreasers, cleaners, paints, adhesives, insecticides and fertilizers among others. Candidates can be provided in the knowledge bank that correspond to different: a. Concentration of active ingredients. b. Potency c. A particular product characteristic d. A particular quality standard e. Products contaminated with chemical substances

Since those attributes are affected by the composition of the liquid, the X- ray image of the liquid will also be different and can be used to distinguish between them. Accordingly, the X-ray inspection, in this fashion can be used to determine for a given product if the product is matches a candidate that corresponds to a predetermined quality standard, its active ingredients or determine its potency. A specific example is the case of gasoline where the X-ray image allows distinguishing between different additives and additives concentration. In other words, the additives in the gasoline that change its octane rating, and its cleaning properties, among others affect the X-ray image of the liquid. Accordingly, by providing in the knowledge bank candidates that correspond to X-ray images of gasoline having different additives/concentrations of additives, can be used to determine what type of additive/concentration of additive a gasoline sample contains.

4. Edible products, such as milk, drinks (spirits, soft drinks, water and juice), syrup, oil, food extracts and cream, among others. Candidates can be provided in the knowledge bank that correspond to different: a. Tastes. b. Whether the product is safe for consumption or unsafe (different candidates corresponding to different possible contaminations, such as contamination by a chemical product, contamination by bacteria, etc). c. A particular quality standard d. A particular concentration of an ingredient, such as degree of alcohol, degree of sugar e. A particular mouth felt texture of the product f. Products contaminated with certain substances, such as bacteria, or chemical contaminants

Since those attributes are affected by the composition/state of the edible liquid, the X-ray image of the liquid will also be different and can be used to distinguish between them. Accordingly, the X-ray inspection, in this fashion can be used to determine for a given product if the product is matches a candidate that corresponds to a predetermined quality standard, or determine its taste or safety. A specific example is: • The case of food that can be contaminated with different types of bacteria. The knowledge bank 400 stores different candidates which are the X-ray images of the food contaminated by bacteria A, bacteria B, bacteria C, etc. if a match is found between the tested sample and anyone of the candidates, this means that the sample is contaminated.

• The case of spirits that changes its chemical composition as it ages and with alcohol concentration. Accordingly, the knowledge bank 400 can contain candidates that correspond to different degrees of aging, thus allowing classifying the wine in different quality categories depending on the age, alcohol concentration, etc.

5. Household products, such as detergents and cleaners. Candidates can be provided in the knowledge bank that correspond to different: a. Degree of potency; b. Concentration or presence of particular chemical products. c. A particular quality standard d. Products contaminated with certain chemical substances

Since those attributes are affected by the composition of the liquid, the X- ray image of the liquid will also be different and can be used to distinguish between them. Accordingly, the X-ray inspection, in this fashion can be used to determine for a given product if the product is matches a candidate that corresponds to a predetermined quality standard, or determine the degree of potency or concentration of particular chemical product. 6. Petroleum products such as crude oil, gasoline, liquefied propane or liquefied natural gas. Candidates can be provided in the knowledge bank that correspond to: a. A particular grade; b. Concentration or presence of particular chemical products, such as additives; c. Presence or concentration of contaminants

In the case of crude oil, the method can be used to determine the its quality or grade. In the case of gasoline, the method can be used to determine the presence or concentration of certain additives such as additives used to increase the octane number, additives used to clean the fuel system of an internal combustion engine and additives used to prevent deposits in the fuel system of an internal combustion engine, among others. In the case of liquefied propane or natural gas the method can be used to distinguish among different density grades.

The flow chart in Figure 19a illustrates another example of implementation of the invention where the characterization of the product is made by reference to the Universal Product Code (UPC) bar code that appears on the product. Nearly all the products that are sold today in the market use a bar code system that facilitate checkout procedures and also help tracking inventories. UPC barcodes originate with the Uniform Product Council that manages the allocation of the barcodes to different manufacturers. A typical bar code that is applied to the product package has generally two components; one is the machine readable part and the other the human readable part. The machine readable part appears as a series of bars while the human readable part is a series of digits appearing below the machine readable bars. A typical UPC bar code has a part that identifies the manufacturer and another part that identifies the actual product within that manufacturer's product line. Since UPC barcodes are used primarily for payment and inventory control purposes they are unique for each product. Accordingly, the UPC barcode constitutes a unique identifier for almost every product that is found in market today. Thθ process at the flowchart of Figure 19a starts at step 1900 where the bar code of the product (container + liquid) that is to be qualified is read. This operation is performed by using a standard bar code reader of a type known in the art. The information obtained as a result of the reading operation is then used to search a knowledge bank 1902 and usually will be sufficient to uniquely identify the product among the plurality of products stored in the knowledge bank 1902.

The structure of the knowledge bank is shown in Figure 19b. The information in the knowledge bank 1908 can be organized as a table. Each entry of the table is associated with a certain liquid product. The examples shown relate to medicinal products having different concentration. Specifically, the first row relates to a product having a concentration of active ingredient A (low concentration), the second row relates to a product having a concentration of active ingredient B (intermediate concentration), the third row relates to a product having a concentration of active ingredient C(high concentration) and the fourth row relates to a product having a concentration of active ingredient D (super high concentration). The four products are otherwise identical, in other words only the concentration of active ingredients changes. Each entry of the knowledge bank is identified by the UPC bar code applied on the product by its manufacturer. Since bar codes are unique, that entry conveniently constitutes a key on the basis of which the knowledge bank 1908 can be searched. In the specific example of implementation shown at Figure 19b, the knowledge bank 1908 has six data fields for each entry. The data fields are as follows:

1) The UPC bar code that is expressed in any suitable format.

2) The density of the liquid. The density may be the real density (as measured by standard techniques) or the density as assessed as a result of an X-ray scan, or both. In this example, only one density value is shown assuming that the real density and the one obtained as a result of an X-ray scan are the same.

3) The effective atomic number of the liquid as measured by X-rays. 4) Optionally, container features, such as visual characteristics that distinguish the container. Examples include the dimensions of the container (height and transverse dimensions), type of container (screw cap, can or other), general container shape (cylindrical, rectangular cross-section, etc), and unique visual features such as ridges or projections on the walls, among many others. One possibility is to store in this data field a 3d image of the product that would show the product from different sides. With the appropriate image viewer, the operator can, therefore be provided with a complete image of the product that was found to match the barcode search operation. The container features also include information on the wall thickness and the material from which the wall is made such as to allow compensating the X- ray image data for the attenuation by the container walls.

5) The diffraction/back scatter signature.

6) The product identification in terms of active ingredient concentration.

It should be recognized that the structure of the knowledge bank 1908 can include more information about liquid products or less information, without departing from the spirit of the invention.

Referring back to Figure 19a, step 1904 determines the response of the liquid in the container to penetrating radiation, X-rays in particular. This can be done in the same way as described previously under the first example of implementation. In short, step 1904 will derive parameters of the liquid from the X-ray scan, such as density, effective atomic number, and diffraction/back scatter signature, among others. Optionally, this can be done by referring or using information stored in the knowledge bank 1908, such as for example the thickness of the container wall and the material from which the container wall is made such as to perform X-ray image compensation for the attenuation of the X-rays by the container wall.

Next, the comparison step 1906 qualifies the liquid product. This is done by comparing parameters of the liquid product as extracted from the knowledge bank to those measured by the X-ray scan. Assume for the sake of this example, that at step 1900 the bar code on the container was correctly read and the search step 1902 identified an entry in the knowledge bank on the basis of the bar code. The comparison step 1906 will then read the data associated with this entry, such as the density and effective atomic number of the liquid, the container features, diffraction/back scatter signature and product information. Next, step 1906 will compare the parameters such as the density, effective atomic number and/or diffraction/backscatter signature to the parameters that were assessed by the X-ray inspection.

The results of the comparison are passed to step 1910 that performs the product qualification. If there is a match between the parameters read from the knowledge bank and those measured by the X-ray inspection machine, then the process assumes that the container that is being inspected contains a liquid that is consistent with the label on the container.

On the other hand, if no match is found between the parameters read from the knowledge bank and those measured by the X-ray inspection machine, the logic concludes that the liquid in the container is different from what the label says. This is a strong indication that there is a mismatch between the liquid and packaging.

This feature could be particularly to determine, during a production run, if the liquid has been correctly labelled. In the case of medicines, where such errors can be fatal to a patient, the ability to quickly test at the manufacturing or distribution site, all or at least a significant part of the product population can bring significant advantages.

Figure 20 is a block diagram of the equipment used to implement the method described in Figure 19a. The installation is very similar to the set-up described in connection with Figure 1 and for that reason whenever possible similar reference numbers will be used. The main distinction resides in the addition of a bar code reader 2000 that generates a bar code signal on output 2002 conveying the bar codes scanned by the reader 2000. The output 2002 connects to the processing module 200.

In this example of implementation the bar code reader 2000 is separate from the X-ray apparatus 100. Specifically, the bar code reader 2000 may be a handheld reader of the type commonly used at checkout payment stations, in stores. Alternatively, the bar code reader 2000 may be a stationary device that has a reading window. The container is presented in front of the reading window to allow the bar code to be read.

In the case of a hand held bar code reader 2000, the operator 130 would scan the liquid product whose threat level is to be assessed such as to read the bar code. Once the bar code is acquired, the knowledge bank 1908 is searched by the processing module 200 to locate the entry associated with that code. If the entry in the knowledge bank 1908 is identified, information about the entry can be displayed on the operator console 350. For instance one or more container features can be visually shown on the console 350, such as a three-dimensional image of the container, allowing the operator to visually confirm that the entry in the knowledge bank 1908 indeed matches the container that was scanned.

Next, the operator 130 processes the container as discussed earlier. In particular, the liquid product is placed in the tray and the tray put on the conveyor belt of the X-ray apparatus 100. The X-ray scan is performed and the results are passed to the processing module 200. The processing module will process the X-ray image data to extract the response of the liquid in the container to the X-rays. The response is compared to the parameters stored in the previously identified knowledge bank 1908 entry.

In the instance where the bar code reader is a fixed device, it can be integrated in the X-ray apparatus such that the bar code on each container is read as the liquid product is put on the conveyor belt. This may require positioning the containers in the tray in such a way as to leave the bar codes exposed. The reader will appreciate that many options exist to position the bar code reader in a way to suit a wide variety of possible applications.

In a possible variant, the bar code reader can be replaced with a Radio Frequency Identification (RFID) tags reader, suitable for liquid products that use such RFID tags for identification purposes. More specifically, RFID tags have an antennae and a small electronic circuit holding the information to supply when the RFID tag is interrogated. RFID tags can be read over relatively short distances (10 feet or less) and the reading does not have to be in the line of sight of the reader. In this type of application the liquid product to be scanned may be passed close to an RFID tag reader that will gather the identifying information. For instance, the RFID tag reader may be integrated to the X-ray apparatus 100 adjacent the conveyor of the X-ray apparatus. As the liquid product is put in the tray on the conveyor the liquid product will pass close enough the RFID tag reader for the reading operation to take place.

Although various embodiments have been illustrated, this was for the purpose of describing, but not limiting, the invention. Various modifications will become apparent to those skilled in the art and are within the scope of this invention, which is defined more particularly by the attached claims.

Claims

CLAIMS:

1. A method to perform quality inspection of a liquid material held in a container, the method including: a. subjecting the liquid material to X-ray radiation to generate X-ray image data representing an X-ray image of the container holding the liquid; b. comparing information derived from the X-ray image data to a plurality of database records, each database record including reference information corresponding to a particular level of quality, among several possible levels or quality of the liquid material; c. releasing information indicative of the level of quality of the liquid material, the level of quality conveyed by the released information corresponding to the database record matched to the X-ray image data by the comparing.

2. A method as defined in claim 1 , including the step of compensating the X-ray image data for attenuation due to the container.

3. A method as defined in claim 2, wherein the liquid material is a food product.

4. A method as defined in claim 3, wherein the food product is selected from the group consisting of milk, soft drink, syrup, juice and alcoholic beverage.

5. A method as defined in claim 2, wherein the liquid material is a medicinal product.

6. A method as defined in claim 5, wherein the medicinal product is selected from the group consisting of aurally ingestible medicinal product, medicinal product for cutaneous application and injectable medicinal product.

7. A method as defined in claim 6, wherein the aurally ingestible medicinal product is cough syrup.

8. A method as defined in claim 6, wherein the medicinal product for cutaneous application is selected from the group consisting of eye drops, ear drops and nasal spray.

9. A method as defined in claim 2, wherein the liquid product is a personal care product for direct application to the body.

10. A method as defined in claim 9, wherein the personal care product for direct application to the body is selected from the group consisting of shampoo, hair treatment preparations, UV protection preparations, skin cream and cosmetics.

11.A method as defined in claim 2, wherein the liquid product is an industrial product.

12. A method as defined in claim 1 1 , wherein the industrial product is selected in the group consisting of solvent, degreaser, petroleum product, cleaner, paint, adhesive, insecticide and fertilizer.

13. A method as defined in claim 2, wherein the liquid product is a household product.

14. A method as defined in claim 2, wherein the liquid product is a petroleum product.

15. A method as defined in claim 14, wherein the liquid product is selected from the group consisting of crude oil, gasoline, liquefied natural gas and liquefied propane.

16.An apparatus performing quality inspection of a liquid material held in a container, the apparatus including: a. an X-ray inspection station for subjecting the liquid material to X- ray radiation to generate X-ray image data representing an X-ray image of the container holding the liquid; b. a computer based processing unit for comparing information derived from the X-ray image data to a plurality database records, each database record including reference information corresponding to a particular level of quality, among several possible levels or quality of the liquid material; c. an output for releasing information indicative of the level of quality of the liquid material, the level of quality conveyed by the released information corresponding to the database record matched to the X-ray image data by the comparing.

17. A method to determine if a liquid material held in a container is safe for human consumption, the method including: a. subjecting the liquid material to X-ray radiation to generate X-ray image data representing an X-ray image of the container holding the liquid; b. comparing information derived from the X-ray image data to a plurality database records, the plurality of database records, including: i. a first database record including reference information corresponding to a liquid material that is safe for human consumption; ii. a second database record including reference information corresponding to a liquid material that is not safe for human consumption; c. releasing information on the safeness of the liquid product for human consumption on the basis of the results of the comparing.

18. A method as defined in claim 17, including the step of compensating the X-ray image data to compensate the X-ray image data for attenuation due to the container.

19. A method as defined in claim 18, wherein the liquid product is selected from the group consisting of milk, soft drink, syrup, juice and alcoholic beverage.

20. A method as defined in claim 19, wherein the second database record includes reference information corresponding to liquid material with a contaminant which renders the liquid material unsafe for human consumption.

21.A method as defined in claim 20, wherein the contaminant is bacteria.

22.A method as defined in claim 20, wherein the contaminant is a chemical compound.

23.An apparatus to determine if a liquid material held in a container is safe for human consumption, the apparatus including: a. an inspection station for subjecting the liquid material to X-ray radiation to generate X-ray image data representing an X-ray image of the container holding the liquid; b. a computer based processing unit comparing information derived from the X-ray image data to a plurality database records, the plurality of database records, including: i. a first database record including reference information corresponding to a liquid material that is safe for human consumption; ii. a second database record including reference information corresponding to a liquid material that is not safe for human consumption; c. an output for releasing information on the safeness of the liquid product for human consumption on the basis of the results of the comparing.

24. A method to determine a level of sugar content of a liquid material held in a container, the method including: a. subjecting the liquid material to X-ray radiation to generate X-ray image data representing an X-ray image of the container holding the liquid; b. comparing information derived from the X-ray image data to a plurality database records, each database record including reference information corresponding to a particular level of sugar, among several possible levels or sugar in the liquid material;

25. A method as defined in claim 24, including the step of compensating the X-ray image data for attenuation due to the container.

26.An apparatus to determine a level of sugar content of a liquid material held in a container, the method including: a. an inspection station for subjecting the liquid material to X-ray radiation to generate X-ray image data representing an X-ray image of the container holding the liquid; b. a processing unit for comparing information derived from the X-ray image data to a plurality database records, each database record including reference information corresponding to a particular level of sugar, among several possible levels or sugar in the liquid material; c. an output for releasing information indicative of the level sugar content in the liquid material, the level of sugar content conveyed by the released information corresponding to the database record matched to the X-ray image data by the comparing.

27. A method to determine a level of alcohol content of a liquid material held in a container, the method including: a. subjecting the liquid material to X-ray radiation to generate X-ray image data representing an X-ray image of the container holding the liquid; b. comparing information derived from the X-ray image data to a plurality database records, each database record including reference information corresponding to a particular level of alcohol, among several possible levels or alcohol of the liquid material; c. releasing information indicative of the level of alcohol of the liquid material, the level of alcohol conveyed by the released information corresponding to the database record matched to the X-ray image data by the comparing.

28. A method as defined in claim 27, including the step of compensating the X-ray image data for attenuation due to the container.

29.An apparatus to determine a level of alcohol of a liquid material held in a container, the method including: a. an inspection station for subjecting the liquid material to X-ray radiation to generate X-ray image data representing an X-ray image of the container holding the liquid; b. a processing unit for comparing information derived from the X-ray image data to a plurality database records, each database record including reference information corresponding to a particular level of alcohol, among several possible levels or alcohol of the liquid material; c. an output for releasing information indicative of the alcohol content in the liquid material, the level of alcohol conveyed by the released information corresponding to the database record matched to the X-ray image data by the comparing.

30. A method to determine the taste of a liquid material held in a container, the method including: a. subjecting the liquid material to X-ray radiation to generate X-ray image data representing an X-ray image of the container holding the liquid; b. comparing information derived from the X-ray image data to a plurality database records, each database record including reference information corresponding to a different taste, among several possible tastes; c. releasing information indicative of the taste of the liquid material, the taste conveyed by the released information corresponding to the database record matched to the X-ray image data by the comparing.

31. A method as defined in claim 30, including the step of compensating the X-ray image data for attenuation due to the container.

32.An apparatus to determine the taste of a liquid material held in a container, the method including: a. subjecting the liquid material to X-ray radiation to generate X-ray image data representing an X-ray image of the container holding the liquid; b. comparing information derived from the X-ray image data to a plurality database records, each database record including reference information corresponding to a different taste, among several possible tastes; c. releasing information indicative of the taste of the liquid material, the taste conveyed by the released information corresponding to the database record matched to the X-ray image data by the comparing.

33. A method to determine a the mouth felt texture of a liquid material held in a container, the method including: a. subjecting the liquid material to X-ray radiation to generate X-ray image data representing an X-ray image of the container holding the liquid; b. comparing information derived from the X-ray image data to a plurality database records, each database record including reference information corresponding to a different mouth felth texture, among several possible mouth felt textures; c. releasing information indicative of the mouth felt texture of the liquid material, the mouth felt texture conveyed by the released information corresponding to the database record matched to the X-ray image data by the comparing.

34. A method as defined in claim 33, including the step of compensating the X-ray image data for attenuation due to the container.

35. An apparatus to determine the mouth felt texture of a liquid material held in a container, the method including: a. subjecting the liquid material to X-ray radiation to generate X-ray image data representing an X-ray image of the container holding the liquid; b. comparing information derived from the X-ray image data to a plurality database records, each database record including reference information corresponding to a different mouth felt texture, among several possible mouth felt textures; c. releasing information indicative of the mouth felt texture of the liquid material, the mouth felt texture conveyed by the released information corresponding to the database record matched to the X-ray image data by the comparing.

36.A method to determine if a container bearing a liquid product identification holds a liquid matching that product identification, the method including: a. subjecting the liquid material to X-ray radiation to derive a response of the liquid to X-ray radiation; b. inspecting the container to derive product identification information; c. providing a knowledge bank containing responses of different liquids to X-rays, each response being mapped to a corresponding product identification information; d. searching the knowledge bank to determine if an entry exists in the knowledge bank matching the product identification information acquired by the inspecting and also matching the response of the liquid to X-ray radiation acquired by the subjecting; e. using the results of the searching to determine if the liquid in the container matches the product identification on the container.

37. A method as defined in claim 39, including the step of compensating the response of the liquid to X-ray radiation for attenuation due to the container.