CN105203547A - Cloth flaw detection method and device based on intelligent visual sensor - Google Patents
Cloth flaw detection method and device based on intelligent visual sensor Download PDFInfo
- Publication number
- CN105203547A CN105203547A CN201510532513.1A CN201510532513A CN105203547A CN 105203547 A CN105203547 A CN 105203547A CN 201510532513 A CN201510532513 A CN 201510532513A CN 105203547 A CN105203547 A CN 105203547A
- Authority
- CN
- China
- Prior art keywords
- image
- vision sensor
- threshold value
- controller
- wavelet transformation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 72
- 239000004744 fabric Substances 0.000 title claims abstract description 69
- 230000000007 visual effect Effects 0.000 title description 2
- 238000000034 method Methods 0.000 claims abstract description 70
- 238000001914 filtration Methods 0.000 claims abstract description 55
- 230000007547 defect Effects 0.000 claims abstract description 39
- 230000009466 transformation Effects 0.000 claims abstract description 39
- 230000000877 morphologic effect Effects 0.000 claims abstract description 32
- 230000011218 segmentation Effects 0.000 claims abstract description 23
- 230000008569 process Effects 0.000 claims description 48
- 238000003709 image segmentation Methods 0.000 claims description 17
- 230000005540 biological transmission Effects 0.000 claims description 13
- 238000000354 decomposition reaction Methods 0.000 claims description 11
- 238000012360 testing method Methods 0.000 claims description 7
- 238000007781 pre-processing Methods 0.000 claims description 6
- 230000015572 biosynthetic process Effects 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 5
- 238000012545 processing Methods 0.000 abstract description 3
- 238000009940 knitting Methods 0.000 abstract description 2
- 238000004458 analytical method Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 238000007689 inspection Methods 0.000 description 4
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 230000015556 catabolic process Effects 0.000 description 2
- 230000008878 coupling Effects 0.000 description 2
- 238000010168 coupling process Methods 0.000 description 2
- 238000005859 coupling reaction Methods 0.000 description 2
- 238000006731 degradation reaction Methods 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000007797 corrosion Effects 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 238000006073 displacement reaction Methods 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000003628 erosive effect Effects 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 238000007665 sagging Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Abstract
The embodiment of the invention provides a method and a device for detecting cloth flaws based on an intelligent vision sensor. The method mainly comprises the following steps: the intelligent vision sensor collects the image of the cloth cover of the warp knitting machine and transmits the collected image to the controller; the controller carries out two-dimensional wavelet transformation processing on the image data; and denoising the image after the two-dimensional wavelet transform by adopting a morphological filtering method with directivity. The embodiment of the invention provides a cloth defect detection device which achieves the purposes of miniaturization, low cost and high reliability. The detection speed of the cloth flaw can reach 10fps, the detection rate can reach 95%, and a satisfactory result is obtained. By the detection method of direct binary segmentation in the wavelet high-frequency coefficient domain, linear flaws are prevented from being corroded. By adopting a morphological filtering method with directivity, random noise is filtered while a flaw area is reserved.
Description
Technical field
The present invention relates to media communication technical field, particularly relate to a kind of Fabric Defect detection method based on intelligent vision sensor and device.
Background technology
Defect Detection is significant in guarantee fabric quality.Tradition Defect Detection rely on naked eyes carry out, inefficiency and due to the easy tired loss of naked eyes higher.The Fabric Defect detection that develops into of machine vision technique provides new resolving ideas, and over nearly 20 years, Chinese scholars has been carried out a lot of useful exploration and developed some real-time cloth pick-up units.Abouelela proposes a kind of vision inspection apparatus, is made up of camera, video frequency collection card and computing machine, and digitized image carries out binaryzation operation after pre-service, thus detects flaw position.Saeidi proposes a kind of vision inspection apparatus for CircularKnittingMachine, and be that the CMOS camera of 640x320 and personal computer form by resolution, detection algorithm adopts Garbor small echo.Rocco proposes a kind of online real-time vision detection method based on neural network, and can detect common flaw and classify, flaw recall rate reaches 93%.Mak proposes a kind of detection method based on machine vision, and construct prototype system in laboratory, this system is made up of light source, line scan camera (linescancamera), capture card (framegrabber) and computing machine, detection algorithm adopts Gaborwavlet, achieves satisfied result.Sun proposes a kind of self-adapting detecting device of Based PC NN neural network, is that the area array cameras (areascancamera) of 800x600 and computing machine form by resolution.
Fabric Defect pick-up unit above based on machine vision is all Based PC frame structures, namely system is made up of light source, camera, capture card and computing machine, in the core that this PC framework Computer is computing, view data is transferred in PC by capture card, runs various detection algorithm and output detections result.
The shortcoming of the cloth pick-up unit of above-mentioned Based PC is that cost is high, fluctuation of service, power consumption are high, volume is installed greatly, not easily.
Summary of the invention
The embodiment provides a kind of Fabric Defect detection method based on intelligent vision sensor and device, to realize improving Fabric Defect detection efficiency.
To achieve these goals, this invention takes following technical scheme.
According to an aspect of the present invention, provide a kind of Fabric Defect detection method based on intelligent vision sensor, comprising:
Intelligent vision sensor gathers the view data of the cloth cover of tricot machine, by the image data transmission of collection to controller;
Described controller carries out two-dimensional wavelet transformation process to described view data;
The directive morphologic filtering method of tool is adopted to carry out denoising to the view data after described two-dimensional wavelet transformation process.
Preferably, described intelligent vision sensor gathers the view data of the cloth cover of tricot machine, by the image data transmission of collection to controller, comprising:
Flush bonding processor and imageing sensor are integrated in vision sensor, above the cloth cover of the tricot machine that multiple vision sensor is installed, each vision sensor covers the cloth cover of setting width, each vision sensor all gathers the image of covered cloth cover, by the image transmitting of collection to controller.
Preferably, described controller also comprises before carrying out two-dimensional wavelet transformation process to described view data: controller carries out pre-service to the image that each intelligent vision sensor transmissions is come, and described pre-service comprises equilibrium treatment and/or filter preprocessing.
Preferably, described controller carries out two-dimensional wavelet transformation process to described view data, comprising:
Described controller carries out two-dimensional wavelet transformation process to described view data, obtain multiple subband, wherein, subband LH2 represents the high-frequency components in the low frequency of horizontal direction after two layers of wavelet decomposition and vertical direction, carries out binarization segmentation process to the image that subband LH2 is formed;
Select the threshold value T that initial, this threshold value gets the maximum gray scale of image and the average of minimal gray of described subband LH2 formation, and utilizing the Image Segmentation Using that threshold value T is formed described subband LH2, is two parts by Iamge Segmentation: gray-scale value is greater than the image-region G of T
1the image-region G of T is less than or equal to gray-scale value
2, calculate G
1and G
2the gray average u of the pixel comprised
1and u
2, obtain new threshold value
recycle new threshold value T to Image Segmentation Using;
Repeat above threshold calculations, utilize threshold value to the treatment step of Image Segmentation Using, the difference of the threshold value T calculated until double is less than setting value, obtains the image after binarization segmentation process.
Preferably, the described directive morphologic filtering method of employing tool carries out denoising to the image after described two-dimensional wavelet transformation process, comprising:
Arrange the directive morphologic filtering template of tool, described morphologic filtering template comprises vertical detection Filtering Template and diagonal angle detection filter template;
Described vertical detection Filtering Template and diagonal angle detection filter template is utilized to carry out denoising to the image after described two-dimensional wavelet transformation process, obtain the testing result of Fabric Defect, when applying vertical detection Filtering Template, when current pixel is consistent with its upper and lower two neighbors, then described current pixel is set to 255; When applying diagonal angle detection filter template, when current pixel is consistent to the neighbor of two on angular direction with it, then described current pixel is set to 255.
According to another aspect of the present invention, provide a kind of Fabric Defect pick-up unit based on intelligent vision sensor, comprising: intelligent vision sensor, controller;
Described intelligent vision sensor, for gathering the view data of the cloth cover of tricot machine, by the image data transmission of collection to controller;
Described controller, for carrying out two-dimensional wavelet transformation process to described view data, adopts the directive morphologic filtering method of tool to carry out denoising to the image after described two-dimensional wavelet transformation process.
Preferably, described vision sensor is integrated with flush bonding processor and imageing sensor, above the cloth cover of the tricot machine that multiple vision sensor is installed, each vision sensor covers the cloth cover of setting width, each vision sensor all gathers the image of covered cloth cover, by the image transmitting of collection to controller.
Preferably, described controller comprises:
Pretreatment module, also comprised before carrying out two-dimensional wavelet transformation process to described view data: controller carries out pre-service to the image that each intelligent vision sensor transmissions is come, and described pre-service comprises equilibrium treatment and/or filter preprocessing;
Two-dimensional wavelet transformation process, for carrying out two-dimensional wavelet transformation process to described controller to described view data, obtain multiple subband, wherein, subband LH2 represents the high-frequency components in the low frequency of horizontal direction after two layers of wavelet decomposition and vertical direction, carries out binarization segmentation process to the image that subband LH2 is formed;
Select the threshold value T that initial, this threshold value gets the maximum gray scale of image and the average of minimal gray of described subband LH2 formation, and utilizing the Image Segmentation Using that threshold value T is formed described subband LH2, is two parts by Iamge Segmentation: gray-scale value is greater than the image-region G of T
1the image-region G of T is less than or equal to gray-scale value
2, calculate G
1and G
2the gray average u of the pixel comprised
1and u
2, obtain new threshold value
recycle new threshold value T to Image Segmentation Using;
Repeat above threshold calculations, utilize threshold value to the treatment step of Image Segmentation Using, the difference of the threshold value T calculated until double is less than setting value, obtains the image after binarization segmentation process.
Preferably, described controller also comprises:
Morphologic filtering module, for arranging the directive morphologic filtering template of tool, described morphologic filtering template comprises vertical detection Filtering Template and diagonal angle detection filter template;
Described vertical detection Filtering Template and diagonal angle detection filter template is utilized to carry out denoising to the image after described two-dimensional wavelet transformation process, obtain the testing result of Fabric Defect, when applying vertical detection Filtering Template, when current pixel is consistent with its upper and lower two neighbors, then described current pixel is set to 255; When applying diagonal angle detection filter template, when current pixel is consistent to the neighbor of two on angular direction with it, then described current pixel is set to 255.
The technical scheme provided as can be seen from the embodiment of the invention described above, the embodiment of the present invention proposes a kind of pick-up unit based on intelligent vision sensor newly.The program is different from the Vision Builder for Automated Inspection based on computer system, achieves the object of miniaturization, cost degradation and high reliability.The detection speed of Fabric Defect can reach 10fps, and the detection efficiency of Fabric Defect can reach 95%, obtains satisfied achievement.
The aspect that the present invention adds and advantage will part provide in the following description, and these will become obvious from the following description, or be recognized by practice of the present invention.
Accompanying drawing explanation
In order to be illustrated more clearly in the technical scheme of the embodiment of the present invention, below the accompanying drawing used required in describing embodiment is briefly described, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skill in the art, under the prerequisite not paying creative work, other accompanying drawing can also be obtained according to these accompanying drawings.
The schematic flow sheet of a kind of cloth cover flaw detection method based on intelligent vision sensor that Fig. 1 provides for the embodiment of the present invention;
Fig. 2 be view data after two-dimensional wavelet transformation, be broken down into the schematic diagram of multiple subband;
Fig. 3 is morphologic filtering template schematic diagram of the prior art;
Fig. 4 is the tool directive morphologic filtering template schematic diagram that the embodiment of the present invention proposes;
The structural representation of a kind of Fabric Defect pick-up unit based on intelligent vision sensor that Fig. 5 provides for the embodiment of the present invention;
The structural representation of a kind of controller that Fig. 6 provides for the embodiment of the present invention, pretreatment module 61, two-dimensional wavelet transformation process 62, morphologic filtering module 63;
The structural representation of a kind of vision sensor that Fig. 7 provides for the embodiment of the present invention.
Embodiment
Be described below in detail embodiments of the present invention, the example of described embodiment is shown in the drawings, and wherein same or similar label represents same or similar element or has element that is identical or similar functions from start to finish.Being exemplary below by the embodiment be described with reference to the drawings, only for explaining the present invention, and can not limitation of the present invention being interpreted as.
Those skilled in the art of the present technique are appreciated that unless expressly stated, and singulative used herein " ", " one ", " described " and " being somebody's turn to do " also can comprise plural form.Should be further understood that, the wording used in instructions of the present invention " comprises " and refers to there is described feature, integer, step, operation, element and/or assembly, but does not get rid of and exist or add other features one or more, integer, step, operation, element, assembly and/or their group.Should be appreciated that, when we claim element to be " connected " or " coupling " to another element time, it can be directly connected or coupled to other elements, or also can there is intermediary element.In addition, " connection " used herein or " coupling " can comprise wireless connections or couple.Wording "and/or" used herein comprises one or more arbitrary unit listing item be associated and all combinations.
Those skilled in the art of the present technique are appreciated that unless otherwise defined, and all terms used herein (comprising technical term and scientific terminology) have the meaning identical with the general understanding of the those of ordinary skill in field belonging to the present invention.Should also be understood that those terms defined in such as general dictionary should be understood to have the meaning consistent with the meaning in the context of prior art, unless and define as here, can not explain by idealized or too formal implication.
For ease of the understanding to the embodiment of the present invention, be further explained explanation below in conjunction with accompanying drawing for several specific embodiment, and each embodiment does not form the restriction to the embodiment of the present invention.
Embodiment one
Along with the raising of DSP embedded processor processing power, detection algorithm and imageing sensor are integrated, form embedded intelligence vision sensor and become possibility.Just under this background, the embodiment of the present invention proposes a kind of Fabric Defect pick-up unit based on intelligent vision sensor, and is applied to the Defect Detection of tricot machine cloth cover, achieves satisfied achievement.
The modal flaw of tricot machine is cracked ends, the flaw formed after namely being disconnected by warp.In multi-bar warp knitting machine, be very unconspicuous by singly disconnecting with yarn the cracked ends flaw formed, particularly very thin yarn, such as 20D (referring to that the weight of the yarn of 9000 meters is 20 grams).Detect the work that this flaw is very challenging property.In order to improve recall rate, the embodiment of the present invention takes following measures: the high resolution imageing sensor 1) using 2,000,000 pixels, and accuracy of detection can reach 0.5mm; For the feature of tricot machine flaw, propose one and carry out adaptive threshold fuzziness based on wavelet decomposition territory, instead of based on reconstructed image; Propose a kind of morphologic filtering with set direction, while filtering noise, retain defect areas.
Traditional industrial camera only can gather image, does not have processing power to image.If high performance flush bonding processor and imageing sensor are combined, detection algorithm is transplanted in flush bonding processor, then constitutes intelligent vision sensor.The advantage of vision sensor is apparent, and such as volume is little, be easy to install, low in energy consumption, low price etc., and the fail operation of single-sensor does not affect the operation of miscellaneous equipment.
The embodiment of the present invention adopts wavelet analysis as the basis of detection algorithm, and the schematic flow sheet of the cloth cover flaw detection method based on intelligent vision sensor that the embodiment of the present invention proposes as shown in Figure 1, comprises following treatment step:
Step S110, each intelligent vision sensor gather the image of the cloth cover of tricot machine, by the image transmitting of collection to controller.
Step S120, controller carry out equilibrium and filter preprocessing to the image that each intelligent vision sensor transmissions is come.
There is more noise in the image that cmos image sensor obtains, and cause the uneven of brightness of image due to the inconsistent meeting of light, these problems all can impact analysis result.Therefore need to do pretreatment work to improve picture quality before graphical analysis.Image semantic classification comprises equilibrium and filtering, and the image after equilibrium treatment needs to carry out filtering process, and by the comparison to several filtering mode, the embodiment of the present invention selects the medium filtering of 3x3 template.
Step S130, two-dimensional wavelet transformation process is carried out to pretreated view data.
Wavelet transformation, owing to having good local time's frequency domain characteristic, is widely used in signal analysis.Square integrable spatial function f (t) ∈ L
2(R) continuous wavelet transform is defined as:
Wherein ψ (t) ∈ L
2(R) and meet
be called one " mother wavelet ", in above formula
it is the Fourier transform of ψ (t).Continuous wavelet inverse transformation is defined as
Wherein
Do in yardstick and displacement to formula (3.3) discrete, obtain
ψ
j,k(t)=a
0 -j/2ψ(a
0 -jt-kb
0),j,k∈Z(3.6)
Then corresponding discrete wavelet transformer is changed to
View data is after two-dimensional wavelet transformation, be broken down into multiple subbands as shown in Figure 2, LL represents the low frequency component in level and vertical direction, LH represents the high-frequency components in the low frequency of horizontal direction and vertical direction, HL represents the low frequency component in the high frequency of horizontal direction and vertical direction, HH represents the high-frequency components of level and vertical direction, and subscript 1 represents one deck wavelet decomposition, and subscript 2 represents two layers of wavelet decomposition.Example: subband LL1 represents the low frequency component after one deck wavelet decomposition in level and vertical direction, subband HH2 represents the high-frequency components of level and vertical direction after two layers of wavelet decomposition, and subband LH2 represents the high-frequency components in the low frequency of horizontal direction after two layers of wavelet decomposition and vertical direction.The arrangement mode of each subband shown in above-mentioned Fig. 2 is fixing, is a kind of convenient diagram method for expressing sanctified by usage understood.
Step S140, to small echo high frequency coefficient LH2 form image carry out binarization segmentation process.
Cracked ends flaw due to tricot machine is all in the vertical direction, therefore can obtain obvious unwanted visual characteristic from LH2 coefficient, and HL2 and HH2 comprises is random noise.
In order to improve the real-time of detection, reduce noise to the interference of rebuilding image, the present invention carries out binarization segmentation to the image that small echo high frequency coefficient LH2 is formed, instead of carries out image reconstruction.First select the threshold value T that initial, this threshold value gets the maximum gray scale of image and the average of minimal gray.Utilizing threshold value T to Image Segmentation Using, is two parts by Iamge Segmentation: gray-scale value is greater than the image-region G of T
1the image-region G of T is less than or equal to gray-scale value
2.Then, G is calculated
1and G
2the gray average u of the pixel comprised
1and u
2, obtain new threshold value
recycle new threshold value T to Image Segmentation Using.Repeat above step, the difference of the T calculated until double is less than setting value.
Step S150, the directive morphologic filtering method of employing tool carry out denoising to the image after binarization segmentation.
Mathematical morphology is based upon on the basis of set theory, may be used for the removal of picture noise.Basic morphology operations comprises corrosion and expands.Wire on tricot machine flaw, minimum flaw only may have the width of 1 pixel, if adopt the template of conventional nxn to carry out erosion operation, so defect areas also can be corroded.The flaw occurred in field of view edge is not vertical, but has the oblique line of certain angle.If adopt the template of nx1, the flaw of edge also can be corroded.In order to address this problem, the embodiment of the present invention proposes the directive morphologic filtering method of a kind of tool, ensures that oblique line removes random noise while not being corroded.
In image after binarization segmentation, the pixel value of defect areas and noise is 255, is shown as white, and other parts are 0, are shown as black.For 3x3 template, morphologic filtering is carried out to bianry image.As shown in Figure 3, hollow dots represents current processed pixel to morphologic filtering template of the prior art, and its gray-scale value is 255.Only have when adjacent 8 pixels have same gray-scale value with current pixel, this pixel is just set to 255, otherwise is set to 0.Obviously very thin wire flaw will by this template filtering.The directive template of tool that the embodiment of the present invention proposes as shown in Figure 4.In Fig. 4, template a is vertical detection template, and when current pixel and its neighbouring pixel two pixels are consistent, this pixel is set to 255.Template b and c is all diagonal angle detection template, and when consistent to the pixel value of 3 on angular direction, current pixel is set to 255.The relation of three templates is the relations of " or ", as long as namely meet a template current point be just set to 255.All be retained at the flaw of the sagging direct sum oblique line of this template strategy, and random noise is by filtering.
After image after binarization segmentation being carried out to the denoising of the directive morphologic filtering method of tool, just can obtain the testing result of Fabric Defect.
Embodiment two
The structural representation of a kind of Fabric Defect pick-up unit based on intelligent vision sensor that the embodiment of the present invention provides as shown in Figure 5, comprising: intelligent vision sensor, controller.Fig. 5 is the applicable cases of vision sensor on tricot machine of embodiment of the present invention exploitation, multiple (such as 6) vision sensor is installed in above the cloth cover of the tricot machine of setting width (such as 210 cun wide), and each vision sensor covers 90cm cloth cover.Each vision sensor all can work alone, and all each vision sensors export to be cascaded and are connected on controller.
Described intelligent vision sensor, for gathering the view data of the cloth cover of tricot machine, by the image data transmission of collection to controller; Be integrated with flush bonding processor and imageing sensor, above the cloth cover of the tricot machine that multiple vision sensor is installed, each vision sensor covers the cloth cover of setting width, and each vision sensor all gathers the image of covered cloth cover, by the image transmitting of collection to controller.
Described controller, for carrying out two-dimensional wavelet transformation process to described view data, adopts the directive morphologic filtering method of tool to carry out denoising to the image after described two-dimensional wavelet transformation process.
The structural representation of a kind of controller that the embodiment of the present invention provides as shown in Figure 6, comprising:
Pretreatment module 61, also comprised before carrying out two-dimensional wavelet transformation process to described view data: controller carries out pre-service to the image that each intelligent vision sensor transmissions is come, and described pre-service comprises equilibrium treatment and/or filter preprocessing;
Two-dimensional wavelet transformation process 62, for carrying out two-dimensional wavelet transformation process to described controller to described view data, obtain multiple subband, wherein, subband LH2 represents the high-frequency components in the low frequency of horizontal direction after two layers of wavelet decomposition and vertical direction, carries out binarization segmentation process to the image that subband LH2 is formed;
Select the threshold value T that initial, this threshold value gets the maximum gray scale of image and the average of minimal gray of described subband LH2 formation, and utilizing the Image Segmentation Using that threshold value T is formed described subband LH2, is two parts by Iamge Segmentation: gray-scale value is greater than the image-region G of T
1the image-region G of T is less than or equal to gray-scale value
2, calculate G
1and G
2the gray average u of the pixel comprised
1and u
2, obtain new threshold value
recycle new threshold value T to Image Segmentation Using;
Repeat above threshold calculations, utilize threshold value to the treatment step of Image Segmentation Using, the difference of the threshold value T calculated until double is less than setting value, obtains the image after binarization segmentation process.
Morphologic filtering module 63, for arranging the directive morphologic filtering template of tool, described morphologic filtering template comprises vertical detection Filtering Template and diagonal angle detection filter template;
Described vertical detection Filtering Template and diagonal angle detection filter template is utilized to carry out denoising to the image after described two-dimensional wavelet transformation process, obtain the testing result of Fabric Defect, when applying vertical detection Filtering Template, when current pixel is consistent with its upper and lower two neighbors, then described current pixel is set to 255; When applying diagonal angle detection filter template, when current pixel is consistent to the neighbor of two on angular direction with it, then described current pixel is set to 255.
The structured flowchart of a kind of vision sensor that the embodiment of the present invention provides as shown in Figure 7, is made up of cmos image sensor, ISP processor module, DSP embedded processor, SDRAM data-carrier store, FLASH program storage, Ethernet interface and relay.Image-forming component adopts the cmos sensor of 2,000,000 pixels, and resolution is 1600x1200, gets wherein 1600x200 data process for raising detection speed.
Detection algorithm is the core of vision sensor, and in algorithms selection, we consider two factors: arithmetic speed and the susceptibility to flaw.Because Algorithms of Wavelet Analysis is very responsive to wire flaw, and operand is less.
The installation and operation on the tricot machine of certain factory of the cloth cover Defect Detection device based on intelligent vision sensor that the present invention proposes.210 " cloth cover on be evenly distributed 6 sensors, each sensor covers the width of 900mm.Each sensor can work alone, and the result of detection comprises quantity and the position coordinates of flaw.Testing result and cloth cover image can be sent to PC by Ethernet, and show in the client software of PC.
In sum, the embodiment of the present invention proposes a kind of Fabric Defect pick-up unit based on intelligent vision sensor newly.This device is different from the Vision Builder for Automated Inspection based on computer system, achieves the object of miniaturization, cost degradation and high reliability.The detection speed of Fabric Defect can reach 10fps, and verification and measurement ratio can reach 95%, obtains satisfied achievement.
Intelligent vision sensor in the embodiment of the present invention is made up of High Performance DSP processor and 2,000,000 pixel high-definition image sensors, image clearly can be obtained, and there is powerful arithmetic capability, by the detection method at the direct binarization segmentation in small echo high frequency coefficient territory, avoid wire flaw and be corroded.By adopting the directive morphologic filtering method of tool, filtering random noise while reservation defect areas.
One of ordinary skill in the art will appreciate that: accompanying drawing is the schematic diagram of an embodiment, the module in accompanying drawing or flow process might not be that enforcement the present invention is necessary.
As seen through the above description of the embodiments, those skilled in the art can be well understood to the mode that the present invention can add required general hardware platform by software and realizes.Based on such understanding, technical scheme of the present invention can embody with the form of software product the part that prior art contributes in essence in other words, this computer software product can be stored in storage medium, as ROM/RAM, magnetic disc, CD etc., comprising some instructions in order to make a computer equipment (can be personal computer, server, or the network equipment etc.) perform the method described in some part of each embodiment of the present invention or embodiment.
Each embodiment in this instructions all adopts the mode of going forward one by one to describe, between each embodiment identical similar part mutually see, what each embodiment stressed is the difference with other embodiments.Especially, for device or system embodiment, because it is substantially similar to embodiment of the method, so describe fairly simple, relevant part illustrates see the part of embodiment of the method.Apparatus and system embodiment described above is only schematic, the wherein said unit illustrated as separating component or can may not be and physically separates, parts as unit display can be or may not be physical location, namely can be positioned at a place, or also can be distributed in multiple network element.Some or all of module wherein can be selected according to the actual needs to realize the object of the present embodiment scheme.Those of ordinary skill in the art, when not paying creative work, are namely appreciated that and implement.
The above; be only the present invention's preferably embodiment, but protection scope of the present invention is not limited thereto, is anyly familiar with those skilled in the art in the technical scope that the present invention discloses; the change that can expect easily or replacement, all should be encompassed within protection scope of the present invention.Therefore, protection scope of the present invention should be as the criterion with the protection domain of claim.
Claims (9)
1., based on a Fabric Defect detection method for intelligent vision sensor, it is characterized in that, comprising:
Intelligent vision sensor gathers the view data of the cloth cover of tricot machine, by the image data transmission of collection to controller;
Described controller carries out two-dimensional wavelet transformation process to described view data;
The directive morphologic filtering method of tool is adopted to carry out denoising to the view data after described two-dimensional wavelet transformation process.
2. the Fabric Defect detection method based on intelligent vision sensor according to claim 1, is characterized in that, described intelligent vision sensor gathers the view data of the cloth cover of tricot machine, by the image data transmission of collection to controller, comprising:
Flush bonding processor and imageing sensor are integrated in vision sensor, above the cloth cover of the tricot machine that multiple vision sensor is installed, each vision sensor covers the cloth cover of setting width, each vision sensor all gathers the image of covered cloth cover, by the image transmitting of collection to controller.
3. the Fabric Defect detection method based on intelligent vision sensor according to claim 1, it is characterized in that, described controller also comprises before carrying out two-dimensional wavelet transformation process to described view data: controller carries out pre-service to the image that each intelligent vision sensor transmissions is come, and described pre-service comprises equilibrium treatment and/or filter preprocessing.
4. the Fabric Defect detection method based on intelligent vision sensor according to claim 3, it is characterized in that, described controller carries out two-dimensional wavelet transformation process to described view data, comprising:
Described controller carries out two-dimensional wavelet transformation process to described view data, obtain multiple subband, wherein, subband LH2 represents the high-frequency components in the low frequency of horizontal direction after two layers of wavelet decomposition and vertical direction, carries out binarization segmentation process to the image that subband LH2 is formed;
Select the threshold value T that initial, this threshold value gets the maximum gray scale of image and the average of minimal gray of described subband LH2 formation, and utilizing the Image Segmentation Using that threshold value T is formed described subband LH2, is two parts by Iamge Segmentation: gray-scale value is greater than the image-region G of T
1the image-region G of T is less than or equal to gray-scale value
2, calculate G
1and G
2the gray average u of the pixel comprised
1and u
2, obtain new threshold value
recycle new threshold value T to Image Segmentation Using;
Repeat above threshold calculations, utilize threshold value to the treatment step of Image Segmentation Using, the difference of the threshold value T calculated until double is less than setting value, obtains the image after binarization segmentation process.
5. the Fabric Defect detection method based on intelligent vision sensor according to claim 4, is characterized in that, the described directive morphologic filtering method of employing tool carries out denoising to the image after described two-dimensional wavelet transformation process, comprising:
Arrange the directive morphologic filtering template of tool, described morphologic filtering template comprises vertical detection Filtering Template and diagonal angle detection filter template;
Described vertical detection Filtering Template and diagonal angle detection filter template is utilized to carry out denoising to the image after described two-dimensional wavelet transformation process, obtain the testing result of Fabric Defect, when applying vertical detection Filtering Template, when current pixel is consistent with its upper and lower two neighbors, then described current pixel is set to 255; When applying diagonal angle detection filter template, when current pixel is consistent to the neighbor of two on angular direction with it, then described current pixel is set to 255.
6. based on a Fabric Defect pick-up unit for intelligent vision sensor, it is characterized in that, comprising: intelligent vision sensor, controller;
Described intelligent vision sensor, for gathering the view data of the cloth cover of tricot machine, by the image data transmission of collection to controller;
Described controller, for carrying out two-dimensional wavelet transformation process to described view data, adopts the directive morphologic filtering method of tool to carry out denoising to the image after described two-dimensional wavelet transformation process.
7. the Fabric Defect pick-up unit based on intelligent vision sensor according to claim 6, it is characterized in that, described vision sensor is integrated with flush bonding processor and imageing sensor, above the cloth cover of the tricot machine that multiple vision sensor is installed, each vision sensor covers the cloth cover of setting width, each vision sensor all gathers the image of covered cloth cover, by the image transmitting of collection to controller.
8. the Fabric Defect pick-up unit based on intelligent vision sensor according to claim 7, it is characterized in that, described controller comprises:
Pretreatment module, also comprised before carrying out two-dimensional wavelet transformation process to described view data: controller carries out pre-service to the image that each intelligent vision sensor transmissions is come, and described pre-service comprises equilibrium treatment and/or filter preprocessing;
Two-dimensional wavelet transformation process, for carrying out two-dimensional wavelet transformation process to described controller to described view data, obtain multiple subband, wherein, subband LH2 represents the high-frequency components in the low frequency of horizontal direction after two layers of wavelet decomposition and vertical direction, carries out binarization segmentation process to the image that subband LH2 is formed;
Select the threshold value T that initial, this threshold value gets the maximum gray scale of image and the average of minimal gray of described subband LH2 formation, and utilizing the Image Segmentation Using that threshold value T is formed described subband LH2, is two parts by Iamge Segmentation: gray-scale value is greater than the image-region G of T
1the image-region G of T is less than or equal to gray-scale value
2, calculate G
1and G
2the gray average u of the pixel comprised
1and u
2, obtain new threshold value
recycle new threshold value T to Image Segmentation Using;
Repeat above threshold calculations, utilize threshold value to the treatment step of Image Segmentation Using, the difference of the threshold value T calculated until double is less than setting value, obtains the image after binarization segmentation process.
9. the Fabric Defect pick-up unit based on intelligent vision sensor according to claim 8, it is characterized in that, described controller also comprises:
Morphologic filtering module, for arranging the directive morphologic filtering template of tool, described morphologic filtering template comprises vertical detection Filtering Template and diagonal angle detection filter template;
Described vertical detection Filtering Template and diagonal angle detection filter template is utilized to carry out denoising to the image after described two-dimensional wavelet transformation process, obtain the testing result of Fabric Defect, when applying vertical detection Filtering Template, when current pixel is consistent with its upper and lower two neighbors, then described current pixel is set to 255; When applying diagonal angle detection filter template, when current pixel is consistent to the neighbor of two on angular direction with it, then described current pixel is set to 255.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510532513.1A CN105203547A (en) | 2015-08-26 | 2015-08-26 | Cloth flaw detection method and device based on intelligent visual sensor |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510532513.1A CN105203547A (en) | 2015-08-26 | 2015-08-26 | Cloth flaw detection method and device based on intelligent visual sensor |
Publications (1)
Publication Number | Publication Date |
---|---|
CN105203547A true CN105203547A (en) | 2015-12-30 |
Family
ID=54951341
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510532513.1A Pending CN105203547A (en) | 2015-08-26 | 2015-08-26 | Cloth flaw detection method and device based on intelligent visual sensor |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105203547A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110286131A (en) * | 2019-07-09 | 2019-09-27 | 安徽富煌科技股份有限公司 | A kind of textile flaw automatic detection analysis prompt system and method |
CN110389130A (en) * | 2019-07-04 | 2019-10-29 | 盎古(上海)科技有限公司 | Intelligent checking system applied to fabric |
CN111366586A (en) * | 2020-04-28 | 2020-07-03 | 杭州华纺布艺有限公司 | Cloth flaw detection method and system, storage medium and cloth inspecting machine |
CN112801930A (en) * | 2019-10-28 | 2021-05-14 | 上海铼钠克数控科技股份有限公司 | Method and equipment for detecting warp and weft defects of fabric in backlight based on machine vision |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP0757245A2 (en) * | 1995-08-03 | 1997-02-05 | TZN Forschungs- und Entwicklungszentrum Unterlüss GmbH | Apparatus for detecting streaky surface defects |
US20020054694A1 (en) * | 1999-03-26 | 2002-05-09 | George J. Vachtsevanos | Method and apparatus for analyzing an image to direct and identify patterns |
US20020109112A1 (en) * | 2001-02-09 | 2002-08-15 | Guha Sujoy D. | Web inspection system |
CN1760437A (en) * | 2005-11-10 | 2006-04-19 | 东华大学 | Automatic system for assessing grade of cloth inspection objectively |
CN101063660A (en) * | 2007-01-30 | 2007-10-31 | 蹇木伟 | Method for detecting textile defect and device thereof |
CN101923059A (en) * | 2010-04-15 | 2010-12-22 | 北京经纬纺机新技术有限公司 | Automatic cloth inspection method and system |
CN104198498A (en) * | 2014-09-12 | 2014-12-10 | 河海大学常州校区 | Method and device for detecting cloth flaws based on adaptive orthogonal wavelet transform |
-
2015
- 2015-08-26 CN CN201510532513.1A patent/CN105203547A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP0757245A2 (en) * | 1995-08-03 | 1997-02-05 | TZN Forschungs- und Entwicklungszentrum Unterlüss GmbH | Apparatus for detecting streaky surface defects |
US20020054694A1 (en) * | 1999-03-26 | 2002-05-09 | George J. Vachtsevanos | Method and apparatus for analyzing an image to direct and identify patterns |
US20020109112A1 (en) * | 2001-02-09 | 2002-08-15 | Guha Sujoy D. | Web inspection system |
CN1760437A (en) * | 2005-11-10 | 2006-04-19 | 东华大学 | Automatic system for assessing grade of cloth inspection objectively |
CN101063660A (en) * | 2007-01-30 | 2007-10-31 | 蹇木伟 | Method for detecting textile defect and device thereof |
CN101923059A (en) * | 2010-04-15 | 2010-12-22 | 北京经纬纺机新技术有限公司 | Automatic cloth inspection method and system |
CN104198498A (en) * | 2014-09-12 | 2014-12-10 | 河海大学常州校区 | Method and device for detecting cloth flaws based on adaptive orthogonal wavelet transform |
Non-Patent Citations (6)
Title |
---|
LI YUNDONG ET AL.: "Online Fabric Defect Inspection Using Smart Visual Sensors", 《SENSORS》 * |
MAK K.L. ET AL.: "Fabric defect detection using morphological filters", 《IMAGE AND VISION COMPUTING》 * |
刘伟斌等: "一种方向性纹理织物疵点的检测方法", 《华侨大学学报(自然科学版)》 * |
徐增波等: "基于二维连续小波变换的织物疵点检测", 《中国纺织大学学报》 * |
李勇等: "基于机器视觉的坯布自动检测技术", 《纺织学报》 * |
李立轻等: "图像处理用于织物疵点自动检测的研究进展", 《东华大学学报(自然科学版)》 * |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110389130A (en) * | 2019-07-04 | 2019-10-29 | 盎古(上海)科技有限公司 | Intelligent checking system applied to fabric |
CN110286131A (en) * | 2019-07-09 | 2019-09-27 | 安徽富煌科技股份有限公司 | A kind of textile flaw automatic detection analysis prompt system and method |
CN112801930A (en) * | 2019-10-28 | 2021-05-14 | 上海铼钠克数控科技股份有限公司 | Method and equipment for detecting warp and weft defects of fabric in backlight based on machine vision |
CN111366586A (en) * | 2020-04-28 | 2020-07-03 | 杭州华纺布艺有限公司 | Cloth flaw detection method and system, storage medium and cloth inspecting machine |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Wang et al. | Deep learning for asphalt pavement cracking recognition using convolutional neural network | |
CN105203547A (en) | Cloth flaw detection method and device based on intelligent visual sensor | |
Muresan et al. | Demosaicing using optimal recovery | |
CN110268190A (en) | A kind of underground pipe gallery leakage detection method based on the processing of static Infrared Thermogram | |
CN106548182A (en) | Based on deep learning and the causal analytic pavement distress survey method and device of master | |
CN110033431B (en) | Non-contact detection device and detection method for detecting corrosion area on surface of steel bridge | |
CN103487729A (en) | Electrical equipment defect detection method based on fusion of ultraviolet video and infrared video | |
CN103593670A (en) | Copper sheet and strip surface defect detection method based on-line sequential extreme learning machine | |
CN102183525A (en) | Bridge apparent state automatic detection device and method based on charge coupled device (CCD) array photographic technique | |
CN1948603A (en) | Method of identifying woven fabric defect | |
Gou et al. | Remote sensing image super-resolution reconstruction based on nonlocal pairwise dictionaries and double regularization | |
Shen et al. | Deep joint rain and haze removal from a single image | |
CN110807749A (en) | Single image raindrop removing method based on dense multi-scale generation countermeasure network | |
CN110826429A (en) | Scenic spot video-based method and system for automatically monitoring travel emergency | |
CN105931246A (en) | Fabric flaw detection method based on wavelet transformation and genetic algorithm | |
CN112232237A (en) | Vehicle flow monitoring method, system, computer device and storage medium | |
CN115908354A (en) | Photovoltaic panel defect detection method based on double-scale strategy and improved YOLOV5 network | |
Serir et al. | No-reference blur image quality measure based on multiplicative multiresolution decomposition | |
CN106702870A (en) | Road crack detecting system | |
CN103778611A (en) | Switch weighting vector median filter method utilizing edge detection | |
Van Ruitenbeek et al. | Convolutional Neural Networks for vehicle damage detection | |
CN104700405A (en) | Foreground detection method and system | |
CN106056608A (en) | Image dot-line defect detection method and device | |
CN107590785A (en) | A kind of Brillouin spectrum image-recognizing method based on sobel operators | |
CN103455990A (en) | Image fusion method with visual attention mechanism and PCNN combined |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20151230 |