CN103606194A - Space, heaven and earth integration situation expression engine and classification and grading target browsing method thereof - Google Patents

Abstract

The invention relates to a space, heaven and earth integration situation expression engine and a classification and grading target browsing method thereof. The method comprises the following steps of target classification: determining whether the target belongs to a ball-type target or a non-ball type target; using a universal viewpoint control method if the target belongs to a non-ball type target; distance grading: determining whether far viewpoint browsing or near spherical browsing is used if the browsing belongs to the ball-type target; using the universal viewpoint control method if the far viewpoint browsing is used; using a special viewpoint control method if the near spherical browsing is used. The universal viewpoint control method can be used for the far viewpoint of the ball-type target and viewpoint control of other non-ball type target. A main distinction of the ball-type target and the non-ball type target is characterized in that when browsing is performed on the ball-type target, the viewpoint is not entered into an internal portion of the ball; the viewpoint control method used during the near spherical browsing is different from the control method used during the far viewpoint browsing.

Description

Empty Incorporate situation is expressed engine and classify and grading target browsing method thereof

Technical field

The present invention relates to a kind of empty Incorporate situation and express engine and classify and grading target browsing method thereof.

Background technology

Along with the development of technology, the mankind's activity is no longer confined to ground, and space becomes the another important place of mankind's activity.In order to capture the important manor of this piece, space, increasing satellite launch lift-off, makes to have reached thousands of with the satellite losing efficacy in-orbit, and space environment sharply worsens.For guaranteeing space safety, safeguard spatial benefit, various countries have set up the roomage state sensory perceptual system of oneself one after another.Roomage state sensory perceptual system is the systems engineering of a bulky complex, its packet obtaining is containing all kinds of multi-source heterogeneous data such as space-based, grounds, how these data are effectively used, for strategy and tactics decision-making and other space operations provide information service to become a problem in the urgent need to address.

In order better to grasp Changing Pattern and the interaction relationship thereof of the Various types of data such as extraterrestrial target, space environment of roomage state system senses, need to develop empty Incorporate situation and express engine, to arrive greatly whole solar each major planet, little of tellurian Yi Ge land station and high resolving power landform around thereof, the Various types of data such as visible extraterrestrial target and sightless space environment are unified to show.Realize the naturally smooth and easy displaying of all kinds of situation data, what first need solution is exactly viewpoint control problem, and the quality that viewpoint is controlled directly affects user and situation expressed to the experience of engine.

Viewpoint is unique interface that user observes Visual Scene, is the fundamental of virtual reality.For simple Visual Scene, show, the function that general using developing instrument provides itself can meet the demands, as the gluLookAt () function of OpenGL.The viewpoint control method that needs more to have " feeling of immersion " for complicated Visual Scene, as the viewpoint control method of the Leica Virtual Explorer systems such as GoogleEarth, WorldWind, what the viewpoint of this class visualization engine was controlled mainly solution is how better to realize ground roaming, viewpoint be take the earth all the time as focus, can be rotated, convergent-divergent and being rotated around ground point, near-earth when roaming local horizon is level all the time, cannot tilt; Also have the emphasis of a class as STK in the visualization engine of extraterrestrial target emulation, the target that this engine is paid close attention to user is placed in the center of visual field always, target is rotated and convergent-divergent, be not suitable for browsing of ground scene, viewpoint can be switched between different target, but switch, adopt Direct Transform mode, there is no the process of conversion, very lofty; An other class is the interplanetary visualization engine as Celestia, this class visualization engine viewpoint can be carried out smooth switching between different target, but its focus of mainly paying close attention to is all kinds of astronomical targets, when viewpoint furthers earth isosphere surface, there will be local horizon to tilt, cannot be around problems such as sphere point are rotated, thus the viewing experience of ground surface or terrain affected.A few class viewpoint control methods above all cannot well meet the demand that empty Incorporate situation is expressed engine.

The principal feature of expressing engine due to empty Incorporate situation has that data type is many, space scale and variable in distance large, the feature such as relate to more than coordinate system, in order to make the viewpoint control method of design can better adapt to empty Incorporate situation, express, each principal feature of below empty Incorporate situation being expressed is carried out detailed analysis.

1.1 data types are many

Roomage state perception data is the expression object that sky Incorporate situation is expressed engine, roomage state perception data numerous and complicated.According to the type of Monitoring Data, perception data can be divided into two large classes, and the one, extraterrestrial target data, comprise all targets on Earth's orbit: all kinds of big or small targets such as moving satellite, inert satellite, space junk in-orbit; The 2nd, space environment data, comprise the Various types of data such as solar activity related data, terrestrial magnetic field, ionosphere, space radiation band, high energy particle.Except perception data, also having class data is data of roomage state sensory perceptual system itself, and as ground and the distribution situation of space-based surveillance own, running status etc., these class data can be classified as a class with extraterrestrial target data from expression-form.The geographic information data that other class data are earth itself and whole solar each major planet data and star background data.Formed thus the basic data of a complete situation expression system that comprises the sky world, above in these data, partial data itself is intuitively visible, as data such as extraterrestrial targets, these data will be reproduced really, partial data itself is abstract sightless, therefore need to carry out Visualization to these data, make sightless become visible.So various data need to be unified to show under same framework, and this controls and brought unprecedented difficulty for viewpoint.Fig. 1 is the nomogram that sky Incorporate situation is expressed engine.

1.2 space scales and variable in distance are large

Empty Incorporate situation express engine except data type how, also having an important feature is exactly to need object space yardstick and the variable in distance of expression huge, to radius in the whole solar system, be 696 greatly, the maximum target sun of 300 kms is little of one several meters large police radars on earth surface; As far as apart from sun Neptune farthest from the sun 4,504,300,000 km, Mercury nearest to the sun also has 57,910,000 km, how distance between ground two targets can realize so Precise Representation of the target of large scale span closely in one meter, and the seamless switching between the target of so large distance is the gordian technique that must solve.

1.3 to relate to coordinate system many

All kinds of targets that empty Incorporate situation is expressed in engine are not simple stacks, but according to strict time-space relationship model, allow target in the corresponding time in correct place, must be based on they corresponding coordinate systems to the description of all kinds of Target space positions.Because the data class of roomage state perception is various, related to a plurality of subjects, and in different subjects, the description of research object has been had to conventional coordinate system, conventional heliocentric ecliptic coordinate system during as the motion of each major planet in describing the solar system; While describing the satellite motion of a certain planet, conventionally adopt the inertial coordinates system of this planet, the description of satellite itself is adopted to satellite body coordinate system; While describing the fixed target on planetary surface, adopt the fixed coordinate system of this planet; For the description of space environment data, as radiation belt often adopts geomagnetic coordinate system, magnetic field of the earth data are accustomed to adopting the earth's core solar magnetospheric system of coordinates.

As can be seen here, the coordinate system that relates to is various is the another feature that sky Incorporate situation is expressed engine, will be relating to so data unified displaying together of multi-coordinate, what must solve is the mutual transfer problem between these coordinate systems, the reference coordinate of the comprehensive all kinds of Situation Awareness data of the present invention has built coordinate transformation relation figure as shown in Figure 2 for this reason, this transformational relation has covered reference coordinate conventional within the scope of the solar system, all kinds of orbit coordinates relevant to extraterrestrial target, body coordinate, the geocentrical various magnetosphere reference coordinates relevant to space environment etc.Making in order as viewpoint of coordinate relation controlled and laid a good foundation.

It is various that empty Incorporate situation is expressed the expressed data type of engine, and space scale and distance be large, it is many to relate to coordinate system, and viewpoint is controlled will reach satisfied object, must design for these features.Mainly there are the following problems for prior art:

Problem one, the Visualization that has comprised all kinds of Situation Awareness data such as planet, spacecraft, ground surface or terrain, space environment due to empty Incorporate engine, if expressing in engine, situation only adopt a kind of viewpoint controlling party rule of browsing objective to be difficult to meet the demands, as the viewpoint control method of browsing spacecraft there will be while using browsing of ground surface or terrain local horizon tilt, cannot be around the problem of a series of affect users experience such as landform focus is rotated

Problem two, comprehensively analyzes viewpoint control method, and browsing of viewpoint mainly contains two class actions: a class is around a direction Vector Rotation certain angle, and an other class is by scale value, to carry out displacement while carrying out convergent-divergent.Express in engine take the situation that desktop computer is platform, viewpoint is controlled and mainly by mouse interactive interface, is realized, generally, directly mouse amount of movement and roller rolling amount are mapped to the mutual control that can realize viewpoint in the anglec of rotation of viewpoint and scale value, this method mapping relations are simple, be convenient to realize, mouse is moving how many, viewpoint is with regard to the corresponding angle of corresponding rotation, roller turns how many, and viewpoint is how many with regard to corresponding convergent-divergent, and the conversion of mouse stops viewpoint change and just stops at once, this mutual effect is stiff, there is no innervation.

Problem three, in the time viewpoint need to being switched to another one target from the target of current concern, can adopt Immediate Mode, and user arranges after new associated objects, occurs immediately new associated target in scene.This viewpoint switch mode realizes simply, but lack the transition of viewpoint from a position to another one position, seems more stiff, not lively, and while being especially switched to the earth satellite from the whole solar system, it is comparatively lofty that this method seems.

Problem four, the large feature that empty Incorporate situation is expressed engine is exactly that space-time span is huge, the meticulous expression that wants to realize all kinds of extraterrestrial targets under large scale like this just must overcome due to the excessive jitter problem causing of floating number round-off error, shake can cause some part of object by uncertain cutting, triangle gridding break, the series of problems such as texture entanglement.

Summary of the invention

The object of this invention is to provide a kind of empty Incorporate situation and express engine and classify and grading target browsing method thereof, in order to address the above problem one the single viewpoint control method of employing, affect the problem that user experiences.

For achieving the above object, the solution of the present invention comprises:

Empty Incorporate situation is expressed the classify and grading target browsing method of engine, comprises the steps:

Target classification: judgement target belongs to ball target and still belongs to non-ball target; If target belongs to non-ball target, adopt general viewpoint control method;

Apart from classification: if ball target, judgement is that High eye point is browsed or nearly sphere is browsed, if High eye point is browsed, adopts general viewpoint control method; If nearly sphere is browsed, adopt special viewpoint control method; General viewpoint control method comprises: around the viewpoint control method that target is rotated, convergent-divergent and direction of visual lines are controlled;

Special viewpoint control method comprises: the viewpoint control method of local horizon level, viewpoint are not passed through the viewpoint control method of landform, the viewpoint control method that viewpoint is rotated around the place ahead point.

General viewpoint control method and special viewpoint control method also comprise the viewpoint control method with inertia, and the described viewpoint control method with inertia is as follows:

$V_{\mathrm{angle}}^n=V_{\mathrm{angle}}^{n-1}\×D_{\mathrm{angle}}$

$V_{\mathrm{factor}}^n=V_{\mathrm{factor}}^{n-1}\×D_{\mathrm{factor}}---\left(11\right)$

V _angle=F(△x,△y)

V _factor=G(△d) （9）

In formula (9), V _anglethe angular velocity of rotation that direction vector is determined in viewpoint winding, V _factorit is the scaling speed of viewpoint; F (△ x, △ y) is that (△ x, △ y) is to V _angle, the mapping function set; G (△ d) is that △ d is to V _factor, the mapping function set, △ x and △ y are the vertical rotary amount of movement that horizontally rotates amount of movement and input of input, the convergent-divergent rolling amount of △ d input; In formula (11)

the angular velocity of rotation of current scene frame,

the angular velocity of rotation of a upper scene frame,

the scaling speed of present frame,

the scaling speed of previous frame, D _angle, D _factorbe respectively damping due to rotation coefficient and convergent-divergent ratio of damping, their value is between 0 to 1.

Empty Incorporate situation is expressed engine, comprising: target classification module: judgement target belongs to ball target and still belongs to non-ball target; If target belongs to non-ball target, adopt general viewpoint control method; Apart from diversity module: if ball target, judgement is that High eye point is browsed or nearly sphere is browsed, if High eye point is browsed, adopts general viewpoint control method; If nearly sphere is browsed, adopt special viewpoint control method; General viewpoint control method comprises: around the viewpoint control method that target is rotated, convergent-divergent and direction of visual lines are controlled; Special viewpoint control method comprises: the viewpoint control method of local horizon level, viewpoint are not passed through the viewpoint control method of landform, the viewpoint control method that viewpoint is rotated around the place ahead point.

General viewpoint control method and special viewpoint control method also comprise the viewpoint control method with inertia, and the described viewpoint control method with inertia is as follows:

$V_{\mathrm{angle}}^n=V_{\mathrm{angle}}^{n-1}\×D_{\mathrm{angle}}$

$V_{\mathrm{factor}}^n=V_{\mathrm{factor}}^{n-1}\×D_{\mathrm{factor}}---\left(11\right)$

V _angle=F(△x,△y)

V _factor=G(△d) （9）

In formula (9), V _anglethe angular velocity of rotation that direction vector is determined in viewpoint winding, V _factorit is the scaling speed of viewpoint; F (△ x, △ y) is that (△ x, △ y) is to V _angle, the mapping function set; G (△ d) is that △ d is to V _factor, the mapping function set, △ x and △ y are the vertical rotary amount of movement that horizontally rotates amount of movement and input of input, the convergent-divergent rolling amount of △ d input; In formula (11)

the angular velocity of rotation of current scene frame,

the angular velocity of rotation of a upper scene frame,

the scaling speed of present frame, the scaling speed of previous frame, D _angle, D _factorbe respectively damping due to rotation coefficient and convergent-divergent ratio of damping, their value is between 0 to 1.

If expressing in engine, situation only adopt a kind of viewpoint controlling party rule of browsing objective to be difficult to meet the demands, as the viewpoint control method of browsing spacecraft there will be while using browsing of ground surface or terrain local horizon to tilt, cannot be around the problem of a series of users' of impact experience such as landform focus is rotated, the comprehensive geometrical property of analyzing Various types of data in situation expression engine, as shown in Figure 4, the present invention adopts the viewpoint control method of classify and grading to carry out browsing of realize target, expression data is divided into ball and non-ball, the reason that is divided into this two class is: for ball targets such as planets, except normal rotation, convergent-divergent, outside the operation such as translation, the special requirements such as inside that also have viewpoint can not enter planet.For ball target, because carrying out nearly sphere, browse by needs, also need viewpoint control method to carry out classification processing, viewpoint adopts different viewpoint control methods when distance objective different distance.

Although the present invention carries out classify and grading to viewpoint control method, but the viewpoint control method of different target classification is some general implementation method still, the one, around target, rotate, the 2nd, convergent-divergent, the 3rd, direction of visual lines is controlled, above-mentioned three kinds of viewpoint control methods all belong to prior art, and embodiment is shown in its introduction.

When nearly sphere is browsed, during as the browse displays of Earth's Terrain, in order to allow scene browse the experience of life that more meets people, viewpoint is controlled and had some special requirements, such as, guarantee that local horizon is level, viewpoint can not be passed through the landform of fluctuating, viewpoint can be around certain some rotation in the place ahead etc., their difference are the viewpoint control method of horizontal line level accordingly, and viewpoint is not passed through the viewpoint control method of landform, and viewpoint is around the viewpoint control method of the place ahead point rotation, this several method also belongs to prior art, and embodiment is shown in its introduction.

Further, in order to make operating interactive more have innervation, can also adopt the viewpoint control method with inertia, because rotational speed is along with the frame number of playing up declines gradually, set rational ratio of damping, when input stops transformation, velocity amplitude does not also become 0, corresponding viewpoint change also will carry out, until speed is 0 to stop transformation, this viewpoint control mode, can have the sensation of inertia to a kind of target of people from scene viewing experience, so the present invention calls this mapping mode with the viewpoint of inertia and controls.As for determining of damping coefficient, mainly according to user's demand and experience, determine, the present invention finds, when ratio of damping is got 0.92 left and right, can obtain good viewing experience by practice.

Accompanying drawing explanation

Fig. 1 is nomogram;

Fig. 2 is space coordinate conversion relation;

Fig. 3 is the tree-shaped structural drawing of roomage state data;

Fig. 4 is classify and grading viewpoint control method;

Fig. 5 is the horizontal direction viewpoint schematic diagram that rotates;

Fig. 6 is eye coordinate conversion schematic diagram;

Fig. 7 is viewpoint local coordinate system define method schematic diagram;

Fig. 8 is continuous viewpoint conversion regime schematic diagram;

Fig. 9 is spherical linear interpolation schematic diagram;

Figure 10 is that target is at the schematic diagram of viewpoint and heliocentric ecliptic coordinate system;

Figure 11 is the frustum schematic diagram of perspective projection;

Figure 12 is that perspective dynamic state of parameters is set schematic diagram;

Figure 13 is solar system global view;

Figure 14 is Mars view;

Figure 15 is global TEC distribution view;

Figure 16 is underground browse view;

Figure 17 is gps satellite view.

Embodiment

Below in conjunction with accompanying drawing, the present invention will be further described in detail.

Empty Incorporate situation is expressed engine, by terrestrial space, the farther moon, planet and even ionosphere, radiating layer etc. are visual, comprise that one of them is greatly to the whole solar system, little of ground Yi Ge land station, and land station's landform around, can catalogue to extraterrestrial target, all kinds of situation data such as space environment unify the space environment simulation platform of showing simultaneously.

On the basis of existing technology, in order to solve the problems of pointing out in background technology, the present invention has proposed respectively multiple solution.Following embodiment has only provided the corresponding method of expression engine.

The invention provides a kind of classify and grading target browsing method

For solving situation expression engine, only adopt a kind of viewpoint controlling party rule of browsing objective to be difficult to meet the demands, affect the problem that user experiences, comprise that step is as follows:

Target classification: judgement target belongs to ball target and still belongs to non-ball target; If target belongs to non-ball target, adopt general viewpoint control method;

Apart from classification: if ball target, judgement is that High eye point is browsed or nearly sphere is browsed, if High eye point is browsed, adopts general viewpoint control method; If nearly sphere is browsed, adopt special viewpoint control method;

General viewpoint control method comprises: around the viewpoint control method that target is rotated, convergent-divergent and direction of visual lines are controlled; Special viewpoint control method comprises: the viewpoint control method of local horizon level, viewpoint are not passed through the viewpoint control method of landform, the viewpoint control method that viewpoint is rotated around the place ahead point.

Lower mask body is introduced above-mentioned various viewpoint control method.

Around target, rotate

As shown in Figure 5, at the residing coordinate system O of object observing ₁in-XYZ, viewpoint is rotated after λ degree in the horizontal direction, and viewpoint P rotates to P ' position in figure after Z axis rotates λ degree, and transformation relation is:

P′=R _z(λ)P （1）

In formula

$R_z\left(\mathrm{\λ}\right)=\left[\begin{array}{ccc}\cos \mathrm{\λ}& -\sin \mathrm{\λ}& 0\\ \sin \mathrm{\λ}& \cos \mathrm{\λ}& 0\\ 0& 0& 1\end{array}\right]---\left(2\right)$

In like manner can obtain viewpoint in the vertical direction and rotate around object observing, the transformation for mula rotating around X-axis is:

P′=R _x(λ)P （3）

In formula

$R_x\left(\mathrm{\λ}\right)=\left[\begin{array}{ccc}1& 0& 0\\ 0& \cos \mathrm{\λ}& -\sin \mathrm{\λ}\\ 0& \sin \mathrm{\λ}& \cos \mathrm{\λ}\end{array}\right]---\left(4\right)$

Convergent-divergent is controlled

Object observing is carried out to convergent-divergent can and be realized away from object observing by viewpoint shift-in, and the convergent-divergent motion process of viewpoint can be summarized as:

P′=factor·P （5）

Wherein P is the original position of viewpoint, position after P ' viewpoint convergent-divergent, factor is zoom factor, and factor>0, when factor>l, realization be to dwindle effect; When factor<1, realization be amplification effect.

Direction of visual lines is controlled

In situation of the present invention, express in engine, for guaranteeing that observer can see object observing, must guarantee that the Z-direction of eye coordinate is pointed to object observing.As shown in Figure 6: O ₂-XYZ coordinate system is eye coordinate.For guaranteeing that Z axis points to object observing, eye coordinate is first around rotate-H of Z axis angle, and then by coordinate system around rotate-p of X-axis angle, now Z axis sensing object observing.

After direction of visual lines conversion above, object observing will be placed in field of view center, now, eye coordinate is rotated and can realizes observed object in vertical and translation effect horizontal direction around X-axis and Y-axis respectively again, around Z axis, rotates and can realize whole scene around the rotate effect at viewpoint center.

When above viewpoint control method can be for the High eye point of ball target and the viewpoint of other non-ball target control, the key distinction for ball target and non-ball target is: when ball target is browsed, viewpoint does not generally enter spheroid inside, and the control method of the viewpoint control method that adopts while browsing of nearly sphere while being different from High eye point.

For ball target, the method that prevents viewpoint target approach inside is when convergent-divergent is controlled, to add following Rule of judgment:

|△P′|>R

（6）

Wherein | △ P ' | represent that viewpoint is to the distance of target's center, R is the radius of sphere target.

When nearly sphere is browsed, during as the browse displays of Earth's Terrain, in order to allow scene browse the experience of life that more meets people, viewpoint is controlled and had some special requirements, such as, guarantee that local horizon is level, viewpoint can not be passed through the landform of fluctuating, viewpoint can be around certain some rotation in the place ahead etc., their difference are the viewpoint control method of horizontal line level accordingly, viewpoint is not passed through the viewpoint control method of landform, viewpoint is around the viewpoint control method of the place ahead point rotation, for without loss of generality, below just to take our the familiar earth be example, introduce in detail respectively the concrete viewpoint control method of the specific (special) requirements when realizing these nearly spheres and browsing:

Local horizon level

First define a local coordinate system O ' _g-X ' Y ' Z ', the define method of this coordinate system as shown in Figure 7, with the line O in viewpoint center and the earth's core _oo _eintersection point G is true origin with ground, and the direction that the earth's core orders to G of take is Z ' direction of principal axis, and the direct north at G point place of take is Y ' direction of principal axis, and the due east direction at G point place of take is X ' direction of principal axis.

Local horizon, in level in visual field, must guarantee eye coordinate O when allowing near the ground browsing _o-X _oy _oz _ox _oaxle and local coordinate system Z ' axle are all the time in plumbness, and now, the pitching of viewpoint can be passed through around X _othe rotation of axle realizes, and the left and right of viewpoint is browsed and only need be allowed eye coordinate O _o-X _oy _oz _oz ' direction of principal axis around local coordinate system is rotated and can realizes.

Viewpoint is not passed through landform

Realize viewpoint and do not pass through landform, except meeting the Rule of judgment of formula (6), when near the ground browsing, also must calculate in real time the sea level elevation of viewpoint position, guarantee that its height value, higher than the ground point height value on corresponding longitude and latitude, meets the condition of formula (7):

H _o>H _g （7）

H wherein _orepresent sea level elevation corresponding to viewpoint position, H _grepresent the corresponding topocentric height value of longitude and latitude identical with viewpoint.

Viewpoint is rotated around the place ahead point

In the landform of carrying out near the ground, browse, when running into interested impact point, during as a certain seat mountain, a certain seat building, may need this terrain object to carry out around observation, in order to realize this function, the present invention be take point of interest and is set up a local coordinate system, the define method of this coordinate system and the coordinate system O ' of preceding office portion as initial point _g-X ' Y ' Z ' is identical, and just coordinate origin is different.Establish after local coordinate system, in this coordinate system, adopt the represented spinning solution of formula (1) to carry out viewpoint position conversion, allow viewpoint carry out posture changing around the Z axis rotation of local coordinate system, can complete like this viewpoint and rotate around point of interest, conventionally get the intersection point of sight line and the earth as point of interest.

The present invention also provides a kind of viewpoint control method with inertia.

The comprehensive viewpoint control method of analyzing, browsing of viewpoint mainly contains two class actions: a class is around a direction Vector Rotation certain angle, and an other class is by scale value, to carry out displacement while carrying out convergent-divergent.△ x and △ y are the vertical rotary amount of movement that horizontally rotates amount of movement and input of input, the convergent-divergent rolling amount of △ d input, the situation that is platform at desktop computer is expressed in engine, viewpoint is controlled and mainly by mouse interactive interface, is realized, the two class actions that the movement of mouse and the rolling of mouse roller are controlled corresponding to viewpoint.When mouse moves, be mainly that the position of mouse changes, in the present embodiment, input equipment is mouse, and △ x and △ y are respectively the horizontal and vertical amount of movement of mouse on screen, and △ d is roller corresponding convergent-divergent rolling amount while rolling.In prior art, directly mouse amount of movement and roller rolling amount are mapped to the mutual control that can realize viewpoint in the anglec of rotation of viewpoint and scale value, adopt formula (8) to shine upon.

Angle=F(△x,△y)

factor=G(△d) （8）

In formula Angle be viewpoint around the anglec of rotation of a direction vector, F (△ x, △ y) is (△ x, △ y) to the mapping function of Angle; Scale value when factor is viewpoint convergent-divergent, G (△ d) is that △ d is to the mapping function of factor.

Above this method, mapping relations are simple, be convenient to realize, but mutual effect are stiff, there is no innervation, mouse is moving how many, and viewpoint is with regard to the corresponding angle of corresponding rotation, roller turns how many, and viewpoint is how many with regard to corresponding convergent-divergent, and the conversion of mouse stops viewpoint change and just stops at once.More natural in order to allow scene browse, there is innervation, as the formula (9), the present invention is mapped to mouse amount of movement and roller rolling amount on the angular velocity of rotation and scaling speed of viewpoint, rather than in the original anglec of rotation and zoom factor, that is to say, in formula (9), V _anglethe angular velocity of rotation that direction vector is determined in viewpoint winding, V _factorit is the scaling speed of viewpoint; F (△ x, △ y) is that (△ x, △ y) is to V _angle, the mapping function set; G (△ d) is that △ d is to V _factor, the mapping function set.

V _angle=F(△x,△y)

V _factor=G(△d) （9）

The corresponding anglec of rotation and scale value can represent with following formula:

Angle=V _angle×△t

factor=V _factor×△t （10）

In formula (10), △ t gets the time interval of drawing adjacent two frame scenes.After mouse moves, through type (9) can be determined the angular velocity of viewpoint rotation, through type (10) can determine that current scene frame, with respect to the viewpoint anglec of rotation of a upper scene frame, only can't complete viewpoint by this two step and control, because each mouse definite angular velocity of rotation V while moving _anglebe a constant, viewpoint will be rotated down always like this, and for this reason, the present invention introduces ratio of damping D, before every frame scene drawing, velocity amplitude be adopted to damping conversion, be shown below:

$V_{\mathrm{angle}}^n=V_{\mathrm{angle}}^{n-1}\&times;D_{\mathrm{angle}}$

$V_{\mathrm{factor}}^n=V_{\mathrm{factor}}^{n-1}\&times;D_{\mathrm{factor}}---\left(11\right)$

In formula (11)

the angular velocity of rotation of current scene frame,

the angular velocity of rotation of a upper scene frame,

the scaling speed of present frame,

the scaling speed of previous frame, D _angle, D _factorvalue between 0 to 1.Pass through like this playing up of multiframe, angular velocity of rotation will move closer in 0, thereby stops the rotation conversion.

Because rotational speed is along with the frame number of playing up declines gradually, set rational ratio of damping, when mouse stops transformation, velocity amplitude does not also become 0, corresponding viewpoint change also will carry out, until speed is 0 to stop transformation, and this viewpoint control mode, from scene viewing experience, can there is the sensation of inertia to a kind of target of people, so the present invention calls this mapping mode with the viewpoint of inertia and controls.As for determining of damping coefficient, mainly according to user's demand and experience, determine, the present invention finds, when ratio of damping is got 0.92 left and right, can obtain good viewing experience by practice.The conversion process of convergent-divergent and rotation are similar, repeat no more here.

The present invention also provide a kind of viewpoint across yardstick seamless handover method.

In the time viewpoint need to being switched to another one target from the target of current concern, can adopt Immediate Mode, user arranges after new associated objects, occurs immediately new associated target in scene.This viewpoint switch mode realizes simply, but lack the transition of viewpoint from a position to another one position, seems more stiff, not lively, and while being especially switched to the earth satellite from the whole solar system, it is comparatively lofty that this method seems.In order to give very natural seamless switching process of user, the present invention improves viewpoint switch mode, except providing Immediate Mode switching mode, continuous mode viewpoint switch is also provided, be that viewpoint does not directly arrive target location, but from current location, move to target location along certain path.The span of expressing engine due to situation of the present invention is larger, has comprised whole solar Situation Awareness target, so the present invention's employing is along the conversion regime of straight line, can either reach the effect of seamless switching, is convenient to again realize.Conversion regime along straight line arrives appointed place in the given time along straight line path continuous transformation.Straight linear interpolation formula is comparatively simple, as shown in the formula:

$P=\frac{B-A}{|B-A|}\&times;\mathrm{Dis}---\left(12\right)$

In formula (12), P is interpolation point position vector, and A is the start position vector of viewpoint switch, and B is the final position vector of viewpoint switch, and Dis is the air line distance that interpolation is put starting point.When view distance target is far away, adopt along straight line conversion regime.Far refer to: as from the solar system overall situation to each celestial body, more closely refer to: the situation of advancing around the earth, such as viewpoint is transferred to U.S. overhead from Over China.

When view distance target nearer, as from earth overhead a bit to another of the earth overhead time, if still adopt the mode of straight line, in viewpoint transfer process, be easy to through the earth, conversion effect is had a greatly reduced quality (as shown in Figure 8, dotted line represents through the earth), the invention provides another along the conversion regime of sphere for this reason, adopt the problem that can avoid passing through the earth along the mode of sphere conversion.

Realize sphere roaming, must use spherical linear interpolation techniques, this interpolation is carried out along the shortest arc line on sphere between two key points.As shown in Figure 9, on a two dimensional surface, vectorial A and vectorial B angle are ω, and vectorial P and vectorial A angle are θ, and P can be obtained by the spherical linear interpolation between A, B:

P=αA+βB （13）

Given three conditions below we can be in the hope of α and β:

$\left\{\begin{array}{c}\left|P\right|=1\\ A\&CenterDot;P=\cos \mathrm{\&theta;}\\ A\&CenterDot;B=\cos \mathrm{\&omega;}\end{array}\right.---\left(14\right)$

Obtain thus:

$P=\frac{\sin (\mathrm{\&omega;}-\mathrm{\&theta;})}{\sin \mathrm{\&omega;}}A+\frac{\sin \mathrm{\&theta;}}{\sin \mathrm{\&omega;}}B---\left(15\right)$

By being used in combination of viewpoint control of the present invention and view point switching method, can realize comprehensive observation and seamless switching that empty Incorporate situation is expressed all kinds of targets in engine.

The present invention also provides a kind of jitter eliminating method.

The large feature that empty Incorporate situation of the present invention is expressed engine is exactly that space-time span is huge, the meticulous expression that wants to realize all kinds of extraterrestrial targets under large scale like this just must overcome due to the excessive jitter problem causing of floating number round-off error, shake can cause some part of object by uncertain cutting, triangle gridding break, the series of problems such as texture entanglement.

In integrated expression engine, because the yardstick of expressed target has very large gap, little only have the land station of several meters to ground, reaches 1,392 greatly to diameter, and the sun of 000 km will be eliminated the setting that shake just relates to the perspective projection parameter of d engine.It shown in Figure 11, is the frustum in perspective projection, frustum is the shape after pyramid is truncated, frustum region is the viewing area in three-dimensional scenic, image video card can be nearly cutting face in frustum between 0 to 1 in to the range mappings between cutting face far away to depth buffer district when carrying out Image Rendering, and the depth value of any object between frustum has all become certain number between 0 to 1.If nearly cutting face is larger to the distance between cutting face far away, because round-off error can cause shake equally.The target size difference that empty Incorporate situation expression engine of the present invention is expressed is huge, only set and when a unified perspective projection parameter that can be suitable for all yardsticks (nearly cutting face and cutting face far away) must cause too small objective expression, occur the phenomenons such as shake, therefore the present invention adopts the dynamic setting method of perspective projection parameter, according to the size of different targets, set nearly cutting face and the cutting face far away of perspective projection, guarantee that both can completely was expressed target, can guarantee precision again.

Concrete method to set up, as Figure 12, is first calculated the radius R of the outer maximum circumsphere of object and the distance dis that target's center arrives viewpoint center, and so back-pushed-type (21) calculates nearly cutting face zNear and cutting face zFar far away:

zNear=dis-R

ZFar=dis+R （21）

Further, because shake also relates to floating number problem, so mainly contain at present two kinds of thinkings in processing floating number problem: the one, by more high-precision floating number, reduce round-off error, but the GPU of main flow is 32 floating numbers at present, even if use double-precision floating point type data when calculate outside, when importing GPU into and show, still can be truncated into the data of single precision.Therefore be mainly to utilize another thinking to solve this problem at present, utilize less floating number to carry out object drafting, this thinking according to being the feature that near floating-point initial point has less round-off error and higher precision.The anti-jitter of realizing the second thinking must utilize local coordinate to realize, be about to the large floating number coordinate conversion of target from heliocentric ecliptic coordinate system shows in the local coordinate of little floating number, the demand that all kinds of planet Situation Awareness datum targets in comprehensive empty Incorporate and the earth show, the present invention adopts the local coordinate of initial point centered by viewpoint to realize jitter elimination.

As shown in figure 10, S-XYZ is the coordinate system that adopts large floating number, in empty Incorporate engine, is heliocentric ecliptic coordinate system, and O-X ' Y ' Z ' is viewpoint local coordinate system, and P point is impact point, establishes the coordinate of P point in S-XYZ for (X, Y, Z) ^t, the coordinate of P point in O-X ' Y ' Z ' is (△ X, △ Y, △ Z) ^t, observed object by day heart ecliptic coordinate to transform to the formula of coordinate of viewpoint local coordinate system as follows:

$\left[\begin{array}{c}\mathrm{\&Delta;X}\\ \mathrm{\&Delta;Y}\\ \mathrm{\&Delta;Z}\end{array}\right]=M\left[\begin{array}{c}X\\ Y\\ Z\end{array}\right]+T---\left(20\right)$

M in formula (20), T is respectively viewpoint local coordinate and ties up to rotation and the transposed matrix under heliocentric ecliptic coordinate system.

Calculate target after viewpoint local coordinate system, then take viewpoint as initial point carries out the drafting of target, can realize the object that replaces large floating number with less floating number, thereby complete the effect of anti-jitter.Comprehensive utilization two kinds of jitter eliminating methods above, can realize good anti-jitter effect, obtain the situation of natural reality more and express.

Terrain rendering under integrated frame

Empty Incorporate situation is expressed engine and the maximum of existing engine and is distinguished and be that gound-mapping is differentiated in the whole world to be integrated into and to have accurate time-space relationship more, in the middle of the unified platform that comprises the sun, each major planet, all kinds of roomage state perception datas.The pyramid models of differentiating of general employing are at present organized global terrain data more, this technology is first pre-stored by longitude and latitude layering different layer and piece the whole world, when viewpoint moves, to depending on Intrapyramidal terrain block, from internal memory, not unload, by new terrain block, replace, when viewpoint to the earth near time, use meticulous terrain block to replace coarse terrain block, away from time, with coarse landform is fast, replace meticulous terrain block, finally realize the Fast rendering of global terrain data.Method is in the past mainly for ground visual needs, do not consider the earth autobiography, revolution and with the relation of other planets, data dispatch only need to be considered the concrete coordinate of each landform tile under ground is admittedly when playing up.For method in the past can be applied under integrated expression framework of the present invention, must be by the coordinate conversion of original earth coordinates to the lower coordinate of day heart ecliptic system while carrying out data dispatch.Current data and method of partition are generally to take longitude and latitude as unit, first must be transformed into earth fixed coordinate system, then be transformed into earth inertial coordinates system, be finally transformed into heliocentric ecliptic coordinate system and realize the accurate drafting of corresponding terrain data in whole empty Incorporate situation displaying scene.

The present invention also provides the viewpoint method to set up that a kind of target is relevant.

Empty Incorporate situation is expressed engine finally will become the system that can show all kinds of visible and sightless roomage state perception datas of the whole solar system, although the data class of expressing is various, space scale and change in location are huge, but are not irregular following.From time-space relationship, we can find that Various types of data has a father node, as each major planet, around the sun, make revolution motion, and the sun can be regarded the father node of each major planet as; The moon and Earth's orbit spacecraft are around earth movements, and the earth can be regarded the father node of the moon and all kinds of Earth's orbit spacecrafts as; The ground target such as land station, although the earth is static relatively, the earth still can be regarded the father node of these targets as, these corresponding targets are exactly the child node of the earth.The rest may be inferred, whole solar Various types of data just can be organized with a tree structure as shown in Figure 3, thereby form an orderly situation and express scene, in this scene tree, each spatial alternation of father node can be delivered to child node, in conjunction with coordinate system transformation relation above, situation express each destination node of scene if guarantee it and father node between spatial alternation relation be correct, can guarantee that whole roomage state expresses accurately expressing in order of each Target space position in engine.

Viewpoint can be regarded a virtual observer in three-dimensional scenic as, if allow this observer be placed in scene according to the absolute position in space, although also can realize the observation to scene, but in the whole solar system, all kinds of planets and extraterrestrial target are all in the middle of continuous motion, if realize the relative static of observer and extraterrestrial target, observer must experience the spatial alternation the same with extraterrestrial target, so not only wastes computational resource, and easily makes mistakes.In order to overcome this problem, the present invention expresses observer all kinds of targets with situation are associated, observer is used as to an extraterrestrial target, for observer specifies a father node, in the time of need to paying close attention to like this which target, just specify the father node that this target is observer, observer is only responsible for and the space conversion of paying close attention to target, as relatively static problem above, as long as guarantee that observer's invariant position in father node fixed coordinate system all can solve easily.Now, by changing position and the attitude of observer in father node, can realize and browsing paying close attention to the arbitrarily angled of target.Simultaneously according to the observation person's association be the coordinate system type of father node, can also realize some special results of browse, as observer is associated with the inertial coordinates system of the earth, the spinning motion of the observable earth, the fixed coordinate system that is associated with the earth can be observed the effect that the stars change in positions.

According to said method, we have developed empty Incorporate situation and have expressed engine, if Figure 13 is viewpoint and the sun solar global view while being associated, have shown the circumsolar ruuning situation of each major planet in figure.By view point switching method of the present invention, viewpoint naturally can be transitioned into any planet, if Figure 14 is that viewpoint switch is near view Mars.Figure 15 is the view that viewpoint switch distributes to Earthwatch whole world TEC.Figure 16 is the effect of viewpoint when landform is browsed in somewhere at the earth's surface, and the DEM resolution of figure mesorelief is 8.5 meters, and image respectively rate is 1 meter.Figure 17 is viewpoint view while being associated with gps satellite, and lines are tracks of this satellite.

Claims (4)

1. empty Incorporate situation is expressed the classify and grading target browsing method of engine, it is characterized in that, comprises the steps:

Target classification: judgement target belongs to ball target and still belongs to non-ball target; If target belongs to non-ball target, adopt general viewpoint control method;

Apart from classification: if ball target, judgement is that High eye point is browsed or nearly sphere is browsed, if High eye point is browsed, adopts general viewpoint control method; If nearly sphere is browsed, adopt special viewpoint control method; General viewpoint control method comprises: around the viewpoint control method that target is rotated, convergent-divergent and direction of visual lines are controlled;

Special viewpoint control method comprises: the viewpoint control method of local horizon level, viewpoint are not passed through the viewpoint control method of landform, the viewpoint control method that viewpoint is rotated around the place ahead point.

2. empty Incorporate situation according to claim 1 is expressed the classify and grading target browsing method of engine, it is characterized in that, general viewpoint control method and special viewpoint control method also comprise the viewpoint control method with inertia, and the described viewpoint control method with inertia is as follows:

$V_{\mathrm{angle}}^n=V_{\mathrm{angle}}^{n-1}\&times;D_{\mathrm{angle}}$

$V_{\mathrm{factor}}^n=V_{\mathrm{factor}}^{n-1}\&times;D_{\mathrm{factor}}---\left(11\right)$

V _angle=F(△x,△y)

V _factor=G(△d) （9）

In formula (9), V _anglethe angular velocity of rotation that direction vector is determined in viewpoint winding, V _factorit is the scaling speed of viewpoint; F (△ x, △ y) is that (△ x, △ y) is to V _angle, the mapping function set; G (△ d) is that △ d is to V _factor, the mapping function set, △ x and △ y are the vertical rotary amount of movement that horizontally rotates amount of movement and input of input, the convergent-divergent rolling amount of △ d input; In formula (11)

the angular velocity of rotation of current scene frame,

the angular velocity of rotation of a upper scene frame,

the scaling speed of present frame,

the scaling speed of previous frame, D _angle, D _factorbe respectively damping due to rotation coefficient and convergent-divergent ratio of damping, their value is between 0 to 1.

3. the classify and grading target browsing method of empty Incorporate situation expression engine described in employing claim 1, is characterized in that, comprising:

Target classification module: judgement target belongs to ball target and still belongs to non-ball target; If target belongs to non-ball target,

Adopt general viewpoint control method;

Apart from diversity module: if ball target, judgement is that High eye point is browsed or nearly sphere is browsed, if High eye point is browsed, adopts general viewpoint control method; If nearly sphere is browsed, adopt special viewpoint control method;

General viewpoint control method comprises: around the viewpoint control method that target is rotated, convergent-divergent and direction of visual lines are controlled;

Special viewpoint control method comprises: the viewpoint control method of local horizon level, viewpoint are not passed through the viewpoint control method of landform, the viewpoint control method that viewpoint is rotated around the place ahead point.

4. empty Incorporate situation is expressed the classify and grading target browsing method of engine according to claim 3, it is characterized in that, general viewpoint control method and special viewpoint control method also comprise the viewpoint control method with inertia, and the described viewpoint control method with inertia is as follows:

$V_{\mathrm{angle}}^n=V_{\mathrm{angle}}^{n-1}\&times;D_{\mathrm{angle}}$

$V_{\mathrm{factor}}^n=V_{\mathrm{factor}}^{n-1}\&times;D_{\mathrm{factor}}---\left(11\right)$

V _angle=F(△x,△y)

V _factor=G(△d) （9）

In formula (9), V _anglethe angular velocity of rotation that direction vector is determined in viewpoint winding, V _factorit is the scaling speed of viewpoint; F (△ x, △ y) is that (△ x, △ y) is to V _angle, the mapping function set; G (△ d) is that △ d is to V _factor, the mapping function set, △ x and △ y are the vertical rotary amount of movement that horizontally rotates amount of movement and input of input, the convergent-divergent rolling amount of △ d input; In formula (11)

the angular velocity of rotation of current scene frame,

the angular velocity of rotation of a upper scene frame,

the scaling speed of present frame,

the scaling speed of previous frame, D _angle, D _factorbe respectively damping due to rotation coefficient and convergent-divergent ratio of damping, their value is between 0 to 1.

Images