阅读说明：本技术 一种真空环境下的分离体飞行运动参数多目视觉测量系统 (Multi-view vision measurement system for flight motion parameters of separating body in vacuum environment ) 是由 高越 丁沛 龙雪 李红薇 赵卓茂 胡鑫 薛锋 于杨 李越 黄强 蔡劭佳 于 2020-12-03 设计创作，主要内容包括：本发明涉及一种真空环境下分离体飞行运动参数多目视觉测量系统,其使用多台高速相机分别拍摄被测试件飞行区域的不同部分,通过对所有相机画面进行融合计算,可以获得试件在任一时刻的飞行运动参数。该测量系统测量精度仅与单台相机拍摄视场有关,能够根据实际拍摄视场大小,在多个观察窗口合理设置相机个数及拍摄区域,从而实现在较大试件飞行区域情况下实现全程高精度测量。(The invention relates to a multi-view vision measuring system for flight motion parameters of a separating body in a vacuum environment. The measurement accuracy of the measurement system is only related to the shooting field of view of a single camera, the number of the cameras and the shooting area can be reasonably set in a plurality of observation windows according to the size of the actual shooting field of view, and therefore the whole-course high-accuracy measurement can be achieved under the condition of a large test piece flight area.)

1. A separation body flight motion parameter multi-view vision measuring system under a vacuum environment is characterized by comprising:

-a high-speed camera for photographing different flight areas of the test piece;

-adjusting means for adjusting the position and orientation of the high speed camera;

-camera markers enabling an image processing algorithm to identify the image coordinate positions of the separate bodies in the image;

-calibration means to calibrate parameters of the high speed cameras including lens focal length, lens distortion, relative position and relative orientation between the respective high speed cameras;

-a light source providing high brightness illumination for a high speed camera;

the control center comprises a synchronous controller and an image processing controller, the synchronous controller enables the high-speed camera to be synchronously exposed, and the image processing controller performs fusion calculation on the pictures of the high-speed camera to obtain the motion parameters of the test piece.

2. The system according to claim 1, wherein the photographic marks are speckle photographic marks having higher recognition accuracy than cross-mark photographic marks.

3. A system for multi-vision measurement of flying parameters of a separated body under vacuum environment as claimed in claim 1, wherein said calibration means is a calibration plate, and the pattern of said calibration plate comprises chessboard grids and dot grids.

4. The system of claim 1, further comprising a baffle, a test piece and a tool thereof, and a recovery bracket arranged in the vacuum tank, wherein the test piece and the tool thereof are arranged at one end of the vacuum tank, the baffle is arranged at the rear side of the test piece and the tool thereof, the recovery bracket is arranged at the position of the other end of the vacuum tank opposite to the test piece and the tool thereof, and the test piece is connected to a control center outside the vacuum tank through a cable.

5. The system according to claim 1, wherein the number of the light sources is three, the number of the high-speed cameras is five, a first light source, a fourth high-speed camera and a fifth high-speed camera are arranged in the vacuum tank, the first light source is arranged between the baffle and the recycling bracket, and the fourth high-speed camera and the fifth high-speed camera are respectively arranged on two sides of the first light source; a second light source, a third light source, a first high-speed camera, a second high-speed camera and a third high-speed camera are arranged outside the vacuum tank; no. two light sources and No. three light sources are close to respectively the baffle with retrieve the support setting, set gradually high-speed camera, No. two high-speed cameras, No. three high-speed cameras between No. two light sources to No. three light sources.

6. The system of claim 5, wherein the high-speed cameras I, II and III are all in oblique shooting mode, and the high-speed cameras I, II and III are all in flat shooting mode.

7. The system of claim 6, wherein the high-speed camera is a Vision Research Phantom V12.1 high-speed camera with a resolution of 1280X 800 pixels, a full-frame maximum shooting speed of 6200 frames/sec, and the lens is a Zeiss dispagon lens, wherein 3 flat-shot high-speed cameras are 21mm fixed-focus lenses, and 2 oblique-shot high-speed cameras are 35mm fixed-focus lenses.

8. A multi-view vision measuring system for flight motion parameters of a separation body under a vacuum environment as claimed in claim 7, wherein five high-speed cameras are located at different positions and different viewing angles, the measuring process is divided into a plurality of flight areas, the test piece can be observed in each flight area only by at least two of the five high-speed cameras, and the observation results in each flight area are subjected to fusion calculation to represent the complete motion process of the test piece.

9. The system for multi-vision measurement of flying motion parameters of a separation body in a vacuum environment as claimed in claim 8, wherein five high-speed cameras are used to shoot the surface measuring points of the test piece from different angles, and according to the perspective imaging principle of the high-speed cameras, the measuring points are located in the ith camera S_iHas a two-dimensional coordinate of p in the image_i(x_i,y_i) I is a natural number from 1 to 5, which is associated with the three-dimensional coordinate P (X) of the measuring point in physical space_w,Y_w,Z_w) There is the following relationship, where w is the three-dimensional coordinate in physical space:

in the above formula: the result of the multiplication on the left side of the formula is a 3x1 vector,the third term is vector and represents the Z coordinate value of the measured point in the optical center coordinate system of the high-speed camera; f. ofⁱIs the focal length of the high-speed camera,physical size in x and y directions for a single pixel of a camera light sensor, fⁱAndcoupling states in mathematical models, commonly usedReplacing;is the coordinates of the center of the high-speed camera image,the internal parameters of the high-speed camera are only related to the mechanical structure inside the high-speed camera and the lens used during shooting; rotation matrixAnd translation vectorThe method comprises the following steps of (1) coordinate transformation parameters of a physical space three-dimensional coordinate system and a camera optical center coordinate system, wherein a rotation matrix only has three independent parameters, and the rotation matrix and a translation vector are external parameters of a camera and only relate to the position and the orientation of the high-speed camera during shooting; the internal parameters and the external parameters are obtained by a high-speed camera parameter calibration method before the test; if the two-dimensional coordinates (x) of the measuring points in the N high-speed camera images are obtained by an image processing methodⁱ,yⁱ) Then, N equations can be listed, in order to calculate the three-dimensional coordinate P (X) of the measured point_w,Y_w,Z_w) And (3) ensuring N (N is more than or equal to 2), and obtaining the unique solution of the equation set by a least square method after the N equations are combined.

10. The system according to claim 9, wherein the matching of the images of the plurality of high-speed cameras is realized by a digital image correlation method, specifically: selecting an image subregion from a reference image, searching an optimal image subregion from a target image according to a correlation coefficient calculation formula in the following formula, and considering the two image subregions as most similar subregions when the correlation coefficients of the two subregions reach a minimum value, thereby completing an image matching process, wherein the correlation coefficient calculation formula is as follows:

wherein M is the size of the image subarea, M ranges from 25 to 35, and f_a(x, y) and f_a(x, y) are the image brightness values of the reference image and the target image at point (x, y), respectively, C_coorFor the correlation coefficient, the smaller the correlation coefficient, the higher the similarity between the two subregions.

Technical Field

The invention relates to a multi-view vision measurement system for flight motion parameters of a separating body in a vacuum environment, and belongs to the field of vacuum tests.

Background

During the on-orbit operation of the space station, a separation body needs to be released to complete a specific accompanying flight task, and the separation body has to meet the following two requirements in design: (1) because the space station is flying in the space orbit when the separating body is released, the atmospheric pressure is close to vacuum, and the whole space station is in a weightless state, the separating body must be normally released in a vacuum weightless environment. (2) After the release of the separating body, it is necessary to keep a certain distance from the space station to complete the subsequent action, and therefore, the releasing of the separating body needs to have a certain speed.

In order to assess the reliability and motion stability of a separating body released in a vacuum weightlessness environment, most of the prior vacuum cabin weightlessness release tests are adopted, and the test process is as follows: and placing the release device of the test piece on a falling platform in the vacuum chamber, closing the vacuum chamber, and pumping out air to the designated atmospheric pressure. During the test, the control equipment is started, so that the test piece and the falling platform fall integrally, and the separating body is released in the falling process. And shooting the release process of the separating body through a small observation window outside the vacuum chamber by using a high-speed camera to obtain the flight process and the motion parameters of the separating body. Such vision-based flight motion parameter measurement methods are typically based on monocular and binocular vision measurements. The monocular vision measurement is to shoot the flying motion process of the separating body from the vertical flying direction by using a single camera, the method firstly linearly marks the size of each pixel of the camera corresponding to the actual space by using simple marking tools such as a shooting scale and the like, then obtains a separating body image according to shooting, and calculates the motion parameters of the separating body image in the flying plane. The binocular vision measuring method comprises the steps of utilizing two cameras to shoot a flying process of a separating body at a certain angle, firstly, calibrating internal parameters such as focal length, image center and lens distortion of the two cameras and external parameters such as relative position and relative orientation by means of calibration tools such as a shooting calibration plate, then matching image coordinates of measuring points of the separating body according to images shot by the two cameras, and calculating three-dimensional motion parameters of the separating body. The method is more complex than a monocular vision measurement process, but has better measurement precision, and can obtain a three-dimensional result. In the two methods, the resolution and the view field size of the camera play a crucial role in the final measurement accuracy, that is, the accuracy is worse when the shooting view field is larger under the condition of a certain resolution of the camera. In addition, because the observation window of the vacuum test chamber is small, the observation field of view of the camera is limited, each observation window can only allow one camera to shoot a part of the test piece flight area, and the motion parameters of the whole flight process of the separation body cannot be measured in one test.

Disclosure of Invention

Technical problem to be solved

Aiming at the defects and requirements in the prior art, the invention provides a multi-view vision measurement system for flight motion parameters of a separating body in a vacuum environment. The measurement accuracy of the measurement system is only related to the shooting field of view of a single camera, the number of the cameras and the shooting area can be reasonably set in a plurality of observation windows according to the size of the actual shooting field of view, and therefore the whole-course high-accuracy measurement can be achieved under the condition of a large test piece flight area. The measurement system is successfully applied to a vacuum weightlessness release test of a certain type of equipment separating body, and fully meets the test requirements.

(II) technical scheme

A separation body flight motion parameter multi-view vision measuring system under a vacuum environment comprises:

-a high-speed camera for photographing different flight areas of the test piece;

-adjusting means for adjusting the position and orientation of the high speed camera;

-camera markers enabling an image processing algorithm to identify the image coordinate positions of the separate bodies in the image;

-calibration means to calibrate parameters of the high speed cameras including lens focal length, lens distortion, relative position and relative orientation between the respective high speed cameras;

-a light source providing high brightness illumination for a high speed camera;

the control center comprises a synchronous controller and an image processing controller, the synchronous controller enables the high-speed camera to be synchronously exposed, and the image processing controller performs fusion calculation on the pictures of the high-speed camera to obtain the motion parameters of the test piece.

The photographic mark adopts a speckle photographic mark, and has higher identification precision than a cross mark photographic mark.

The calibration device is a calibration plate, and the pattern of the calibration plate comprises chessboard grids and dot grids.

The multi-view vision measurement system further comprises a baffle, a test piece, a tool and a recovery support which are arranged in the vacuum tank, the test piece and the tool are arranged at one end of the vacuum tank, the baffle is arranged on the rear side of the test piece and the tool, the recovery support is arranged at the position, opposite to the test piece and the tool, of the other end of the vacuum tank, and the test piece is connected to a control center outside the vacuum tank through a cable.

The high-speed cameras comprise five, a first light source, a fourth high-speed camera and a fifth high-speed camera are arranged in the vacuum tank, the first light source is arranged between the baffle and the recovery bracket, and the fourth high-speed camera and the fifth high-speed camera are respectively arranged on two sides of the first light source; a second light source, a third light source, a first high-speed camera, a second high-speed camera and a third high-speed camera are arranged outside the vacuum tank; no. two light sources and No. three light sources are close to respectively the baffle with retrieve the support setting, set gradually high-speed camera, No. two high-speed cameras, No. three high-speed cameras between No. two light sources to No. three light sources.

The fourth high-speed camera and the fifth high-speed camera are in oblique shooting type, and the first high-speed camera, the second high-speed camera and the third high-speed camera are in flat shooting type.

The high-speed camera is a Phantom V12.1 high-speed camera of Vision Research company, the resolution of the high-speed camera is 1280X 800 pixels, the maximum shooting speed of a full picture is 6200 frames/second, a Distagon lens of ZEISS company is selected as the lens, wherein 3 flat-shooting type high-speed cameras adopt 21mm fixed-focus lenses, and 2 oblique-shooting type high-speed cameras adopt 35mm fixed-focus lenses.

The positions and the visual angles of the five high-speed cameras are different, the measuring process is divided into a plurality of flight areas, the test piece can be observed in each flight area only by at least two of the five high-speed cameras, and the observation results in each flight area are subjected to fusion calculation to represent the complete motion process of the test piece.

Five high-speed cameras are used for shooting the measuring point on the surface of the test piece from different angles, and according to the perspective imaging principle of the high-speed cameras, the measuring point is arranged in the ith camera S_iHas a two-dimensional coordinate of p in the image_i(x_i,y_i) I is a natural number from 1 to 5, which is associated with the three-dimensional coordinate P (X) of the measuring point in physical space_w,Y_w,Z_w) There is the following relationship, where w is the three-dimensional coordinate in physical space:

in the above formula: the result of the multiplication on the left side of the formula is a 3x1 vector,the third term is vector and represents the Z coordinate value of the measured point in the optical center coordinate system of the high-speed camera; f. ofⁱIs the focal length of the high-speed camera,physical size in x and y directions for a single pixel of a camera light sensor, fⁱAndcoupled states in the mathematical modelIs usually usedReplacing;is the coordinates of the center of the high-speed camera image,the internal parameters of the high-speed camera are only related to the mechanical structure inside the high-speed camera and the lens used during shooting; rotation matrixAnd translation vectorThe method comprises the following steps of (1) coordinate transformation parameters of a physical space three-dimensional coordinate system and a camera optical center coordinate system, wherein a rotation matrix only has three independent parameters, and the rotation matrix and a translation vector are external parameters of a camera and only relate to the position and the orientation of the high-speed camera during shooting; the internal parameters and the external parameters are obtained by a high-speed camera parameter calibration method before the test; if the two-dimensional coordinates of the measuring points in the N high-speed camera images are obtained by an image processing methodThen N of the above equations can be listed for calculating the three-dimensional coordinates P (X) of the measured point_w,Y_w,Z_w) And (3) ensuring N (N is more than or equal to 2), and obtaining the unique solution of the equation set by a least square method after the N equations are combined.

The method comprises the following steps of (1) realizing image matching of a plurality of high-speed cameras by adopting a digital image correlation method, specifically: selecting an image subregion from a reference image, searching an optimal image subregion from a target image according to a correlation coefficient calculation formula in the following formula, and considering the two image subregions as most similar subregions when the correlation coefficients of the two subregions reach a minimum value, thereby completing an image matching process, wherein the correlation coefficient calculation formula is as follows:

wherein M is the size of the image subarea, M ranges from 25 to 35, and f_a(x, y) and f_a(x, y) are the image brightness values of the reference image and the target image at point (x, y), respectively, C_coorFor the correlation coefficient, the smaller the correlation coefficient, the higher the similarity between the two subregions.

(III) advantageous effects

The measuring system can realize the accurate measurement of the flight motion parameters released by the separating body in the vacuum weightlessness environment and can complete the measurement tasks under various release speeds. The measurement system completes the vacuum weightlessness release performance assessment test of a certain type of separating body, provides powerful support for type development work, and the experimental result shows that: the system has the shooting speed of 6000 frames/second, the measurement view field of 3m (X direction) multiplied by 2m (Y direction) multiplied by 0.5m (Z direction), the displacement measurement precision of 1mm and the highest measurement test piece movement speed of 50 m/s. The precision of the measuring system is only related to the shooting field of view of a single camera, the number of the cameras and the shooting area can be reasonably set in a plurality of observation windows according to the size of the actual shooting field of view, and therefore the whole-course high-precision measurement can be achieved under the condition of a large test piece flight area. The method can be applied to flight motion parameter/deformation measurement tests of various separating bodies in an expanded way by adjusting the number of the cameras and the shooting mode.

Drawings

FIG. 1 is a schematic view of a multi-view visual measurement system for flight motion parameters of a separation body in a vacuum environment.

Fig. 2 is a schematic diagram of a multi-vision measurement.

In the figure, 1-baffle; 2-test piece and its tooling; 3-recovering the scaffold; 4-high speed camera number four; 5-light source number one; 6-five high speed camera; 7-light source number two; 8-high speed camera number one; 9-high speed camera number two; 10-high speed camera No. three; 11-light source number three; 12-a cable; 13-control center.

Detailed Description

The invention relates to a multi-view vision measuring system for flight motion parameters of a separating body in a vacuum environment, which comprises:

-a high-speed camera for photographing different flight areas of the test piece;

-adjusting means for adjusting the position and orientation of the high speed camera;

-camera markers enabling an image processing algorithm to identify the image coordinate positions of the separate bodies in the image;

-calibration means to calibrate parameters of the high speed cameras including lens focal length, lens distortion, relative position and relative orientation between the respective high speed cameras;

-a light source providing high brightness illumination for a high speed camera;

a control center 13, including a synchronous controller and an image processing controller, wherein the synchronous controller makes the high-speed camera synchronously exposed, and the image processing controller performs fusion calculation on the pictures of the high-speed camera to obtain the motion parameters of the test piece.

The photographic mark adopts a speckle photographic mark, and has higher identification precision than a cross mark photographic mark.

The calibration device is a calibration plate, and the pattern of the calibration plate comprises chessboard grids and dot grids.

The multi-view vision measurement system further comprises a baffle 1, a test piece and a tool 2 thereof and a recovery support 3 which are arranged in the vacuum tank, the test piece and the tool 2 thereof are arranged at one end of the vacuum tank, the baffle 1 is arranged on the rear side of the test piece and the tool 2 thereof, the recovery support 3 is arranged at the position, opposite to the test piece and the tool 2 thereof, of the other end of the vacuum tank, and the test piece is connected to a control center 13 outside the vacuum tank through a cable.

The three light sources are included, the high-speed cameras comprise five high-speed cameras, a first light source 5, a fourth high-speed camera 4 and a fifth high-speed camera 6 are arranged in the vacuum tank, the first light source 5 is arranged between the baffle plate 1 and the recovery support 3, and the fourth high-speed camera 4 and the fifth high-speed camera 6 are respectively arranged on two sides of the first light source 5; a second light source 7, a third light source 11, a first high-speed camera 8, a second high-speed camera 9 and a third high-speed camera 10 are arranged outside the vacuum tank; no. two light sources 7 and No. three light source 11 are close to respectively baffle 1 with retrieve the setting of support 3, set gradually high-speed camera 8, No. two high-speed camera 9, No. three high-speed camera 10 between No. two light sources 7 to No. three light sources 11.

The fourth high-speed camera 4 and the fifth high-speed camera 6 are in oblique shooting type, and the first high-speed camera 8, the second high-speed camera 9 and the third high-speed camera 10 are in flat shooting type.

The high-speed camera is a Phantom V12.1 high-speed camera of Vision Research company, the resolution of the high-speed camera is 1280X 800 pixels, the maximum shooting speed of a full picture is 6200 frames/second, a Distagon lens of ZEISS company is selected as the lens, wherein 3 flat-shooting type high-speed cameras adopt 21mm fixed-focus lenses, and 2 oblique-shooting type high-speed cameras adopt 35mm fixed-focus lenses.

The positions and the visual angles of the five high-speed cameras are different, the measuring process is divided into a plurality of flight areas, the test piece can be observed in each flight area only by at least two of the five high-speed cameras, and the observation results in each flight area are subjected to fusion calculation to represent the complete motion process of the test piece.

Five high-speed cameras are used for shooting a measuring point on the surface of the tested piece from different angles, and according to the perspective imaging principle of the high-speed cameras, the measuring point is positioned in the ith camera S_iHas a two-dimensional coordinate of p in the image_i(x_iAnd yi) i is a natural number from 1 to 5, and is a three-dimensional coordinate P (X) of the measuring point in the physical space_w,Y_w,Z_w) There is the following relationship, where w is the three-dimensional coordinate in physical space:

in the above formula: the result of the multiplication on the left side of the formula is a 3x1 vector,the third term is vector and represents the Z coordinate value of the measured point in the optical center coordinate system of the high-speed camera; f. ofⁱIs the focal length of the high-speed camera,physical size in x and y directions for a single pixel of a camera light sensor, fⁱAndcoupling states in mathematical models, commonly usedReplacing;is the coordinates of the center of the high-speed camera image,the internal parameters of the high-speed camera are only related to the mechanical structure inside the high-speed camera and the lens used during shooting; rotation matrixAnd translation vectorThe coordinate transformation parameters of a physical space three-dimensional coordinate system and a camera optical center coordinate system are adopted, wherein the rotation matrix has only three independent parameters, and the rotation matrix and the translation vector are external parameters of the camera and are only related to the position and the orientation of the high-speed camera during shooting. The internal parameters and the external parameters are obtained by a high-speed camera parameter calibration method before the test. If two-dimensional coordinates (x) of the measuring points in N (in this embodiment, N is 5) high-speed camera images are obtained by an image processing methodⁱ,yⁱ) Then, N equations can be listed, in order to calculate the three-dimensional coordinate P (X) of the measured point_w,Y_w,Z_w) And (3) ensuring N (N is more than or equal to 2), and obtaining the unique solution of the equation set by a least square method after the N equations are combined.

The method for realizing multi-camera image matching by adopting a digital image correlation method comprises the following specific steps: selecting an image subregion from a reference image, searching an optimal image subregion from a target image according to a correlation coefficient calculation formula in the following formula, and considering the two image subregions as most similar subregions when the correlation coefficients of the two subregions reach a minimum value, thereby completing an image matching process, wherein the correlation coefficient calculation formula is as follows:

wherein M is the size of the image subarea, M ranges from 25 to 35, and f_a(x, y) and f_a(x, y) are the image brightness values of the reference image and the target image at point (x, y), respectively, C_coorFor the correlation coefficient, the smaller the correlation coefficient, the higher the similarity between the two subregions.