CN103954221B

CN103954221B - The binocular photogrammetric survey method of large-size pliable structure vibration displacement

Info

Publication number: CN103954221B
Application number: CN201410191705.6A
Authority: CN
Inventors: 王聪; 许畅; 高晶波; 张春芳
Original assignee: Harbin Institute of Technology Shenzhen
Current assignee: Harbin Institute of Technology Shenzhen
Priority date: 2014-05-08
Filing date: 2014-05-08
Publication date: 2016-08-17
Anticipated expiration: 2034-05-08
Also published as: CN103954221A

Abstract

The invention discloses a binocular photogrammetry method for the vibration displacement of a large flexible structure, which belongs to the technical field of dynamic analysis and measurement of flexible structures. The invention aims to solve the problem that the existing contact measurement will affect the performance of the flexible structure itself. It first calibrates the industrial camera; then excites the flexible structure under test, and at the same time controls all industrial cameras to trigger synchronously through the single-chip microcomputer to take pictures of the vibration displacement of the measured point of the flexible structure under test; then processes the images captured by all industrial cameras, and uses The corner detection algorithm obtains the pixel coordinates of the measuring point, and then uses the binocular 3D reconstruction technology to obtain the object space coordinates of the measuring point at each moment, so as to obtain the vibration displacement curve of the measured flexible structure. The invention is used for the measurement of the vibration displacement of the flexible structure.

Description

Binocular Photogrammetry Method for Vibration Displacement of Large Flexible Structure

技术领域technical field

本发明涉及大型柔性结构振动位移的双目摄影测量方法，属于柔性结构动力学分析和测量技术领域。The invention relates to a binocular photogrammetry method for vibration displacement of a large flexible structure, and belongs to the technical field of dynamic analysis and measurement of flexible structures.

背景技术Background technique

双目摄像测量的应用主要有对轿车三维外形进行高精度检测，风洞实验中结构的变形，大型天线型面精度测量，地型测绘等领域。这些测量采用的摄像三维重构技术大部分用于静态物体的三维特征测量，而无法对物体的振动进行测量。对于柔性结构的振动，采用传统的加速度传感器的接触式测量方式，不但会对柔性结构的动力学性能带来影响，还由于测量设备复杂，造成误差难以控制及噪声影响较大。The application of binocular camera measurement mainly includes the high-precision detection of the three-dimensional shape of the car, the deformation of the structure in the wind tunnel experiment, the precision measurement of the large antenna surface, and the terrain surveying and mapping. Most of the camera 3D reconstruction techniques used in these measurements are used for 3D feature measurement of static objects, but cannot measure the vibration of objects. For the vibration of the flexible structure, the traditional contact measurement method of the acceleration sensor will not only affect the dynamic performance of the flexible structure, but also cause the error to be difficult to control and the noise to be greatly affected due to the complexity of the measuring equipment.

发明内容Contents of the invention

本发明目的是为了解决现有接触式测量会对柔性结构本身性能造成影响的问题，提供了一种大型柔性结构振动位移的双目摄影测量方法。The purpose of the invention is to solve the problem that the existing contact measurement will affect the performance of the flexible structure itself, and provides a binocular photogrammetry method for the vibration displacement of a large flexible structure.

本发明所述大型柔性结构振动位移的双目摄影测量方法，它包括以下步骤：The binocular photogrammetry method of the large flexible structure vibration displacement of the present invention, it comprises the following steps:

步骤一：标定工业相机，所述工业相机的数量为2-5个；Step 1: Calibrate the industrial camera, the number of the industrial camera is 2-5;

步骤二：激振被测柔性结构，同时通过单片机控制所有工业相机同步触发，进行被测柔性结构被测点振动位移的拍摄；Step 2: Excite the flexible structure under test, and at the same time, control all industrial cameras to trigger synchronously through the single-chip microcomputer, and shoot the vibration displacement of the measured point of the flexible structure under test;

步骤三：对所有工业相机拍摄获得的图像进行处理，利用角点检测算法获得测点的像素坐标，再利用双目三维重构技术获得每一时刻测点的物方坐标，从而得到被测柔性结构的振动位移曲线。Step 3: Process the images captured by all industrial cameras, use the corner detection algorithm to obtain the pixel coordinates of the measuring point, and then use the binocular 3D reconstruction technology to obtain the object space coordinates of the measuring point at each moment, so as to obtain the measured flexible The vibration displacement curve of the structure.

标定工业相机的具体方法为：The specific method of calibrating the industrial camera is as follows:

采用标定板对工业相机进行标定，所述标定板为由黑白相间的正方形构成的7×9的棋盘格，正方形边长为5mm；设置标定板的位置，使每一个工业相机的视场范围覆盖完整的标定板，并采用工业相机拍摄标定板的静止图像，完成标定。The calibration board is used to calibrate the industrial camera. The calibration board is a 7×9 checkerboard consisting of black and white squares, and the side length of the square is 5mm; the position of the calibration board is set so that the field of view of each industrial camera covers Complete the calibration board, and take a still image of the calibration board with an industrial camera to complete the calibration.

工业相机的数量为2个，步骤三中得到被测柔性结构的振动位移曲线的具体方法为：The number of industrial cameras is 2, and the specific method for obtaining the vibration displacement curve of the measured flexible structure in step 3 is:

由2个工业相机同时拍摄被测柔性结构的被测点，对拍摄获得的图像进行处理后，建立方程组如下：The measured points of the flexible structure to be tested are photographed by two industrial cameras at the same time. After processing the captured images, the equations are established as follows:

$\{\begin{matrix} {Z Z}_{c c 11} [\begin{matrix} {u u}_{11} \\ {v v}_{11} \\ 11 \end{matrix}] = = {N N}_{11} {M m}_{11} [\begin{matrix} {X x}_{w w} \\ {Y Y}_{w w} \\ {Z Z}_{w w} \\ 11 \end{matrix}] = = {Q Q}_{11} [\begin{matrix} {X x}_{w w} \\ {Y Y}_{w w} \\ {Z Z}_{w w} \\ 11 \end{matrix}] \\ {Z Z}_{c c 22} [\begin{matrix} {u u}_{22} \\ {v v}_{22} \\ 11 \end{matrix}] = = {N N}_{22} {M m}_{22} [\begin{matrix} {X x}_{w w} \\ {Y Y}_{w w} \\ {Z Z}_{w w} \\ 11 \end{matrix}] = = {Q Q}_{22} [\begin{matrix} {X x}_{w w} \\ {Y Y}_{w w} \\ {Z Z}_{w w} \\ 11 \end{matrix}] \end{matrix},,$

式中，Z_c1为第一工业相机在其光心坐标系下，测点P的Z轴坐标值，(u₁,v₁)为测点P的像素坐标值，(X_w,Y_w,Z_w)为测点P的物方坐标值，N₁为第一工业相机的内参数矩阵，M₁为第一工业相机的外参数矩阵，Q₁为第一工业相机的投影矩阵，Q₁＝N₁M₁；In the formula, Z _c1 is the Z-axis coordinate value of the measuring point P under the optical center coordinate system of the first industrial camera, (u ₁ , v ₁ ) is the pixel coordinate value of the measuring point P, (X _w , Y _w , Z _w ) is the object space coordinate value of the measuring point P, N ₁ is the internal parameter matrix of the first industrial camera, M ₁ is the external parameter matrix of the first industrial camera, Q ₁ is the projection matrix of the first industrial camera, Q ₁ = N ₁ M ₁ ;

Z_c2为第二工业相机在其光心坐标系下，测点P的Z轴坐标值，(u₂,v₂)为测点P的像素坐标值，(X_w,Y_w,Z_w)为测点P的物方坐标值，N₂为第二工业相机的内参数矩阵，M₂为第二工业相机的外参数矩阵，Q₂为第二工业相机的投影矩阵，Q₂＝N₂M₂；Z _c2 is the Z-axis coordinate value of the measuring point P under the optical center coordinate system of the second industrial camera, (u ₂ , v ₂ ) is the pixel coordinate value of the measuring point P, (X _w , Y _w , Z _w ) is the object space coordinate value of measuring point P, N ₂ is the internal parameter matrix of the second industrial camera, M ₂ is the external parameter matrix of the second industrial camera, Q ₂ is the projection matrix of the second industrial camera, Q ₂ =N ₂ _M2 ;

上述光心坐标系表示以工业相机的光心为原点，光轴为Z轴的坐标系；The above-mentioned optical center coordinate system represents a coordinate system with the optical center of the industrial camera as the origin and the optical axis as the Z axis;

对上述方程组进行处理，获得超定方程：The above equations are processed to obtain the overdetermined equation:

$\{\begin{matrix} {Q Q}^{11}_{1111} {X x}_{w w} + + {Q Q}^{11}_{1212} {Y Y}_{w w} + + {Q Q}^{11}_{1313} {Z Z}_{w w} + + {Q Q}^{11}_{1414} - - {u u}_{11} {Q Q}^{11}_{3131} {X x}_{w w} - - {u u}_{11} {Q Q}^{11}_{3232} {Y Y}_{w w} - - {u u}_{11} {Q Q}^{11}_{3333} {Z Z}_{w w} = = {u u}_{11} {Q Q}^{11}_{3434} \\ {Q Q}^{11}_{21 twenty one} {X x}_{w w} + + {Q Q}^{11}_{22 twenty two} {Y Y}_{w w} + + {Q Q}^{11}_{23 twenty three} {Z Z}_{w w} + + {Q Q}^{11}_{24 twenty four} - - {v v}_{11} {Q Q}^{11}_{3131} {X x}_{w w} - - {v v}_{11} {Q Q}^{11}_{3232} {Y Y}_{w w} - - {v v}_{11} {Q Q}^{11}_{3333} {Z Z}_{w w} = = {v v}_{11} {Q Q}^{11}_{3434} \\ {Q Q}^{22}_{1111} {X x}_{w w} + + {Q Q}^{22}_{1212} {Y Y}_{w w} + + {Q Q}^{22}_{1313} {Z Z}_{w w} + + {Q Q}^{22}_{1414} - - {u u}_{22} {Q Q}^{22}_{3131} {X x}_{w w} - - {u u}_{22} {Q Q}^{22}_{3232} {Y Y}_{w w} - - {u u}_{22} {Q Q}^{22}_{3333} {Z Z}_{w w} = = {u u}_{22} {Q Q}^{22}_{3434} \\ {Q Q}^{22}_{21 twenty one} {X x}_{w w} + + {Q Q}^{22}_{22 twenty two} {Y Y}_{w w} + + {Q Q}^{22}_{23 twenty three} {Z Z}_{w w} + + {Q Q}^{22}_{24 twenty four} - - {v v}_{22} {Q Q}^{22}_{3131} {X x}_{w w} - - {v v}_{22} {Q Q}^{22}_{3232} {Y Y}_{w w} - - {v v}_{22} {Q Q}^{22}_{3333} {Z Z}_{w w} = = {v v}_{22} {Q Q}^{22}_{3434} \end{matrix},,$

式中Q¹ _ij表示Q₁的第i行第j列，i＝1，2，3，j＝1，2，3，4；In the formula, Q ¹ _ij represents the i-th row and j-th column of Q ₁ , i=1, 2, 3, j=1, 2, 3, 4;

Q² _ij表示Q₂的第i行第j列，i＝1，2，3，j＝1，2，3，4；Q ² _ij represents the i-th row and j-th column of Q ₂ , i=1, 2, 3, j=1, 2, 3, 4;

整理超定方程，得到：After rearranging the overdetermined equations, we get:

$[\begin{matrix} {Q Q}^{11}_{1111} - - {u u}_{11} {Q Q}^{11}_{3131} & {Q Q}^{11}_{1212} - - {u u}_{11} {Q Q}^{11}_{3232} & {Q Q}^{11}_{1313} - - {u u}_{11} {Q Q}^{11}_{3333} \\ {Q Q}^{11}_{21 twenty one} - - {v v}_{11} {Q Q}^{11}_{3131} & {Q Q}^{11}_{22 twenty two} - - {v v}_{11} {Q Q}^{11}_{3232} & {Q Q}^{11}_{23 twenty three} - - {v v}_{11} {Q Q}^{11}_{3333} \\ {Q Q}^{22}_{1111} - - {u u}_{22} {Q Q}^{22}_{3131} & {Q Q}^{22}_{1212} - - {u u}_{22} {Q Q}^{22}_{3232} & {Q Q}^{22}_{1313} - - {u u}_{22} {Q Q}^{22}_{3333} \\ {Q Q}^{22}_{21 twenty one} - - {v v}_{22} {Q Q}^{22}_{3131} & {Q Q}^{22}_{22 twenty two} - - {v v}_{22} {Q Q}^{22}_{3232} & {Q Q}^{22}_{23 twenty three} - - {v v}_{22} {Q Q}^{22}_{3333} \end{matrix}] [\begin{matrix} {X x}_{w w} \\ {Y Y}_{w w} \\ {Z Z}_{w w} \end{matrix}] = = [\begin{matrix} {u u}_{11} {Q Q}^{11}_{3434} - - {Q Q}^{11}_{1414} \\ {v v}_{11} {Q Q}^{11}_{3434} - - {Q Q}^{11}_{24 twenty four} \\ {u u}_{22} {Q Q}^{22}_{3434} - - {Q Q}^{22}_{1414} \\ {v v}_{22} {Q Q}^{22}_{3434} - - {Q Q}^{22}_{24 twenty four} \end{matrix}],,$

设定 $K = [\begin{matrix} {Q^{1}}_{11} - u_{1} {Q^{1}}_{31} & {Q^{1}}_{12} - u_{1} {Q^{1}}_{32} & {Q^{1}}_{13} - u_{1} {Q^{1}}_{33} \\ {Q^{1}}_{21} - v_{1} {Q^{1}}_{31} & {Q^{1}}_{22} - v_{1} {Q^{1}}_{32} & {Q^{1}}_{23} - v_{1} {Q^{1}}_{33} \\ {Q^{2}}_{11} - u_{2} {Q^{2}}_{31} & {Q^{2}}_{12} - u_{2} {Q^{2}}_{32} & {Q^{2}}_{13} - u_{2} {Q^{2}}_{33} \\ {Q^{2}}_{21} - v_{2} {Q^{2}}_{31} & {Q^{2}}_{22} - v_{2} {Q^{2}}_{32} & {Q^{2}}_{23} - v_{2} {Q^{2}}_{33} \end{matrix}],$ set up $K = [\begin{matrix} {Q^{1}}_{11} - u_{1} {Q^{1}}_{31} & {Q^{1}}_{12} - u_{1} {Q^{1}}_{32} & {Q^{1}}_{13} - u_{1} {Q^{1}}_{33} \\ {Q^{1}}_{twenty one} - v_{1} {Q^{1}}_{31} & {Q^{1}}_{twenty two} - v_{1} {Q^{1}}_{32} & {Q^{1}}_{twenty three} - v_{1} {Q^{1}}_{33} \\ {Q^{2}}_{11} - u_{2} {Q^{2}}_{31} & {Q^{2}}_{12} - u_{2} {Q^{2}}_{32} & {Q^{2}}_{13} - u_{2} {Q^{2}}_{33} \\ {Q^{2}}_{twenty one} - v_{2} {Q^{2}}_{31} & {Q^{2}}_{twenty two} - v_{2} {Q^{2}}_{32} & {Q^{2}}_{twenty three} - v_{2} {Q^{2}}_{33} \end{matrix}],$

$h h = = [\begin{matrix} {X x}_{w w} \\ {Y Y}_{w w} \\ {Z Z}_{w w} \end{matrix}],, U u = = [\begin{matrix} {u u}_{11} {Q Q}^{11}_{3434} - - {Q Q}^{11}_{1414} \\ {v v}_{11} {Q Q}^{11}_{3434} - - {Q Q}^{11}_{24 twenty four} \\ {u u}_{22} {Q Q}^{22}_{3434} - - {Q Q}^{22}_{1414} \\ {v v}_{22} {Q Q}^{22}_{3434} - - {Q Q}^{22}_{24 twenty four} \end{matrix}],,$

则上式变形为Kh＝U，Then the above formula is transformed into Kh=U,

求解获得h的最小二乘解：Solve to obtain the least squares solution for h:

h＝(K^TK)^-1K^TU，h=(K ^T K) ^-1 K ^T U,

由此求解获得随时间变化的测点P的物方坐标值，进而获得被测柔性结构的振动位移曲线。From this solution, the object-space coordinate value of the measuring point P that changes with time is obtained, and then the vibration displacement curve of the measured flexible structure is obtained.

第一工业相机的内参数矩阵N₁与第二工业相机的内参数矩阵N₂的获得方法相同，第一工业相机的外参数矩阵M₁与第二工业相机的外参数矩阵M₂的获得方法相同，第一工业相机的内参数矩阵N₁和外参数矩阵M₁的获得方法为：The internal parameter matrix N ₁ of the first industrial camera is obtained in the same way as the internal parameter matrix N ₂ of the second industrial camera, and the external parameter matrix M ₁ of the first industrial camera is obtained by the external parameter matrix M ₂ of the second industrial camera Similarly, the method for obtaining the internal parameter matrix N ₁ and external parameter matrix M ₁ of the first industrial camera is:

第一工业相机的标定过程中，将拍摄获得的标定板的静止图像利用角点检测算法提取相邻正方形网格的角点，获得标定板中的角点像素坐标，再根据获得的角点像素坐标，求解下面方程：During the calibration process of the first industrial camera, the corner point detection algorithm is used to extract the corner points of the adjacent square grids from the still image of the calibration board obtained by shooting, and the corner point pixel coordinates in the calibration board are obtained, and then according to the obtained corner point pixel coordinates coordinates, solve the following equation:

${Z Z}_{c c} [\begin{matrix} u u \\ v v \\ 11 \end{matrix}] = = [\begin{matrix} f f / / dx dx & 00 & {u u}_{00} \\ 00 & f f / / dy dy & {v v}_{00} \\ 00 & 00 & 11 \end{matrix}] [\begin{matrix} R R & t t \end{matrix}] [\begin{matrix} {X x}_{w w} \\ {Y Y}_{w w} \\ {Z Z}_{w w} \\ 11 \end{matrix}] = = {N N}_{11} {M m}_{11} [\begin{matrix} {X x}_{w w} \\ {Y Y}_{w w} \\ {Z Z}_{w w} \\ 11 \end{matrix}],,$

式中，Z_c为第一工业相机在其光心坐标系下角点的Z轴坐标值，(u,v)为角点的像素坐标值，f为第一工业相机的焦距，(dx，dy)为第一工业相机成像元件的物理宽度，(u₀,v₀)为第一工业相机的成像中心坐标值，R为第一工业相机外参数矩阵中的旋转矩阵，t为第一工业相机外参数矩阵中的平移矩阵；In the formula, Z _c is the Z-axis coordinate value of the corner point of the first industrial camera in its optical center coordinate system, (u, v) is the pixel coordinate value of the corner point, f is the focal length of the first industrial camera, (dx, dy ) is the physical width of the imaging element of the first industrial camera, (u ₀ , v ₀ ) is the coordinate value of the imaging center of the first industrial camera, R is the rotation matrix in the external parameter matrix of the first industrial camera, and t is the first industrial camera the translation matrix in the extrinsic parameter matrix;

求解上式，获得第一工业相机的内参数矩阵N₁和外参数矩阵M₁。Solve the above formula to obtain the internal parameter matrix N ₁ and external parameter matrix M ₁ of the first industrial camera.

本发明的优点：本发明所述大型柔性结构振动位移的双目摄影测量方法，通过两个或两个以上高速工业相机，从不同的角度拍摄柔性结构的振动过程，并利用角点检测算法获得测点的像素坐标，再根据三维重构原理得到测点每一时刻的三维坐标，该三维坐标的时间序列就是位移振动信息。Advantages of the present invention: the binocular photogrammetry method for the vibration displacement of a large flexible structure described in the present invention uses two or more high-speed industrial cameras to photograph the vibration process of the flexible structure from different angles, and uses the corner point detection algorithm to obtain The pixel coordinates of the measuring point, and then the three-dimensional coordinates of the measuring point at each moment are obtained according to the principle of three-dimensional reconstruction. The time series of the three-dimensional coordinates is the displacement vibration information.

本发明解决了大型柔性结构的振动位移测量问题，避免了接触式测量对结构本身性能的影响，降低了测试设备复杂度，同时它将基于图像的三维重构技术应用于动态测量，提出了一种利用数字摄像测量技术对柔性结构进行振动测试的方法。该方法利用两台以上工业摄影相机以同一帧率拍摄被测物体，利用MCS51单片机控制时序达到严格同步。它避免了对柔性结构本身性能的影响，具有测试过程简单，测量结果精度高等特点，可以广泛应用于航天大型柔性结构的振动测量。The invention solves the problem of vibration displacement measurement of large flexible structures, avoids the impact of contact measurement on the performance of the structure itself, and reduces the complexity of testing equipment. At the same time, it applies image-based three-dimensional reconstruction technology to dynamic measurement, and proposes a A method for vibration testing of flexible structures using digital photogrammetry. The method uses more than two industrial photographic cameras to shoot the measured object at the same frame rate, and uses MCS51 single-chip microcomputer to control the time sequence to achieve strict synchronization. It avoids the impact on the performance of the flexible structure itself, has the characteristics of simple testing process and high precision of measurement results, and can be widely used in the vibration measurement of large flexible structures in aerospace.

本发明方法同时具有测量过程快速，无需额外信号放大电路；适应性强，可以根据测量要求改变相机帧率和分辨率，以获得更精确的结果的特点；它受环境噪声影响小；在标定准确的前提下，测量精度较高。The method of the invention has the characteristics of fast measurement process and no need for additional signal amplification circuit; strong adaptability, can change the frame rate and resolution of the camera according to the measurement requirements to obtain more accurate results; it is less affected by environmental noise; accurate in calibration Under the premise, the measurement accuracy is high.

附图说明Description of drawings

图1是标定板的结构示意图；Fig. 1 is the schematic structural diagram of calibration plate;

图2是双目三维重构的原理图；第一工业相机像空间坐标系O₁-x_ly_lz_l与物方坐标系O-xyz重合，图像坐标系为O_l-X_lY_l，有效焦距为f_l；第二工业相机像空间坐标系O₂-x_ry_rz_r，图像坐标系为O_r-X_rY_r，有效焦距为f_r；物方点P在O-xyz中的坐标为(X,Y,Z)，其在第一工业相机像片中对应的像点p在O₁-x_ly_lz_l中的坐标为(x,y,f_l)，P在第二工业相机像片中对应的像点p_r在O₂-x_ry_rz_r中的坐标为(x_r,y_r,f_r)；Figure 2 is the principle diagram of binocular 3D reconstruction; the image space coordinate system O ₁ -x _{ly l z l} _of _the first industrial camera coincides with the object space coordinate system O-xyz, and the image coordinate system is O _l -X _l Y _l , the effective focal length is f _l ; the image space coordinate system of the second industrial camera is O ₂ -x _{ry r z r} _, _the image coordinate system is O _r -X _r Y _r , and the effective focal length is f _r ; the object space point P is at O- The coordinates in xyz are (X, Y, Z), and the coordinates of the corresponding image point p in O ₁ -x _l y _l z _l in the photo of the first industrial camera are (x, y, f _l ), The coordinates of the image point p _r corresponding to P in the photo of the second industrial camera in O ₂ -x _r y _r z _r are (x _r , y _r , f _r );

图3是本发明所述大型柔性结构振动位移的双目摄影测量方法的测量状态示意图；图中1为被测柔性结构，2为第一工业相机，3为第二工业相机，被测柔性结构1上的圆点表示测点。Fig. 3 is the measurement status schematic diagram of the binocular photogrammetry method of vibration displacement of large flexible structure described in the present invention; Among the figure 1 is the measured flexible structure, 2 is the first industrial camera, 3 is the second industrial camera, the measured flexible structure The dots on 1 indicate the measuring points.

具体实施方式detailed description

具体实施方式一：下面结合图1至图3说明本实施方式，本实施方式所述大型柔性结构振动位移的双目摄影测量方法，它包括以下步骤：Specific embodiment one: the present embodiment is described below in conjunction with Fig. 1 to Fig. 3, the binocular photogrammetry method of the vibration displacement of the large-scale flexible structure described in the present embodiment, it comprises the following steps:

本实施方式中，被测柔性结构的位移范围在任一工业相机的视场内，对被测柔性结构的激励方式可以根据测试要求自由选择。对多个工业相机的同步触发，采用MCS51单片机控制时序，保证多个工业相机同时触发和停止；工业相机帧率根据被测柔性结构的振动频率选择，要求满足奈奎斯特采样定理，要求不存在丢帧现象。In this embodiment, the displacement range of the flexible structure under test is within the field of view of any industrial camera, and the excitation method for the flexible structure under test can be freely selected according to test requirements. For the synchronous triggering of multiple industrial cameras, the MCS51 single-chip microcomputer is used to control the timing to ensure that multiple industrial cameras are triggered and stopped at the same time; the frame rate of the industrial cameras is selected according to the vibration frequency of the flexible structure to be tested, and it is required to meet the Nyquist sampling theorem. Frame drops are present.

具体实施方式二：下面结合图1说明本实施方式，本实施方式对实施方式一作进一步说明，本实施方式标定工业相机的具体方法为：Specific implementation mode two: The following describes this implementation mode in conjunction with Figure 1. This implementation mode further explains implementation mode one. The specific method for calibrating an industrial camera in this implementation mode is as follows:

本实施方式中的标定板采用高精度打印机打印，以保证工业相机标定的精度满足测量要求。标定完成后工业相机的安装位置和角度，以及镜头焦距等都不得改变。拍摄振动过程时保证被测物上的测点靶标没有超出任一相机的视场范围。相机帧率的选择为大于所关心的振动频率的3倍，最小分辨率为不低于30w像素，相机与被测物之间的距离不超过3m。所有工业相机的拍摄必须严格同步进行，要求不存在丢帧现象。采用MCS51单片机进行拍摄时序控制，利用开关信号实现各个相机的触发。触发时序精度为1μs。The calibration plate in this embodiment is printed with a high-precision printer to ensure that the calibration accuracy of the industrial camera meets the measurement requirements. After the calibration is completed, the installation position and angle of the industrial camera and the focal length of the lens must not be changed. When shooting the vibration process, ensure that the measuring point target on the measured object does not exceed the field of view of any camera. The frame rate of the camera is selected to be greater than 3 times the vibration frequency concerned, the minimum resolution is not less than 30w pixels, and the distance between the camera and the measured object is not more than 3m. The shooting of all industrial cameras must be strictly synchronized, requiring no frame loss. MCS51 single-chip microcomputer is used to control the shooting sequence, and the trigger of each camera is realized by using the switch signal. The trigger timing accuracy is 1μs.

工业相机拍摄好的一系列图片最后导入计算机中进行处理，处理过程中利用角点检测算法得到每一测点在不同工业相机图像中的像素坐标，再根据已知的相机参数进行坐标变换，得到每一帧时刻测点的物方坐标。物方坐标的时间序列就是测点的位移振动曲线。A series of pictures taken by industrial cameras are finally imported into the computer for processing. During the processing, the pixel coordinates of each measuring point in different industrial camera images are obtained by using the corner detection algorithm, and then the coordinates are transformed according to the known camera parameters to obtain The object space coordinates of the measuring point at each frame time. The time series of object space coordinates is the displacement vibration curve of the measuring point.

具体实施方式三：下面结合图2和图3说明本实施方式，本实施方式对实施方式二作进一步说明，本实施方式所述工业相机的数量为2个，步骤三中得到被测柔性结构的振动位移曲线的具体方法为：Specific embodiment three: the present embodiment will be described below in conjunction with Fig. 2 and Fig. 3, and this embodiment will further explain embodiment two, the quantity of the industrial camera described in this embodiment is 2, obtain the measured flexible structure in step 3 The specific method of the vibration displacement curve is:

则上式变形为Kh＝U，Then the above formula is transformed into Kh=U,

h＝(K^TK)^-1K^TU，h=(K ^T K) ^-1 K ^T U,

本实施方式中，测点可采用反光片标识以便提取。像素坐标的提取方法为角点检测法，可采用预置的计算程序实现，无需每一帧像素点分别提取，效率较高。像素坐标和物方坐标的转换需要利用步骤一中的标定结果。In this embodiment, the measuring points can be marked with reflective sheets for easy extraction. The extraction method of pixel coordinates is the corner detection method, which can be realized by a preset calculation program, without the need to extract pixels of each frame separately, and the efficiency is high. The conversion of pixel coordinates and object space coordinates requires the use of the calibration results in step 1.

具体实施方式四：本实施方式对实施方式三作进一步说明，本实施方式所述第一工业相机的内参数矩阵N₁与第二工业相机的内参数矩阵N₂的获得方法相同，第一工业相机的外参数矩阵M₁与第二工业相机的外参数矩阵M₂的获得方法相同，第一工业相机的内参数矩阵N₁和外参数矩阵M₁的获得方法为：Specific Embodiment 4: This embodiment will further explain Embodiment 3. The internal parameter matrix N ₁ of the first industrial camera described in this embodiment is obtained in the same way as the internal parameter matrix N ₂ of the second industrial camera. The external parameter matrix M1 of the camera is obtained in the same way as the external parameter matrix _M2 _of the second industrial camera, and the internal parameter matrix N1 and external parameter matrix M1 _of the _first industrial camera are obtained by:

具体实施方式五：本实施方式对实施方式一、二、三或四作进一步说明，本实施方式所述单片机为MCS51单片机。Embodiment 5: This embodiment further explains Embodiment 1, 2, 3 or 4. The single-chip microcomputer described in this embodiment is an MCS51 single-chip microcomputer.

本发明方法在标定过程中，首先将工业相机安装到位，调节合适焦距，后将标定板放入相机视场进行拍照。然后在计算机中利用角点检测算法获得标定板中角点像素坐标。再求得相机参数矩阵。然后将所有工业相机连接至单片机控制电路，测试触发和图片存储。本发明方法中，在理论上采用的摄像机数目越多，得到的测量结果越精确。利用不同时刻拍摄得到的图像序列，就可以得到测点的位移振动曲线，完成对测点的三维动态重构。分别提取将每一时刻的三维坐标，便得到测点的三轴位移振动曲线。In the calibration process of the method of the present invention, the industrial camera is firstly installed in place, a suitable focal length is adjusted, and then the calibration plate is put into the field of view of the camera to take pictures. Then use the corner detection algorithm in the computer to obtain the pixel coordinates of the corner points in the calibration board. Then obtain the camera parameter matrix. Then connect all industrial cameras to the single chip microcomputer control circuit, test trigger and picture storage. In the method of the present invention, theoretically, the more cameras are used, the more accurate the measurement results will be. Using the image sequences captured at different times, the displacement vibration curve of the measuring point can be obtained, and the three-dimensional dynamic reconstruction of the measuring point can be completed. The three-dimensional coordinates of each moment are respectively extracted to obtain the three-axis displacement vibration curve of the measuring point.

本发明可以有效地利用工业相机进行柔性结构振动测试，经过试验，在光照充足，标定准确的情况下，采用30w像素相机，拍摄距离为2m的柔性结构，测量精度可达0.1mm。可以满足大部分情况下的柔性结构振动非接触式测量要求。The invention can effectively use the industrial camera to carry out the vibration test of the flexible structure. After the test, under the condition of sufficient light and accurate calibration, a 30W pixel camera is used to shoot the flexible structure with a shooting distance of 2m, and the measurement accuracy can reach 0.1mm. It can meet the non-contact measurement requirements of flexible structure vibration in most cases.

Claims

1. a binocular photogrammetric survey method for large-size pliable structure vibration displacement, it comprises the following steps:

Step one: demarcate industrial camera, the quantity of described industrial camera is 2-5；

Step 2: the tested flexible structure of exciting, synchronizes to trigger by all industrial cameras of Single-chip Controlling simultaneously, carries out tested The shooting of flexible structure measured point vibration displacement；

Step 3: the image obtaining the shooting of all industrial cameras processes, utilizes Corner Detection Algorithm to obtain the picture of measuring point Element coordinate, recycling binocular three-dimensional reconfiguration technique obtains the object coordinates of each moment measuring point, thus obtains tested flexible structure Vibration displacement curve；

Demarcate industrial camera method particularly includes:

Using scaling board to demarcate industrial camera, described scaling board is the chess of be made up of chequered with black and white square 7 × 9 Dish lattice, the square length of side is 5mm；The position of scaling board is set, makes the field range of each industrial camera cover complete Scaling board, and use the rest image of industrial camera shooting scaling board, complete to demarcate；

It is characterized in that, the quantity of industrial camera is 2, obtains the vibration displacement curve of tested flexible structure in step 3 Method particularly includes:

Shot the measured point of tested flexible structure by 2 industrial cameras simultaneously, after the image obtaining shooting processes, build Vertical equation group is as follows:

\{\begin{matrix} Z_{c 1} [\begin{matrix} u_{1} \\ v_{1} \\ 1 \end{matrix}] = N_{1} M_{1} [\begin{matrix} X_{w} \\ Y_{w} \\ Z_{w} \\ 1 \end{matrix}] = Q_{1} [\begin{matrix} X_{w} \\ Y_{w} \\ Z_{w} \\ 1 \end{matrix}] \\ Z_{c 2} [\begin{matrix} u_{2} \\ v_{2} \\ 1 \end{matrix}] = N_{2} M_{2} [\begin{matrix} X_{w} \\ Y_{w} \\ Z_{w} \\ 1 \end{matrix}] = Q_{2} [\begin{matrix} X_{w} \\ Y_{w} \\ Z_{w} \\ 1 \end{matrix}] \end{matrix},

In formula, Z_c1Be the first industrial camera under its photocentre coordinate system, the Z axis coordinate figure of measuring point P, (u₁,v₁) it is measuring point P Pixel coordinate value, (X_w,Y_w,Z_w) it is the object coordinates value of measuring point P, N₁It is the Intrinsic Matrix of the first industrial camera, M₁ It is the outer parameter matrix of the first industrial camera, Q₁It is the projection matrix of the first industrial camera, Q₁=N₁M₁；

Z_c2Be the second industrial camera under its photocentre coordinate system, the Z axis coordinate figure of measuring point P, (u₂,v₂) it is the picture of measuring point P Element coordinate figure, (X_w,Y_w,Z_w) it is the object coordinates value of measuring point P, N₂It is the Intrinsic Matrix of the second industrial camera, M₂For The outer parameter matrix of the second industrial camera, Q₂It is the projection matrix of the second industrial camera, Q₂=N₂M₂；

Above-mentioned photocentre coordinate system represents that optical axis is the coordinate system of Z axis with the photocentre of industrial camera as initial point；

Above-mentioned equation group is processed, it is thus achieved that overdetermined equation:

\{\begin{matrix} {Q^{1}}_{11} X_{w} + {Q^{1}}_{12} Y_{w} + {Q^{1}}_{13} Z_{w} + {Q^{1}}_{14} - u_{1} {Q^{1}}_{31} X_{w} - u_{1} {Q^{1}}_{32} Y_{w} - u_{1} {Q^{1}}_{33} Z_{w} = u_{1} {Q^{1}}_{34} \\ {Q^{1}}_{21} X_{w} + {Q^{1}}_{22} Y_{w} + {Q^{1}}_{23} Z_{w} + {Q^{1}}_{24} - v_{1} {Q^{1}}_{31} X_{w} - v_{1} {Q^{1}}_{32} Y_{w} - v_{1} {Q^{1}}_{33} Z_{w} = v_{1} {Q^{1}}_{34} \\ {Q^{2}}_{11} X_{w} + {Q^{2}}_{12} Y_{w} + {Q^{2}}_{13} Z_{w} + {Q^{2}}_{14} - u_{2} {Q^{2}}_{31} X_{w} - u_{2} {Q^{2}}_{32} Y_{w} - u_{2} {Q^{2}}_{33} Z_{w} = u_{2} {Q^{2}}_{34} \\ {Q^{2}}_{21} X_{w} + {Q^{2}}_{22} Y_{w} + {Q^{2}}_{23} Z_{w} + {Q^{2}}_{24} - v_{2} {Q^{2}}_{32} X_{w} - v_{2} {Q^{2}}_{32} Y_{w} - u_{2} {Q^{2}}_{33} Z_{w} = v_{2} {Q^{2}}_{34} \end{matrix},

Q in formula¹ _ijRepresent Q₁I-th row jth row, i=1,2,3, j=1,2,3,4；

Q² _ijRepresent Q₂I-th row jth row, i=1,2,3, j=1,2,3,4；

Arrange overdetermined equation, obtain:

[\begin{matrix} {Q^{1}}_{11} - u_{1} {Q^{1}}_{31} & {Q^{1}}_{12} - u_{1} {Q^{1}}_{32} & {Q^{1}}_{13} - u_{1} {Q^{1}}_{33} \\ {Q^{1}}_{21} - v_{1} {Q^{1}}_{31} & {Q^{1}}_{22} - v_{1} {Q^{1}}_{32} & {Q^{1}}_{23} - v_{1} {Q^{1}}_{33} \\ {Q^{2}}_{11} - u_{2} {Q^{2}}_{31} & {Q^{2}}_{12} - u_{2} {Q^{2}}_{32} & {Q^{2}}_{13} - u_{2} {Q^{2}}_{33} \\ {Q^{2}}_{21} - v_{2} {Q^{2}}_{31} & {Q^{2}}_{22} - v_{2} {Q^{2}}_{32} & {Q^{2}}_{23} - v_{2} {Q^{2}}_{33} \end{matrix}] [\begin{matrix} X_{w} \\ Y_{w} \\ Z_{w} \end{matrix}] = [\begin{matrix} u_{1} {Q^{1}}_{34} - {Q^{1}}_{14} \\ v_{1} {Q^{1}}_{34} - {Q^{1}}_{24} \\ u_{2} {Q^{2}}_{34} - {Q^{2}}_{14} \\ v_{2} {Q^{2}}_{34} - {Q^{2}}_{24} \end{matrix}],

Set

K = [\begin{matrix} {Q^{1}}_{11} - u_{1} {Q^{1}}_{31} & {Q^{1}}_{12} - u_{1} {Q^{1}}_{32} & {Q^{1}}_{13} - u_{1} {Q^{1}}_{33} \\ {Q^{1}}_{21} - v_{1} {Q^{1}}_{31} & {Q^{1}}_{22} - v_{1} {Q^{1}}_{32} & {Q^{1}}_{23} - v_{1} {Q^{1}}_{33} \\ {Q^{2}}_{11} - u_{2} {Q^{2}}_{31} & {Q^{2}}_{12} - u_{2} {Q^{2}}_{32} & {Q^{2}}_{13} - u_{2} {Q^{2}}_{33} \\ {Q^{2}}_{21} - v_{2} {Q^{2}}_{31} & {Q^{2}}_{22} - v_{2} {Q^{2}}_{32} & {Q^{2}}_{23} - v_{2} {Q^{2}}_{33} \end{matrix}],

h = [\begin{matrix} X_{w} \\ Y_{w} \\ Z_{w} \end{matrix}],

U = [\begin{matrix} u_{1} {Q^{1}}_{34} - {Q^{1}}_{14} \\ v_{1} {Q^{1}}_{34} - {Q^{1}}_{24} \\ u_{2} {Q^{2}}_{34} - {Q^{2}}_{14} \\ v_{2} {Q^{2}}_{34} - {Q^{2}}_{24} \end{matrix}],

Then above formula is deformed into Kh=U,

Solve obtain h least square solution:

H=(K^TK)^-1K^TU,

Thus solve the object coordinates value obtaining time dependent measuring point P, and then obtain the vibration displacement of tested flexible structure Curve.

The binocular photogrammetric survey method of large-size pliable structure vibration displacement the most according to claim 1, it is characterised in that The Intrinsic Matrix N of the first industrial camera₁Intrinsic Matrix N with the second industrial camera₂Preparation method identical, the first work The outer parameter matrix M of industry camera₁Outer parameter matrix M with the second industrial camera₂Preparation method identical, the first industrial camera Intrinsic Matrix N₁With outer parameter matrix M₁Preparation method be:

In the calibration process of the first industrial camera, the rest image of scaling board shooting obtained utilizes Corner Detection Algorithm to extract The angle point of adjacent square grid, it is thus achieved that the corner pixels coordinate in scaling board, further according to the corner pixels coordinate obtained, asks Solve following equation:

Z_{c} [\begin{matrix} u \\ v \\ 1 \end{matrix}] = [\begin{matrix} f / d x & 0 & u_{0} \\ 0 & f / d y & v_{0} \\ 0 & 0 & 1 \end{matrix}] [\begin{matrix} R & t \end{matrix}] [\begin{matrix} X_{w} \\ Y_{w} \\ Z_{w} \\ 1 \end{matrix}] = N_{1} M_{1} [\begin{matrix} X_{w} \\ Y_{w} \\ Z_{w} \\ 1 \end{matrix}],

In formula, Z_cBe the first industrial camera Z axis coordinate figure of angle point under its photocentre coordinate system, (u, v) be angle point pixel sit Scale value, f is the focal length of the first industrial camera, and (dx, dy) is the physical width of the first industrial camera image-forming component, (u₀,v₀) Being the imaging center coordinate figure of the first industrial camera, R is the spin matrix in the outer parameter matrix of the first industrial camera, and t is the Translation matrix in the outer parameter matrix of one industrial camera；

Solve above formula, it is thus achieved that the Intrinsic Matrix N of the first industrial camera₁With outer parameter matrix M₁。

The binocular photogrammetric survey method of large-size pliable structure vibration displacement the most according to claim 1 and 2, its feature exists In, described single-chip microcomputer is MCS51 single-chip microcomputer.