基于单相机的飞行器结构表面应变片三维定位方法

3D Coordinates Measurement of Strain Gauges on Structures Based on the Single Camera

本文介绍了一种基于单相机的应变片三维定位方法,可获取飞行器结构件表面应变片的空间坐标,有助于解决飞行器结构数值仿真计算与试验数据比对的技术难题。为便于使用,本文采用单个单反相机从多个角度对结构表面进行多视觉重叠成像。基于透视多点算法,根据光学编码点在相邻帧的图像坐标,解算当前相机相对于参考坐标系的旋转和平移信息。基于深度学习算法对相机所拍摄图像中的应变片进行自动识别,根据已确定的相机姿态构建相邻帧之间的极线约束,将识别到的应变片进行立体匹配,获得相邻帧之间的立体对应关系。最后利用三角测量原理对应变片三维坐标进行重构。实验验证本文方法的定位精度可以达到0.100mm/m。

In order to solve the problem in comparing the numerical solutions and the experimental results,a method for determining the 3D coordinates of strain gauges glued on the surface of aerospace structures has been developed by means of a single camera.For ease of use,this paper uses a single SLR camera to image the structural surface from multiple angles.Based on the perspective-n-points algorithm,the rotation and translation information of the current camera relative to the reference coordinate system is solved according to the image coordinates of the optical coding points in the adjacent frame.After the strain gauges in the image taken by the camera is automatically recognized based on the deep learning algorithm,the epipolar constraint between adjacent frames is constructed according to the determined camera attitude,and the recognized strain gauge is spatially matched to obtain the stereo correspondence between adjacent frames.Finally,the three-dimensional coordinates of the corresponding variation are reconstructed based on the principle of triangulation.The experimental results show that the positioning accuracy of the introduced method could fetch up to 0.100mm/m.