Research On Three-Dimensional Measurement Algorithm Based On Line Structured Light

With the rise of intelligent manufacturing,the role of three-dimensional measurement technology in industrial production is gradually becoming more prominent.The three-dimensional measurement based on line structured light is favored by more and more researchers with the characteristics of non-contact,high accuracy and real-time.Three-dimensional measurement technology based on line structured light is playing an important role in the fields of industrial processing,on-line inspection,three-dimensional modeling and other applications.In the text,a low-cost three-dimensional measurement system is studied,and the key algorithms in three-dimensional measurement of line structured light are deeply studied.The whole three-dimensional measurement algorithm is implemented using C++ combined with MFC and OpenCV library.The specific research is as follows:(1)The method of camera calibration based on planar target was studied.The radial distortion and tangential distortion of the lens were taken into account.The nonlinear optimization method was used to solve the parameters of the camera,and the calibration experiments were carried out using the checkerboard target and the circular target respectively.Experiments and corresponding calibration accuracy assessment.(2)The method of optical plane calibration based on planar targets was studied.The feature points on the light plane were obtained by using the checkerboard and the principle of cross ratio invariance on the structured light plane.The parameters of the structured light plane equation in the camera coordinate were obtained by least squares fitting.The accuracy of the calibration of the structured light plane was evaluated using the Euclidean distance between the coordinates of the feature point under the measurement model and the calculated the coordinates from the principle of constant cross ratio.(3)For the acquired distortion light stripe image,the image noise is removed by filtering and morphological transformation,and then the distorted light stripe is separated from the image by the threshold segmentation,finally the center of the light stripe is extracted.Discussed various light stripe center extraction methods.For the Hessian matrix method with large computational complexity,an innovative method is proposed that combines the geometric center and the Hessian matrix to extract the center of the light stripe,and combines the separation and symmetry of the two-dimensional Gaussian function.Without loss of accuracy the computational complexity of the algorithm is greatly reduced,and the productivity of the algorithm is significantly improved.(4)The process of the whole three-dimensional measurement algorithm was elaborated.The measurement range and the factors affecting the measurement accuracy of the three-dimensional measurement were analyzed in detail.The width and height measurement experiments were carried out by the self-built platform to verify the validity and correctness of the algorithm.As the system is mainly used in industrial processing to adjust the process parameters according to height,the accuracy requirements are not particularly high.The measurement accuracy of the low-cost measurement system constructed in this paper can meet actual needs and has practical application value.