Research On Calibration Technology Of 3D Measurement System Based On Line Laser

With the continuous development and improvement of the industry,the traditional measurement technology represented by contact is difficult to meet the measurement needs of contemporary complex industrial components.Because of its non-contact characteristics,the line laser 3D measurement technology is not only flexible and high-precision,but also can better protect the measured object,so it is widely favored.The system calibration part directly affects the measurement accuracy and reconstruction effect,which is of great significance to the entire 3D measurement system.According to the characteristics of line laser three-dimensional measurement,the system calibration is divided into three parts-camera calibration,optical plane calibration and mobile platform calibration,and the following researches are carried out separately:(1)According to the principle of triangulation and the mutual conversion relationship between the four coordinate systems involved in the 3D reconstruction process,a mathematical model for the conversion of 2D pixel coordinates to 3D world coordinates is established.(2)Camera calibration: Analyze the camera calibration theory based on the plane checkerboard calibration board,and realize the camera calibration based on Zhang Zhengyou's theory.Aiming at the strict requirements for image brightness in the corner detection function that comes with opencv,this article adopts a brightness adaptive method to reduce the impact of brightness.Finally,the distortion parameters obtained by the camera calibration are used to correct the distortion of the collected images through the distortion model.(3)Optical plane calibration: Based on the principle of cross-ratio in projective geometry,using the existing checkerboard calibration board combined with mathematical principles such as line fitting,matrix singular value decomposition,plane fitting,etc.,to achieve a plane-based checkerboard calibration board The light plane calibration method.Aiming at the data with large errors in the three-dimensional feature points obtained by the cross-ratio principle,this paper proposes a plane fitting method based on the RANSAC algorithm to reduce the impact of these data.(4)Mobile platform calibration: Based on the existing checkerboard calibration board and mobile platform,the three-dimensional coordinates obtained from each frame of image are unified into a parameter coordinate system.Aiming at the fact that the moving direction of the conveyor belt of the mobile platform cannot be parallel to the X or Y coordinate direction of the world coordinate system,a parameter calibration method of the mobile platform based on the motion vector is proposed to complete the three-dimensional reconstruction more accurately.Finally,a line laser-based 3D measurement and reconstruction platform was built using fixed brackets,camera lenses,self-developed 3D cameras,lasers and other equipment.The parameters obtained by camera calibration and light plane calibration were used to convert the two-dimensional image pixel coordinates into three-dimensional world coordinates.The parameters obtained through the calibration of the mobile platform unify all threedimensional coordinate points into a coordinate system,complete the three-dimensional measurement and reconstruction of the surface of the measured object.