Study On The Algorithm Of Fast Calibration For Multi-camera System

The multi-camera system with high precision,large field of view and the advantage of full-field measurement is widely used in a variety of measurement processes.For multi-camera system,camera calibration is the first step of reconstruction measurement.The robustness of calibration algorithm and the accuracy of the results directly affect the accuracy of the reconstruction results.Since the shooting angle and the range of each camera are different,it is difficult for each camera to find a public field of view in a multi-camera system.In the calibration process,to capture the target simultaneously and accomplish calibration process one-time are not easy.Therefore,the new requirements on calibration algorithm are put forward.In this paper,the imaging geometry of multi-camera system and some existed calibration algorithms are studied,and then the improved methods are proposed for the limitations and shortcomings in multi-camera system calibration.The main research works of this paper are as follows:(1)The stereo calibration part based on Zhang's plane template calibration method is improved,and the process of solving the external parameters at once is accomplished.This method uses Zhang's algorithm to calculate the internal parameters and the pose of each camera.Then,the improved algorithm programmed and implement in the MATLAB platform according to the post of each camera relative to the target and the relationship of the point in each coordinate system.The process and steps of solving the external parameters are greatly simplified.(2)The stability of Orthogonal calibration algorithm based on spheres is studied in detail,and the error ratio of repeated experiments was calculated.Firstly,the imaging process of the ball is studied in detail,then the connection between the image of the ball and the image of the absolute conic(IAC)and the dual image of the absolute conic(DIAC)are studied,and linear algorithm of calculating IAC and DIAC is deduced with the orthogonal constraint.The internal parameters of camera can be obtained with IAC.The stability of the method is analyzed by the experiments respectively.The factors of influencing the method are given.(3)The algorithm of DIAC is optimized based on the Orthogonal calibration algorithm and non-linear optimization.Then the method of calculating distortion parameters is given according to the center of the sphere and its projection position.Based on the characteristics of sphere calibration and orthogonal constraints,DIAC can be recognized with the parameters between image of sphere and its center,and DIAC.The non-linear optimization is proposed in order to satisfy the condition of positive definition of DIAC.The camera distortion is calculated based on the obtained spherical coordinates and the spherical projection coordinates.The experiment verifies the correctness of the method and gives experimental results statistics.