Camera Calibration Based on Improved Genetic Algorithm

Camera calibration is considered as an important issue in computer vision, and this technology has been widely applied to 3D structure reconstruction, navigation, visual surveillance, 3D measurement and other fields. Camera calibration of stereo vision is the optimization of multiparameter, nonlinear complicated function which is different to resolve by using conventional optimization method. In this paper, the encoding method of standard genetic algorithm is improved- adaptive adjustment function of variable interval is applied, and P_m, P_c adaptive adjustment is applied. Improved genetic algorithm can not only provide the adequate search space, but also ensure the precise enough encoding precision without altering the length of chromosome string. The experiments on the camera calibration of the binocular vision system indicate that the improved genetic algorithm can find the near optimum in the high-dimension optimization space effectively.