Stereo camera calibration using real coded genetic algorithm

Abstract

In this paper, the problem of obtaining calibration parameters for a stereo camera system is considered. A real coded genetic algorithm is used to solve this problem. The solution of extrinsic and intrinsic parameters of the left and right cameras are encoded in the vectors of real numbers. The Laplace crossover and power mutation operators are used to obtain next generation of the solutions. The solutions are obtained using three different numbers of control points. The robustness of proposed approach is checked in the presence of image pixel noise. Encouraging results are obtained in terms of robustness and accuracy.