A GENETIC ALGORITHM APPROACH TO ACCURATE CALIBRATION OF CAMERA
一种精确档定摄像机的遗传算法方案

A new two-stage calibration method based on the close form solution and genetic algorithm was presented. In the first stage, the distortion parameters were ignored. The internal and external parameters were estimated by least square algorithm. In the second stage, the parameters estimated in the first stage were used as initial values. Taking account of the camera distortions, all parameters were optimized by genetic algorithm to get exact solutions. Because the algorithm optimized the rotation angles directly, the orthonormal constraints were easily satisfied in the present method. Moreover, the approach simplified the calibration process.