A Self-adaptive Bat Algorithm for Camera Calibration

Abstract

To obtain more accurate results of optimization calibration, an improved self-adaptive bat algorithm of camera calibration was proposed. Firstly, step parameters were set adaptively so that the objective function could avoid local minima. Secondly, the improved bat algorithms used in non-linear camera calibration did not need initial estimation values. So the proposed method could solve the problem that traditional optimization algorithms were sensitive to initial value. Furthermore, the self-adaptive bat algorithm combined with the process of camera calibration was used to optimize the camera's intrinsic parameters and the coefficient of radial distortion. Finally, the average re-projection error was analyzed, and the mean absolute error and the standard deviation were also calculated on the cases of different noise level. The experimental evaluation demonstrates that the proposed method was more efficient and accurate.