A new method in simultaneous estimation of Kinect-V2 sensor calibration using shuffled frog leaping algorithm

Abstract

Calibration of color and infra-red cameras is the fundamental step in almost every 2D or 3D image processing applications and pre-processing of images and videos. This article presents a novel method for estimation of 19 parameters of RGB and depth camera calibration in Kinect sensor, simultaneously. The proposed algorithm is based on applying shuffled frog leaping algorithm (SFLA) for deep optimization and estimation all parameters of intrinsic, extrinsic and lens distortions of cameras. This algorithm does not need the initial estimation for optimization and it can avoid trapping into local minima. Using non-direct estimation, we achieve middle computing matrices such as homography matrix, used in pinhole camera model. Re-projection error criteria is defined as the objective function in this algorithm. The radial lens distortion is estimated using SFLA. Kinect version 2 sensor is used in this research and experimental results show the proposed method is more efficient and accurate in compare with traditional numerical solutions.