Application of Robust Regression Estimation in Moving Background Compensation
Cai Qing-Yao, Ding Xi-Bo, Dong Shan-Shan · 2016
The primary task in preprocessing image is moving estimation and compensation of image senor, that is to say, correction problem of image background. In this paper, a new method of motion background compensation based on robust regression is proposed. The background motion velocity is calculated by the estimation of the optical flow field model. Then robust iterative weighted least square method is used to estimate the global motion parameters of the image sensor. This technique solves the problem of image background correction, improve the performance of detection and tracking algorithm and the motion background is compensated by the estimated global motion parameters. Optical flow field technology is mainly used in tracking algorithm. The experimental results show that the proposed method can meet the real-time requirement of motion background compensation.