Model-Free-Adaptive-Control for Moving Object Detection in RGB Video Sequence
Guoqing Sun, Zhongsheng Hou · 2021 IEEE 10th Data Driven Control and Learning Systems Conference (DDCLS) · 2021
Moving object detection in video sequences is an important topic in the field of computer vision. The video background image extraction is the key step in most moving object detection methods. The traditional background image extraction methods are easy to cause model distortion when the video scene cannot meet its assumptions or model conditions. In this paper, the background image extraction method is proposed using model free adaptive control. The method combines RGB three-channel history and current data of pixels to represent and update background image. The proposed method that under different video sequences is compared with the traditional background image extraction methods. The results show that the method can extract the color background image and separated the color foreground image, and the foreground image is closer to the ground truth image of the video target.