Year: 2017
Type: article
Abstract: In this paper, we propose a moving object detection method based on codebook and threeframe difference.The method uses codebook to learn the background and the three-frame difference is employed to extract initial foreground object, then we adopt the processes based on Log edge detection and component filling.It associates the foreground object obtained by codebook with the object, which is obtained by improved three-frame difference to perform logic "AND" operation, and gets the final moving object.Experimental results show that our method can effectively eliminate ghosting, double shadow and cavities of the video sequence, and the proposed method has strong robustness under the variety changing background, as well as obtains the entire moving object.
Authors: Shuai Li, Peixun Liu, Guangliang Han
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W2608788843