Moving Object Detection Using Three-Frame Difference and Background Subtraction

Abstract

Moving object detection is an important research content in video processing. Frames subtraction and background subtraction are commonly used methods to detect moving objects. They are both simple and effective, but have some limitations. Background updating can be adapted to changed environment. Three-frame difference compared to traditional frames subtraction can get more complete object and eliminate noise. After in-depth study of three-frame difference and Surendra background updating based background subtraction, this paper proposes an approach of dynamic threshold based three-frame difference and Surendra background updating based background subtraction. The experiment results show that the approach can detect object quickly and accurately, and adapt to the complex changed environment.