Background Extraction and Update Algorithm Based on Bi-directional Frame Difference

摘要

Background subtraction is the most common method to delect moving targets, whose key step is to extract high-precision background. In this paper, it uses the principle of frame difference method, combined light pre-filter, multi-frame detection, regional expansion, morphology, bi-directional frame difference, and classification and discrimination, and proposes a background extraction and update algorithm based on bi-directional frame difference. The results show that the algorithm has the advantage of fast calculating speed, high accuracy, antinoise and could get ideal background in the still camera environment.