Fast moving object detecting method based on video streaming continuity
基于视频流连续性的快速运动目标检测方法*

This paper proposed a fast moving object detecting algorithm based on the quality of video streaming continuity. In the aspect of background reconstruction, introduced a fast background reconstruction method. It was based on pixel intensity classification method and the assumption of background pixels could be observed in at least 50% time in the training phases. In the aspect of shadow removing, presented RGB chromaticity difference algorithm based on RGB chromaticity ratio. It could remove the shadow of moving object quickly and effectively. The experimental results confirm the validity of the proposed method.