A Foreground-background Segmentation Algorithm for Video Sequences
Moving objects extraction is a crucial part of video surveillance system.This paper presents a foregroundbackground algorithm for motion detection.It is based on traditional adaptive mixture Gaussian model.By dynamically adjusting the parameters and the number of Gaussian components,the computation cost reduced greatly.In order to solve detected moving target based on Gaussian mixture model easily broken,two-way matching method on the basis of frame difference thoughts with a series of image filtering methods are combined.In a stable outdoor detector,the algorithm deals with lighting changes,swaying of leaves,and various noises reliably.The proposed algorithm can identify moving objects more exactly than traditional method.An intrusion detection alarming system can be built to discover the abnormity by using the algorithm to process and analyze the video sequences.
moving object detection background extraction goreground segmentation Gaussian mixture model intrusion detection alarming
Zhou Wei Peng Li HuangYue
Dept.of Internet of Things Engineering Jiangnan University Wuxi,China school of IoT technology Wuxi institute of technology Wuxi,China
国际会议
贵阳
英文
340-343
2015-08-18（万方平台首次上网日期，不代表论文的发表时间）