Moving Target Detection In The Image Of A Sequence Of Activities

Motion targets detection and recognition is the key project of Image Process and Image understanding, it has been widely used in many areas such as automation navigation, Video Surveillance and monitoring, Medicine image analysis and Video image compressing and transmitting. With its application enlarged, its utility gets more and more important.Sequences can be divided into two types by which background is still or not. There are many reports in the first type and these reports mostly base on frame subtraction and background modeling. The other type is rarely researched and far away from matured. We discuss them both in this paper.This is the contexts:To the static background sequences, I introduced the method of how to extract the main edges of the motion targets from differences of frames, the method of motion targets detection by regional growing based on the differences of frames and the improvement of the method of background modeling.To the moving background sequences, I used the method of block match to estimate the motion vectors of pixels and blocks, and then use the vectors to identify the targets, 'cause the targets and the background had the different vectors.The last, I introduced how to design and realize a software of motion targets detection subsystem in a project. With the multithread programming, the subsystem take few resources and can identify and trace the targets exactly. Further more, it provided a better solution of how to use all kinds of algorithms in realities.