Moving Object Detection Method Based on NMI Features Motion Detection Frame Difference

摘要

This paper focuses on motion target front and back disruption or smearing phenomenon this easily produced by inter-frame difference algorithm. In this paper, which use NMI features to extract Image feature and use Frame Difference to detection Moving object. In the method, firstly extracting NMI features of motion the target area, and the standardized vector. Then analyzing data of definition pixel region that produced by frame data difference to capture moving objects in the scene. From the simulation validation, the method can effectively solve the phenomenon of splitting or smear that appear in the frame difference and accurately extract target.