Novel Method of Moving Object Detection Based on Three Frame Differencing and Connectivity Checking

Detection of moving objects in an image sequence is vital for applications of image processing,e.g.,intelligent video surveillance systems and digital TVs.In this paper,a novel algorithm for automated detection of moving objects was proposed and discussed.It is characterized by a fusion of some improving strategies including exploitation of edge gradients,three-frame-differencing,and connectivity checking.The results of theoretical analyses and computer simulation presented show that the algorithm has some advantages over its competitors,e.g.,less computation cost and higher processing speed(e.g.,15 ms per frame for a image of 320×240×24 bits),being more robust with noisy and luminance-varying image sequence.