Moving Target Detection And Tracking In Infrared Image Sequences

With the rapid development of infrared technology, infrared video is widely used in military and civil affairs, such as reconnaissance, surveillance and weapon guidance. Moving objects detection and tracking is one of the most important issues in image processing and computer vision. Object detection involves verifying the presence of an object in image sequences and then segmenting it precisely. Object tracking is to locate the object in image sequences. These two processes are closely related.This paper mainly discusses the approach of moving objects detection and tracking in infrared image sequences with dynamic complex background. Due to dynamic complex background and defects of infrared image, which are low contrast between object and background, blurred edge of object and high amount of noise, objects detection and tracking is very difficult in this case. We emphasize our research on global motion estimation and object detection, and present a practical algorithm of moving objects detection and tracking under dynamic complex background.The major work and achievement of this paper are presented as follows: (1) A fast half-pixel global motion estimation and compensation algorithm is proposed. As this algorithm does not require the computation of high resolution image, it is much faster than the traditional algorithm. (2) Based on the review and analysis of typical object detection methods, especially the temporal difference, a moving objects detection algorithm based on three frame difference is proposed. This algorithm employs many new technologies, such as adaptive frame interval, half-pixel global motion estimation and compensation, adaptive change detection and object repair, therefore it is very practical. (3) A preliminary investigation is made into moving objects tracking, and a multiple moving objects tracking algorithm is presented, which can recognize combination object and single object.Experimental results show that the proposed moving objects detection and tracking algorithm works well in not only infrared image sequences but also visual ones.