Multi-Target Tracking Technique In Particle Filtering Based Background Modeling And Motion Analysis

Multi-target tracking (MTT) is one of the most active research subjects in the area of computer vision research. With the rapid development of modern computer science and information technology, and with the marvelous renovation of image identification, MTT comes out into the open and is of great practical value in the field of military defense, medical research, traffic monitoring, astronomical prediction, intelligent supervision etc.This paper studies the "lost-object" phenomenon in the period of tracking. Based particle filter, we use motion vector in mpeg2 and GMM to model the background in order to locate the position of the moving objects. Thus, the particles with large weight become more. So, we can get tracking correctly. But we also lose the moving object when different objects is sheltered each other. We use the technology above, based objects' velocity to tell the motion direction and combined the older knowledge. At last, we confirm the direction of moving objects. We realize the MMT system and solve the "lost-object" problem during the moving basically.The research results are as follows: 1. The author models each pixel of the image with Gaussian Mixture Model (GMM), to calculate the probability of background pixel in the current image so as to abstract foreground moving objects. 2. We got the motion vector from Mpeg2, and used the way of Area to confirm the weight of particles. 3. In the waving particles, using the way of Motion Analysis to tell the correct of particles. And can tell the motion direction. 4. In the process of evaluating particle weight, we use Bhattacharyya distance to evaluate the likelihood. The experiment shows that the tracker of particle filter based motion vector, motion analysis and background modeling can solve the "lose-object" problem and realize the multi-objects effectively.