A framework of multi-objective particle swarm optimization in motion segmentation problem

Research in motion segmentation and robust tracking have been getting more attention recently. In video sequence, motion segmentation is considered as multi-objective problem. Better representation and processing of the standard image in video sequence, with efficient segmentation algorithm is required. Thus, multi-objective optimization approach is an appropriate method to solve the optimization problem in motion segmentation. In this paper, we present new framework of the video surveillance for optimization of motion segmentation using Multi-objective particle swarm (MOPSO) algorithm. Experiment based on benchmarked test functions of MOPSO and PSO is evaluated to show the result with respect to the coverage metric of the best point of optimization value. The result indicates that MOPSO is highly good in converging towards the Pareto Front and has generated a well-distributed set of non-dominated solution. Hence, is a promising solution in multi-objective motion segmentation problem of video surveillance application.