Moving Target Detection Based on Background Modeling by Multi-level Median Filter

Video surveillance system gets more and more attention and application in security and defense fields. The background modeling and moving target detection are the key technologies of the whole surveillance system. In this paper, a new approach used multi-level median filter for background modeling is proposed. It can model background exactly and effectively and be maintained in real-time, operated easily and performed quickly. After threshold operation, morphology operation was used for removing noise and smoothing the unconnected area. Experiment results show that the proposed method is fast and effective