An efficient approach for motion detecting and tracking

In recent years, the increasing demand for security in home care services has led to further improvements in surveillance activities. This paper proposes an efficient approach for motion detecting and tracking in a surveillance system based on the cellular model. The system can monitor human activities and recognize the abnormal events of an elderly man in an indoor environment. The experimental room should be divided into several unit areas. Each unit area can be considered as a simple cell in the 2D cellular model. The state of each cell can be represented by the changeable number of active pixels within corresponding unit area. The approach groups those cells with nonzero status value to represent the target. Then, the centroid of a moving target is computed by its zeroth and first moments. The variation of the centroid can be tracked to recognize abnormal events, such as the fall of an old man or children. The results of the experiments conducted in this study clearly demonstrate that the proposed approach provides a more efficient and effective detecting mechanism.