In recent years, video surveillance systems have been playing a significantly important role in the human safety and security field by monitoring public or private areas. In this chapter, we have discussed the development of an intelligent surveillance system to detect, track and identify potentially hazardous events that may occur at level crossings (LC). This system starts by detecting and tracking objects on the level crossing. Then, a danger evaluation method is built using hidden Markov model in order to predict trajectories of the detected objects. The trajectories are analyzed with a credibility model to evaluate dangerous situations at level crossings. Synthetics and real data are used to test the effectiveness and the robustness of the proposed algorithms and the whole approach by considering various scenarios within several situations.
Part of the book: Motion Tracking and Gesture Recognition