A Dynamic Adaptive Update System of Tracking Moving Objects Based on Contour Feature

This paper proposes a real-time moving objects detecting algorithm based on computer vision. Contour feature is firstly extracted from the initial frame as the feature to describe object, and then the moving object in the latter frame could be tracked and located by marking the location of this feature, at the same time, the feature is dynamically adaptively updated and optimized online. The slight computation burden and robustness to the noise and illumination disturbance make the algorithm have a good real-time quality and reliability. Experimental results indicate that the system reaches the designing requirements.