Video motion detection algorithm using probabilistic time integrated RANSAC

Video motion detection (VMD) is a critical task demanded for surveillance and monitoring applications for both static and moving camera setup. A new VMD algorithm is proposed: the probabilistic time integrated RANSAC (PTIR) which robustly detects and tracks objects in the difficult long range scenario. The problem of detecting long range moving objects in a wide field of view from a moving camera is difficult especially in poor visibility, cluttered and noisy environments. The suggested method also copes with cases where no time correspondence information is available. The experimental results demonstrate the performance and robustness of the proposed method.