Abstract. Computer vision algorithms have played a vital role in video surveillance systems to detect surveillance events for public safety and security. Even so, a common demerit among these systems is their unfitness to handle divers crowded scenes. In this paper, we have developed algorithms which accommodate some of the challenges encountered in videos of crowded environments to a certain degree. Unlike many approaches that use optical flow, that estimates motion vectors only from two successive frames, we made our descriptor over long-range motion trajectories which is named short trajectorys in the paper. This paper presents a novel video descriptor, referred to as Histogram of Short trajectorys, for detecting abnormal conditions in crowded scenes. Specifically, we extract 2-d histogram from magnitude and orientation matrixes which describe the motion patterns expected in each cubid. We classify frames as normal and abnormal by using machine learning methods.