Abstract: Computer Vision Technology has been widely used in the engineering industry and daily life, but not yet deep into urban design. The advantage of its accurate monitoring can avoid the deviation of the human eye observation, it does not bring the experimental error to the tester as a portable test instrument. Urban street space is one of the most important urban public spaces, carrying the rich daily life of the citizens, is a great place to study the association of behavior and space. The introduction of computer vision technology into urban street space design is a new attempt, different from traditional methods of human eye observation or image analysis. The purpose of this paper is to obtain the accurate behavior spatial association image by analyzing the frame difference between the successive video cameras, and to promote the development of precision design and research of urban public space.