FPGA-based video system for real time moving object detection

Real time detection of moving objects from a video sequence has become an important role in computer vision field. Applications such as traffic control systems, surveillance systems, robotics vision and even more complex tasks such as human face recognition are currently subject of research. Although for some limited video applications the real time movement detection through a serial processor-based system represents a good solution, the use of the inherently concurrent processing and reconfigurability of FPGA´s can be substantially beneficial in processing video algorithms for frames with high resolution. This paper describes the design of a background image subtraction method and its FPGA implementation for detecting a moving object in surveillance video applications with high resolution frames. As proposed system input is a video signal with a resolution of 720 × 480 pixels, and as system output is the frame zone where object movement is taking place. Although the described system is for one object movement, the proposed processing algorithm can be extended for the detection of more objects. According to obtained experimental results the proposed system represents a good option for real time detection of object movements with a speed up to 3 m/s.