Video-Object Detection using Background Subtraction in Spartan 3 FPGA Kit

Abstract

In many real time applications, detection of moving objects is necessary. But while post processing, the computational complexity has to be minimum. This work has hardware based moving object detection using background subtraction method for soft real time application is proposed. It aims in performing the automatic detection of moving objects using FPGA hardware with very high performance as well as low cost. With the help of a background reference image, the moving object can easily be detected using background subtraction method. A reliable background updating model is established based on the available data. This gives an accurate and efficient detection of moving objects. Also Sobel's edge detection algorithm is performed to trace the outline shape of the detected moving object. It is implemented using MATLAB tool and Xilinx ISE in Spartan 3 FPGA kit. RS 232 serial port cable is used for establishing the connection between hardware and PC. Visual basics software is used to view the output in PC.