FPGA Implementation for Background Subtraction Based on Gaussian Mixture Models
Han Jian-hu · Harbin Ligong Daxue xuebao · 2013
Aiming at the real-time problem of moving object detection in video,this paper proposed a FPGA realization method of background subtraction algorithm based on Gaussian mixture model. Firstly,the algorithm model is built for image sequence,and the background model of moving scene is obtained. Then the background model is updated timely according to scene change in moving object segmentation. Considering the real-time requirements of high frame rate and high resolution,the paper modified the Gaussian mixture model for FPGA hardware implementation. The test results show that the algorithm is correct in the hardware implementation,and the real-time performance is improved.