Infrared image moving object detection based on image block reconstruction

Background subtraction is the most popular algorithm for moving object detection at present. It can detect the object position accurately and rapidly. Background extraction is the most important factor that impacts the detection results greatly. With regard to the problem of obtaining background frame in background subtraction algorithm, the dissimilarity discrimination method for detemining the blocks of background was introduced and a new infrared object detection algorithm based on image block reconstruction was presented. The algorithm first carried on the reasonable sub-block processing to the image. Then the dissimilarity of the corresponding blocks of the adjacent frame was used to discriminate the corresponding blocks of background from the frame. At last, the background frame was built through the combination of the blocks of background. Experimental results show that the algorithm has the characteristics of fast operation and strong anti-noise ability and it can detect the moving object effectively.