A novel and efficient moving target detection method is proposed. The method by improving the seven-frame difference and background subtraction to eliminate the moving object detection of "empty" and false targets. Through the optical flow computation model for the weight function and the introduction of universal dynamic image model to establish a new optical flow constraints, it can solve common optical flow field calculation time-consuming and the unestablished problem caused by brightness changes of constraint equation, and also obtain accurate information campaign; Finally, threshold segmentation and morphological processing is completed to extract the target. Analysis indicate that the proposed method can detect and segment the moving target accurately and fastly.