A Moving Object Detection Method Based on Improved Three-frame Difference Algorithm and Edge Information

The moving object detection for expressways plays a significant role in intelligent transportation. A moving object detection approach of that is proposed, which combines the improved three frame difference with edge information. For obtain the marginal information of images, first the moving object's marginal information should be extracted by Sobel operator, and then compares with the results of the improved three frame difference. Finally, a completed detection area of moving object will be dealt with by the OTSU threshold segmentation algorithm and morphology. The experimental result has verified this approach can effectively improve the blank and false edge phenomenon which can be found in the traditional one. Not only that, this approach has less calculation, high real-time ability and great stability.