Study On Pedestrian Detection Based On Video

The traffic event detection based on the video-based traffic detection technology has got more and more attention and approval because of its easy implementing, high accuracy and convenient operation. The pedestrian monitoring technology is one of the main technologies in video-based traffic detection. And it's also an important aspect of traffic event detection.The pedestrian detection algorithms are researched in this paper. The main contents are as follows:(1) This paper is based on block background extraction and binary algorithm. By calculating the average of blocks, each block is extracted background and binary. As to the problems in binary, this paper adopts the morphological filtering to get complete target objects.(2) As to binary regional labeling of the pedestrian target, this paper adopts a method based on regional characteristics. This method uses the area got from regional marker and regional minimum circumscribed rectangular aspect ratio as regional features to judge pedestrian objects.(3) Pedestrian objects are judged by the trajectory of the center of gravity of regional characteristics of pedestrians. Because the speed difference of pedestrians and vehicles is big, we analyses the trajectory of regional gravity, thus we can obtain the speed of the target area. According to the speed, pedestrian goals can be distinguished.(4) The method of tracking regional characteristics is adopted to distinguish pedestrian. Selecting the feature of target and using the matching template to track the target. Then the pedestrian can be judged by analyzing the speed and smooth level of tracking trajectory.This design of the algorithm is on-site experimented in tunnel. Experimental results show that this algorithm can well solve pedestrian problem. Thus the pedestrian in traffic alarms very fast.