Traffic Volume Statistics Algorithm Due To The Vehicle Blocked

Most of the traffic flow parameters statistics methods based on video complete the processing of bodywork obscured or bodywork division, adhesion by tracking system, we raise an algorithm of traffic flow statistics method based on virtual wireframe in this paper. We by splitting virtual wireframe, and after the split virtual line box target connected domain tracking and matching to detect the target vehicle occlusion. The connected domain in order to solve the target vehicle differential binarized hollow, distal large body connected domain splitting, this article uses a series of morphological algorithm processing for virtual line box differential binarization connected domain, in order to improve the accuracy of the traffic flow statistics algorithm proposed proximal virtual line box vehicle traffic to the remote line box statistical error correction. The experimental results show that this article has better results in terms of the target vehicle detection and statistical. The main work of this article as follow:Firstly, we described the core technology of traffic statistics algorithm based on video. The core technology of the traffic flow statistics algorithm described in this paper could be divided into four parts:background extraction, histogram equalization, morphological processing and virtual wireframe detection. And we give a brief analysis and comparison of the advantages and disadvantages of various algorithms, which is focusing on the morphological processing module description.Secondly, we propose a comprehensive traffic statistics algorithm combined morphological processing and virtual the wireframe detection. We do statistics of the occluded targets separately through the Segmentation of virtual wireframe according to the processing in the detection period. And we raise a matching algorithm, which also includes error correction part. And we test the algorithm with a good result that can be applied in practice. In this paper, we program the algorithm based on the actual highway surveillance video. Our research is starting from the actual testing situation. Combined with the related literature, we design a real-time and efficient traffic flow statistical algorithms for the blocked car, which hope to provide a copies of the new ideas of the traffic flow statistics algorithm study.