基于视频分析的车流量统计算法研究

车流量统计作为智能交通系统的关键技术之一，逐渐成为研究的热点。主要研究了基于虚拟线圈的车流量统计算法和基于目标跟踪的车流量统计算法，综合了两种算法的优势，结合Canny算子边缘检测算法、均值背景建模法和背景差分法等技术，提出了一种基于质心的车流量统计综合算法。实验结果表明综合算法在实时性和准确性方面均取得较好的实验结果。

THE ALGORITHM RESEARCH OF TRAFFIC FLOW MEASURING BASED ON VIDEO ANALYSIS

Statistics of the traffic flow, as one of the key technology of intelligent transportation system, gradually become a research hotspot. We mainly studied two kinds of existing algorithms for statistics of the traffic flow： traffic statistics algorithm based on virtual loop and traffic statistics algorithm based on target tracking. Furthermore, we integrated the advantages of both methods and combined with some technology Canny operator edge detection algorithm, the mean background modeling method and background difference method. Finally, we proposed an improved traffic statistics algorithm based on mass. Experimental results show that integrated algorithm achieves good results in real-time and accuracy.