基于机器视觉的道路拥堵状态检测的研究

发布时间：2018-01-29 17:41

本文关键词： 机器视觉道路拥堵检测静止存在采样点车辆模型匹配算法　出处：《小型微型计算机系统》2014年01期 　论文类型：期刊论文

【摘要】：针对道路拥堵检测难的问题提出一种基于机器视觉的道路拥堵状态检测方法.为了能在嵌入式系统上实现道路拥堵状态的视觉自动化检测,首先通过定制道路区域并自动生成均匀分布的采样点,采用以点代面的设计思想来减少图像处理的计算资源和存储资源;其次,将背景差法和帧间差法相结合处理得到反映道路拥堵状态的静止存在采样点,采用车辆模型匹配算法检测道路上的拥堵区域并计算得到排队长度.实验结果表明,本文提出的检测算法具有计算效率高、耗费资源少、检测范围广、鲁棒性强等优点,能快速并准确地检测出在道路拥堵区域和拥堵状态.
[Abstract]:Aiming at the difficulty of road congestion detection, a new road congestion detection method based on machine vision is proposed in order to realize the automatic detection of road congestion on embedded system. Firstly, by customizing the road area and automatically generating uniformly distributed sampling points, the design idea of point replacement plane is adopted to reduce the computing and storage resources of image processing. Secondly, the background difference method and the frame difference method are combined to get the static sampling points which reflect the road congestion. The vehicle model matching algorithm is used to detect the congested area on the road and the queue length is calculated. The experimental results show that the proposed algorithm has the advantages of high computational efficiency, less resource consumption and wide detection range. It has the advantages of strong robustness, and can detect the congested area and congestion state quickly and accurately.
【作者单位】：浙江工业大学信息工程学院;
【基金】：国家自然科学基金项目(61070134)资助
【分类号】：TP391.41;U491.116
【正文快照】： 1引言随着城市化进程的加快和交通量的急剧增长,城市的道路变得越来越拥挤,交通拥堵甚至交通瘫痪时有发生,给城市交通带来了严重挑战.交通拥堵是一个感觉量,在确定交通拥堵时存在模糊性,难以直接测量.目前的研究主要是从采集交通状态信息出发,分析归纳出交通规律,并制定道路拥

【参考文献】