基于OpenCV的车辆检测毕业论文.doc

摘 要 智能交通系统(ITS)是目前世界交通运输领域正在研究和广泛关注的课题。 近年来，智能交通系统的应用给交通运输业带来了巨大的经济效益，对于道路设 计、流量监控和高速公路管理起到了越来越重要的作用。论文所研究的视频车辆 检测技术在ITS中占有很重要的地位，与传统的车辆检测方法相比，视频车辆检测技术不仅具有安装维护便捷且费用较低、可监视范围广等诸多优点，同时可对道路现场图像进行智能化分析和处理。 本文利用OpenCV中对运动物体检测的数据结构、函数库，建立了一个视频车辆分析系统，用于道路上车辆的检测。检测过程是首先对视频图像进行预处理，比如二值化、去噪等；然后进行背景的提取及更新，由于背景差分法是通过当前帧和背景帧相减来提取运动目标，所以实时的背景更新尤为重要，本文采用多帧求平均的方法来更新背景，避免了光照条件和气候环境等带来的不利影响；最后利用背景差分法检测出运动车辆。经过实验验证，该方法可以较准确地检测出车辆目标，检测的成功率可达到90%以上。 关键字： 视频处理 车辆检测 OpenCV TOC \o 1-3 \n \h \u ABSTRACT Intelligent Transportation Systems (ITS) is the subject of study and attention in the worlds transportation sector.In recent years, the application of intelligent transportation systems has brought enormous economic benefits to the transportation industry,and it has played an increasingly important role in the road located design, traffic monitoring and highway management.The video vehicle detection technology in this thesis study plays an important role in ITS.Compared with the traditional method of vehicle detection, video vehicle detection technology has easy installation and maintenance,low cost, wide range that can be monitored and many other advantages , while it can do intelligently analysis and processing to the image of the road scene. This thesis sets up a video vehicle analysis system for detecting vehicles on the road by the moving object detection data structure and function library in OpenCV. Detection processes include:First,video image pre-processes,such as binary,removing noise;Next,background extraction and update, as background subtraction is based on that subtract the current frame from background frame to extract moving target,real-time background update is particularly important, we use multi-frame averaging method to update the background,which avoids the the adverse impact of light conditions and climatic and environmental;Finally, the detection of moving vehicles by background subtraction.Experimental results show that this method can accurately detect t