智能交通中运动目标跟踪与svm识别方法研究-信号与信息处理专业论文.docx

Research on Tracking and Recognition of Moving Object based on SVM in Intelligent TrafficAbstractIntelligent Transportation Systems (ITS) has become the development of the future transportation system in recent years, and it has an extremely important scientific significance and broad application prospects. The core of the research in this area is the moving target detection, tracking and recognition based on video. It deals with the issues of high-tech in many fields such as advanced computer technology, image processing, machine vision, artificial intelligence and pattern recognition and so on, thus creating a real-time, accurate and efficient integrated management system.This paper describes the basic theories and key technologies of moving target detection, tracking and recognition in video sequence images. It focuses on study of the real-time tracking and moving target recognition in traffic video based on SVM.For moving target detection, first, It provides an introduction and summary for the three major detection methods (ie, background subtraction, inter-frame difference, optical flow method), which are currently used, and gives the homologous experimental results. Then, It makes a brief overview of the target detection methods based on adaptive background modeling. In order to extract the moving target more accurately, it has done some operations such as median filtering and mathematical morphology processing later in the experiment.For target tracking, we proposed a method by combing CamShift tracking algorithm and Kalman Filter, we have used the class of functions, data structures and some basic framework by OpenCV to establish a system of moving object tracking. This system includes the following modules such as image frame preprocessing module, the prospect of moving target detection module and moving object tracking module. It has achieved the desired results.At last, for moving target recognition, we introduced a theory of image processing based on SVM