Study Of Support Vector Machine And Its Applications In Video Traffic Information Detection System

Statistical Learning Theory (SLT) is a learning theory which specializes in machine learning with finite samples. As a learning method based on SLT, Support Vector Machine (SVM) has the advantages of global solutions, well adaptively, high generalization ability and maturity in theory. SLT and SVM are the hot-spot in the field of machine learning nowadays.With the development of urban traffic management, the detection of traffic information is paid more and more attention. Video detection is superior to other methods. Video Traffic Information Detection System is a computer processing system using image processing technologies and aims to gather traffic information such as passed vehicle count, vehicle speed and vehicle types.In this thesis, we introduce SVM to the system of Video Traffic Information Detection;the main work is described as follows:(1) We summarize the latest research achievements and development of SLT;present the conceptions of SLT and the principles of SVM;(2) Introduce the superiority of video traffic information detection technologies, present the theories and methods that are used in a Video Traffic Information Detection System;(3) Accomplish a system of video traffic information detection using current method and analyze its function and shortness.(4) Apply the theory of SVM to Video Traffic Information Detection System. Through classifying the traffic flow images in detection area and traces vehicles to get the information of passed vehicle count, vehicle speed and vehicle types. Lastly, we take a comparison between SVM method and current method.