《基于SVM车型识别系统的设计与实现》-毕业论文（设计）.docVIP

PAGE 0 第 PAGE 18 页 共 22 页 陕西理工学院毕业设计 基于SVM车型识别系统的设计与实现 [摘要]车辆自动识别分类技术是智能运输系统的重要组成部分，它对特定地点和时间的车辆进行识别和分类，并以之作为交通管理、收费、调度、统计的依据。要实现我国公路收费自动化、管理规范科学化，车型自动识别方法的研究势在必行。本文研究基于车型图像代数特征的车型识别方法。该方法首先利用背景差分法从背景图像中提取出运动车辆，并对车型图像进行预处理，然后采用特征并行融合的方法用PCA方法，最后通过支持向量机分类器进行车型识别。 [关键词]车型识别，特征融合，特征提取，决策支持向量机 Vehicle Identification System Based On SVM Design And Implementation Abstract: Automatic Vehicle Identification and Classification of Intelligent Transport System technology is an important part of its specific place and time the identification and classification of vehicles, and used as traffic management, fees, scheduling, statistical basis. Chinas road toll to achieve automation, standardized and scientific management, Automatic Vehicle Identification Method imperative. This paper explores the characteristics of the vehicle model image algebra recognition. This method first uses background subtraction to extract from the background image moving vehicles, and vehicle image is preprocessed, and then use the parallel feature fusion method using principal component analysis, and finally through the support vector machine classifier for vehicle identification. Keyword: Vehicle recognition, feature fusion, feature extraction, decision support vector machine 目录 TOC \o 1-3 \h \z \u HYPERLINK \l _Toc263851709 1 引言 PAGEREF _Toc263851709 \h 2 HYPERLINK \l _Toc263851710 1.1车型识别的研究内容及其应用 PAGEREF _Toc263851710 \h 2 HYPERLINK \l _Toc263851711 1.2车型识别的研究现状和前景 PAGEREF _Toc263851711 \h 3 HYPERLINK \l _Toc263851712 1.3本文研究的主要内容及结构安排 PAGEREF _Toc263851712 \h 3 HYPERLINK \l _Toc263851713 2 车辆图像预处理 PAGEREF _Toc263851713 \h 4 HYPERLINK \l _Toc263851714 2.1图像平滑 PAGEREF _Toc263851714 \h 4 HYPERLINK \l _Toc263851715 2.2图像标准化 PAGEREF _Toc263851715 \h 5 HYPERLINK \l _Toc263851716 3 车型特征提取 PAGEREF _Toc263851716 \h 7 HYPERLINK \l _Toc263851717 3.1 PCA方法概述 PAGEREF _Toc263851717 \h 7 HYPERLINK \l _Toc263851718