首页 | 本学科首页   官方微博 | 高级检索  
     检索      

应用SVM的三维车型识别技术
引用本文:陈宏,何小海,吴炜,杨晓敏.应用SVM的三维车型识别技术[J].四川大学学报(自然科学版),2006,43(6):1279-1284.
作者姓名:陈宏  何小海  吴炜  杨晓敏
作者单位:四川大学电子信息学院图象信息研究所,成都,610064
摘    要:提出了一种新的车型识别技术.该技术分成两大部分,第一部分是如何准确地测量车辆的三维信息.首先需要从一系列图象中准确检测出车辆,然后从背景中将车辆完整的分割出来,再使用一个简单的透视变形模型,完成从二维信息到三维信息的转换,获得车辆的长度、宽度和高度特征.第二部分是将这些特征输入一个SVM分类器进行识别处理.实验证明,该方法在车型识别中,具有比较高的识别精度.

关 键 词:图象处理    图象投影模型    三维特征    支持向量机
文章编号:0490-6756(2006)06-1279-06
收稿时间:2006-03-31
修稿时间:2006-03-31

Three-dimensional Characters Recognition for Vehicle Type Identification Using SVM
CHEN Hong,HE Xiao-hai,WU Wei,YANG Xiao-min.Three-dimensional Characters Recognition for Vehicle Type Identification Using SVM[J].Journal of Sichuan University (Natural Science Edition),2006,43(6):1279-1284.
Authors:CHEN Hong  HE Xiao-hai  WU Wei  YANG Xiao-min
Institution:College of Electronics and Information; Sichuan University,College of Electronics and Information; Sichuan University,College of Electronics and Information; Sichuan University,College of Electronics and Information; Sichuan University
Abstract:Presents a new technology for vehicle type identification.This first part is how to get the three-dimensional characters of vehicle from image.First get the vehicle in sequences image and extract the vehicle from background.Then,we transform the two-dimensional image to three-dimensional image with a simple projection model for getting the real length and width of vehicle.At last we input these characters into a SVM classifier.This method have the more precision compared with other methods.
Keywords:image processing  projection model  three-dimensional characters  SVM
本文献已被 CNKI 维普 万方数据 等数据库收录!
点击此处可从《四川大学学报(自然科学版)》浏览原始摘要信息
点击此处可从《四川大学学报(自然科学版)》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号