摘要
基于交通标志的颜色和几何形状特征,采用MATLAB工具设计了一种交通标志检测识别方法。该方法由基于HSV空间的颜色抽取、基于仿射变换的几何形状校正及判定、基于Gabor滤波的特征向量提取和基于SVM的分类识别4部分组成。仿真结果表明,该方法减小了投影失真,有效实现了交通标志的检测与分类,且准确率较高。
A detection and recognition method of traffic signs is implemented with MATLAB based ontraffic signs′ color and shape features. This method consists of four parts which are image segmentationbased on HSV color space,detection and affine transformation correction based on geometry features,extraction of eigenvector with Gabor filter,classification and recognition with support vector machine（SVM）. Experimental results show that the proposed method can reduce projecting distortion,accuratelyclassify and identify traffic signs.
出处
《江汉大学学报（自然科学版）》
2016年第2期119-125,共7页
Journal of Jianghan University：Natural Science Edition
关键词
交通标志检测
HSV颜色空间
仿射变换
GABOR滤波
SVM
traffic signs detection
HSV color space
affine transformation
Gabor filter
support vector machine（SVM）
