基于无参数形状检测子和DT-CWT的交通标志识别  

Traffic Sign Recognition Based on Parameter-free Detector and DT-CWT

作　　者：谷明琴[1,2] 蔡自兴[1] 

机构地区：[1]中南大学信息科学与工程学院,长沙410083 [2]奇瑞前瞻技术科学院,安徽芜湖241009

出　　处：《计算机研究与发展》2013年第9期1893-1901,共9页Journal of Computer Research and Development

基　　金：国家自然科学基金项目(90820302;60805027);教育部高等学校博士学科点专项科研基金项目(195470);湖南省院士基金项目(20010FJ4030);湖南省自然科学基金项目(12JJ6058)

摘　　要：针对车辆行驶环境中难以检测的交通标志,提出了一种检测和识别方法.首先分割交通标志的特征颜色区域,并扩展感兴趣区域,提取区域边缘.然后用直线分割和杂点去除粗略划分边缘,根据直线间顶点处的曲率关系,计算转向角并分类顶点的类型,用无参数形状检测子来检测图像中的圆形、三角形和矩形等.将检测到的候选区域送入形状分类器中,分类形状并排除杂质的干扰,最后通过二元树复小波变换和二维独立分量分析相结合来识别交通标志类型.实验结果表明提出的方法对交通标志被遮挡、光照不均匀、颜色部分失真的情况下,检测率和识别率均较高,并且可以达到实时处理的效果.For the traffic sign which is difficult to detect in traffic environment, a traffic sign detection and recognition algorithm is proposed. First, the main colors of the traffic sign are segmented, the region of interest is expanded and its edge is extract. Then the edge is roughly divided by drawing linear and removing miscellaneous points. Turing angle curvature is computed according to the relations among the curvatures of the vertices. Then the vertex type is classified. The standard shapes such as circular, triangle, rectangle, are detected by parameter-free detector. The candidate regions are sent into the shape classifier to classify the type and exclude the interference. Finally, the type of traffic sign is recognized by dual tree complex wavelet transform and two-dimensional independent component analysis. The experimental results show that the detection and recognition rate of the proposed algorithm is high for the conditions such as traffic signs obscured, uneven illumination, color distortion, and it can achieve the effect of real-time processing.

关 键 词：无参数形状检测子 曲率 形状分类 二元树复小波变换 交通标志识别 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]