提出了浅析特征颜色和sncc的交通标志识别与跟踪方案.docVIP

提出了基于特征颜色和SNCC的交通标志识别与跟踪方案，该方案用于智能车辆的交通标志识别与跟踪．在YCbCr色彩空间对交通场景图像进行颜色阈值分割，提取交通标志所在区域．根据标准交通标志的背景和内景特征颜色，设计标准交通标志的背景和内景颜色特征匹配模板．以背景和内景颜色特征匹配模板作为模板，采用模板匹配技术和SNCC计算匹配度，识别潜在交通标志图像和交通标志．在Matlab／Simulink环境下，基于视频与图像处理模块集和用户自定义函模块构建仿真模型，进行仿真研究，给出了交通场景视频中交通标志的识别和跟踪轨迹仿真结果．结果表明，采用本文的交通标志识别与跟踪方案，计算量少，提高系统的效率，达到了较好的识别效果． 关键词： 智能交通；交通标志识别；简化归一化互相关法；智能车辆；YCbCr空间；模板匹配：跟踪轨迹 Traffic sign detection and recognition of vehicle autonomous navigationhis paper proposes a scheme of traffic signs recognition and tracking based on feature color andSNCC algorithm，which is used in traffic signs recognition and tracking of intelligent vehicles．In YCbCrcolor space，color segmentation of the traffic scene images is calculated，and traffic sign regions areobtmned．According to background and interior color feature of the standard traffic signs，background andinterior color matching templates of traffic signs color feature aredesigned．The background and the interiorcolor matching templates are as the matching templates，the template matching technology and the SNCCalgorithm to calculate matching degree are used to draw and recognize the images of traffic signs．In the Matlab／Simulink environment，the simulation model is established by using video and image processingblockset and User·Defined Functions module．Based on simulation model，recognition results and thetracking of the traffic signs in the traffic scene video aregiven．The experimental results show that thescheme of traffic signs recognition and tracking canimprove the efficiency of the system，and has lesscalculation amount，and has the better recognition effect． Key words： intelligent transportation；traffic sign recognition；simplified normalized cross correlation；intelligent vehicles；YCbCr color space；template matching；tracking 近年来，随着社会的进步，人们对图像及视频的质量要求越来越高。因此，图像视频的处理也越来越受到人们重视。通常，从摄像仪和扫描仪所获得的颜色，以及计算机显示所用的颜色都是RGB色空间。如果不进行转换，在传输和处理过程中，会占用大量的带宽和存储量，不利于传输和处理。为此，人们找到了另一种色空间YCbCr。因为人们对亮度的变化比对色度的变化更为敏感，所以减少带宽引起的颜色损失很小，人眼几乎感觉不出来。基于以上原因，需要找出一种RGB到YCbCr的转换方法。在