复杂环境下基于ROI的车牌提取算法  

ROI-based license plate extraction algorithm under complex environment

作　　者：陈龙[1] 侯普华[1] 

机构地区：[1]上海理工大学机械工程学院,上海200093

出　　处：《信息技术》2013年第11期95-98,共4页Information Technology

摘　　要：感兴趣区域(ROI)是最能体现图像内容的区域,基于ROI的图像特征提取技术有效提高了图像处理和分析的效率,在图像处理与分析领域有着重要的应用。首先简要介绍现有的车牌提取方法,针对目前复杂环境下的车牌ROI提取算法提取效果不明显、计算繁琐、漏检率高等缺点,提出一种采用颜色特征和模板匹配的车牌ROI提取算法,该算法避免大量运算,具有相对于图像平移、尺度变化的低敏感性,并且取得了较好的实验效果。The region of interest （ROI） is the area that best embodies the image content,ROI-based image feature extraction technology improved the efficiency of image processing and analysis,has important applications in the field of image processing and analysis.This paper briefly describes the existing license plate extraction method and then proposed a license plates ROI extraction algorithm using color features and template matching,for the effect of license plate extraction algorithm under complex environment is not obvious,tedious calculation and high undetected shortcomings.This algorithm avoids a large number of operations,and has a low sensitivity with respect to image translation,scale change,and achieved good experimental results.

关 键 词：ROI提取 车牌提取 复杂环境 颜色特征 模板匹配 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]