基于Sobel垂直算子检测的车牌定位算法  

Algorithm based on sobel vertical operator for location of number plate

作　　者：安化萍[1] 丁劲生[1] 高满屯[1] 

机构地区：[1]西北工业大学机电学院

出　　处：《计算机工程与设计》2005年第8期2030-2032,共3页Computer Engineering and Design

基　　金：航空基础科学基金项目(03I53062)

摘　　要：提出一种新的车牌识别的定位算法,对已预处理增强的图像进行Sobel垂直算子检测,用阈值分割出前景,搜索出符合车牌字符特征的候选区域,对候选区域进一步分析定出牌照位置。算法定位速度快,准确率高,对夜间有反光的图片能有效确定牌照的位置。The identification of number plate is the important constituent parts of intelligence traffic system, and it keeps itself in the foreground of widespread availability. A new location algorithm to the identification of number plate was proposed. It detected the preprocessed and enhanced image with sobel vertical operators, segmented the foreground of the image with threshold, search selected regions that accorded with characteristics of the number plate, and analyzed and located the position of the number plate from the selected regions. This location algorithm has high speed and accuracy. It can exactly locate the situation of number plate to reflected image by night.

关 键 词：智能交通 车牌识别 Sobel垂直算子检测 候选区域 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]