摘要
车牌定位是车牌识别系统的关键技术,定位的准确与否直接影响车牌识别的结果。车牌区域具有空间特征——边缘信息丰富,颜色空间信息——伴生与互补特性。若利用边缘特征丰富进行定位,受对比度及噪声影响较大;利用颜色信息受车身颜色及光照影响较大,误识率较高。针对这一特点,提出了一套融合灰度边缘检测与车牌区域特有的颜色特征,准确进行车牌定位的算法。对200幅各种情况下从交通卡口获取的实测彩色图像进行试验,准确定位率为99.5%。
The vehicle license plate location was the key technique of the vehicle license plate recognition system; the accuracy will influence the plate recognition result. The area of license plate has space feature with wonderful edges and colorful feature with concomitant and complement feature. The location accuracy should be influenced by noises or contrast using edges information and by vehicle-body＇ s color or visible light using color feature. For this point the paper shows an arithmetic fusing gray edge detection and plate area color feature to locate license plate. The experiments on 200 automobile RGB real color images that were taken under various conditions from traffic stations show the correct location rate of 99.5%.
出处
《计算机应用研究》
CSCD
北大核心
2007年第8期209-211,共3页
Application Research of Computers
基金
国家自然科学基金资助项目(60541001)
全国优秀博士学位论文作者专项基金项目(200443)
关键词
车牌定位
HSV颜色特征
边缘检测
融合
vehicle license plate location
HSV color feature
edge detection
fusion
