摘要
本文提出的车牌识别技术,在定位环节根据图像清晰度的不同采用不同的边缘检测算子,通过"角点提取"获取候选区位置,通过"模糊外轮廓的汉字识别"进行汉字提取。实验结果表明,本研究对全天候图像的整体定位率大约为97.7%,字符识别率大约为95.6%,整个车牌识别系统的识别率约为93.4%,平均识别时间约为0.5s/幅,识别率有较大的提高。
The license plate recognition technology proposed in this paper adopts different edge detection operators based on difference of image sharpness in the position link, gains the position of candidate area by ＂corner extraction＂, and extracts the Chinese characters through ＂fuzzy outer contour Chinese character recognition＂. Experimental results show that the overall positioning rate of the all-weather image is about 97.7%, the character recognition rate is about 95.6%, the recognition rate of entire license plate recognition system is about 93.4%. and the average recognition time is about 0.5s/web, so the recoznition rate is zreatly improved.
出处
《价值工程》
2015年第4期239-240,共2页
Value Engineering
关键词
车牌识别系统
角点提取
傅里叶描述子
license plate recognition system
corner extraction
Fourier descriptors
