摘要
针对车辆识别过程中的车牌定位、字符识别等阶段干扰因素较多,而国内开发的自动识别系统识别率和误判率没有达到标准要求,提出采用人工神经网络对车牌定位、字符识别等关键阶段进行处理的方法。并对当前车牌识别技术的现状、技术种类及水平进行全面剖析,总结相关识别技术的核心算法。利用神经网络较好的容错能力、自适应及抗干扰能力,可以有效解决车牌信息采集干扰较大、信息不全等问题,表现出良好的应用前景。
As the large disturbance existing in license plate location,character recognition and other processes,and domestic development of the automatic identification system recognition rate and false positive rate do not meet standard requirements,we propose artificial neural network method to solve license plate location,character recognition,and other key stage processing.This paper analyzes the current status of license plate recognition technology,technology type and level,summarizes the core recognition algorithm.Artificial neural network with good fault tolerance,adaptive,and anti-jamming capability,is an effective solution to the license plate information collection interferences,incomplete information and other issues,showing a good prospect.
出处
《宿州学院学报》
2011年第2期61-63,共3页
Journal of Suzhou University
基金
宿州学院智能信息处理实验室开放课题(2010YKF13)
宿州学院大学生科研重点项目(KYLXKZD01)
关键词
车牌识别
图像处理
定位
匹配
神经网络
License Plate Recognition
image processing
location
match
neural network
