BP算法在车牌汉字识别中的应用
pplication of BP Algorithm in Recognition of Chinese Character in License Plate Image
由于汉字本身的复杂性，车牌识别系统识别车牌中的汉字比较困难，这已成为我国智能交通发展的瓶颈之一。针对汉字识别问题，利用学习率可变的动量BP神经网络学习算法，应用于智能交通系统对车牌中汉字的识别。通过图片预处理对车牌照片进行分割，提取出汉字字符作为网络训练的样本，利用样本训练BP网络，并通过样本集和测试集对网络的识别能力进行检测，其中样本集识别的正确率是100%，测试集识别的正确率是92.5%，取得了良好的效果。
ue to the complexity of Chinese characters,it is very difficult to recognize them with a plate recognition system. So it has become one of bottlenecks of China's intelligent transportation development. In order to solve the problem ,apply the momentum BP neural network algorithm of adaptive learning rate to recognize characters in transport intelligence system. Separate charaters from each other with the pretreatment of licence plate images, and extract Chinese characters as examples to train the BP net.After then,test the ability of recognition characters with sample sets and test sets. Among them,the accurate rate to recognize the samples is 100% and the other is 92.5%.Good results have been achieved in the research.
李世银、孙茜
公路运输工程自动化技术、自动化技术设备
神经网络汉字识别动量BP算法智能交通
neural networkChinese character recognitionBP algorithmtransport intelligence
李世银,孙茜.BP算法在车牌汉字识别中的应用[EB/OL].(2011-07-08)[2025-08-02].http://www.paper.edu.cn/releasepaper/content/201107-123.点此复制
评论