马苗, 陈芳, 郭敏, 陈昱莅. 基于改进LeNet-5的街景门牌号码识别方法[J]. 云南大学学报(自然科学版), 2016, 38(2): 197-203. doi: 10.7540/j.ynu.20150560
引用本文: 马苗, 陈芳, 郭敏, 陈昱莅. 基于改进LeNet-5的街景门牌号码识别方法[J]. 云南大学学报(自然科学版), 2016, 38(2): 197-203. doi: 10.7540/j.ynu.20150560
MA Miao, CHEN Fang, GUO Min, CHEN Yu-li. A recognition method based on improved LeNet-5 for street view house numbers[J]. Journal of Yunnan University: Natural Sciences Edition, 2016, 38(2): 197-203. DOI: 10.7540/j.ynu.20150560
Citation: MA Miao, CHEN Fang, GUO Min, CHEN Yu-li. A recognition method based on improved LeNet-5 for street view house numbers[J]. Journal of Yunnan University: Natural Sciences Edition, 2016, 38(2): 197-203. DOI: 10.7540/j.ynu.20150560

基于改进LeNet-5的街景门牌号码识别方法

A recognition method based on improved LeNet-5 for street view house numbers

  • 摘要: 以真实场景中拍摄的街景门牌号码图像数据集SVHN为研究对象,将卷积神经网络与支持向量机相结合,提出了一种基于改进LeNet-5的街景门牌号码快速识别方法.该方法首先对数据进行图像增强预处理,突出有效特征;然后,省去基本LeNet-5中的第3卷积层,并用SVM分类器代替最后输出层中的Softmax分类器,以简化网络结构的同时提高分类效率.在国际公开的SVHN数据集的实验结果表明,改进LeNet-5可以有效识别街景门牌号码,7h便可训练得出结构稳定的网络识别模型,识别率达到90.35%,提高了算法的综合效率.

     

    Abstract: Focusing on the SVHN dataset acquired in real environment,a recognition method based on an improved LeNet-5 model is proposed for street view house numbers,which combines convolutional neural networks with support vector machines.In this method,after original color images are transformed to grayscale images,some images enhancement are employed to highlight the number features.Then,the third convolutional layer is omitted and the softmax classifier of the final output layer is replaced with SVM classifier in the LeNet-5 model,which not only simplifies the network construction,but also helps to obtain a higher recognition rate.Experimental results show that the recognition rate of the suggested method may reach 90.35% after the sample images were trained only 7 hours,which efficiently improved the overall ability of the recognition.

     

/

返回文章
返回