本文已被:浏览 1225次 下载 578次
投稿时间:2020-03-27
投稿时间:2020-03-27
中文摘要: 通过与传统神经网络对比,分析了利用卷积神经网络(CNN)进行车牌号图像识别中的特征提取过程,提出了优化卷积和池化的过程来提高算法的收敛速度和准确率。运用PyCharm环境建立了改进后的车牌号识别模型,并通过实验验证了其正确性与识别速度。通过BP神经网络、传统LeNet-5 CNN和改进后的CNN对相同的字符集进行对比分析实验,得出了改进后的CNN模型的优势。
Abstract:Compared with the traditional neural network,this paper analyzes the feature extraction process in image recognition of autombile plate number using convolutional neural network,and proposes the process of optimizing convolution and pooling to improve the convergence speed and accuracy of the algorithm.The improved license plate recognition model is established in PyCharm environment,and its correctness and recognition speed are verified.By comparing the same character set with BP neural network,conventional LeNet-5 CNN and the modified convolution neural network,the advantages of the proposed model are obtained.
文章编号:20204006 中图分类号:TP399 文献标志码:
基金项目:
作者 | 单位 | |
刘永雪 | 上海电力大学 计算机科学与技术学院 | 407843800@qq.com |
李海明 | 上海电力大学 计算机科学与技术学院 |
引用文本:
刘永雪,李海明.卷积神经网络的优化在车牌号识别上的运用[J].上海电力大学学报,2020,36(4):351-356.
LIU Yongxue,LI Haiming.Optimization of Convolutional Neural Network for License Plate Recognition[J].Journal of Shanghai University of Electric Power,2020,36(4):351-356.
刘永雪,李海明.卷积神经网络的优化在车牌号识别上的运用[J].上海电力大学学报,2020,36(4):351-356.
LIU Yongxue,LI Haiming.Optimization of Convolutional Neural Network for License Plate Recognition[J].Journal of Shanghai University of Electric Power,2020,36(4):351-356.