摘要
针对现有的车牌识别方法存在车牌无法定位且车牌字符无法正确分割等情况,提出了一种基于卷积神经网络的车牌识别技术。首先,设计了一套图像处理流程实现车牌定位和字符分割,然后,利用提出的卷积神经网络对车牌字符集进行训练、识别。所提方法可以达到98.54%以上的准确率,极大提高适用性和准确率。
In view of the existing license plate recognition methods,the license plate cannot be located and the license plate characters cannot be correctly separated.This paper proposes a license plate recognition technology based on convolutional neural network.Firstly,a set of image processing flow is designed to realize license plate location and character segmentation.Then,the proposed convolutional neural network is used to train and recognize the license plate character set.The proposed method can achieve more than 98.54%accuracy,greatly improving the applicability and accuracy.
作者
李涛涛
Li Taotao(School of Mechanical Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China)
出处
《农业装备与车辆工程》
2021年第5期119-121,共3页
Agricultural Equipment & Vehicle Engineering
关键词
车牌识别
卷积神经网络
深度学习
车牌定位
license plate recognition
convolutional neural network
deep learning
license plate location
