基于学习矢量量化神经网络的交通标识识别技术研究
Research on Traffic Sign Recognition Technology Based on Learning Vector Quantitative Neural Network
丁志成 1赵雷雷1
作者信息
- 1. 郑州工业应用技术学院机电工程学院,河南郑州 451100
- 折叠
摘要
传统交通标识识别技术需要大量数据做训练且对输入数据的要求较高,导致识别结果误差较大,为此,提出基于学习矢量量化神经网络的交通标识识别技术.将预处理交通标识图像作为待识别数据集,构建学习矢量量化神经网络模型,输入待识别图像进行学习与训练后,输出交通标识类别的识别结果,实现交通标识识别.实验结果表明,所设计技术的交通标识识别率高达98.77％,验证了该技术的有效性与正确性.
Abstract
Because the traditional traffic sign recognition technology needs a lot of data for training and requires high input data,which leads to a large error in recognition results.Therefore,a traffic sign recognition technology based on learning vector quantization neural network is proposed.The preprocessed traffic sign images are taken as the data set to be recognized,and the learning vector quantization neural network model is constructed.After inputting the images to be recognized for learning and training,the recognition results of traffic sign categories are output to realize the recognition of traffic sign.The experimental results show that the recognition rate of traffic signs under the designed technology is as high as 98.77％,which proves the effectiveness and correctness of the technology.
关键词
学习矢量量化神经网络/交通标识/标识识别Key words
learning vector quantitative neural network/traffic recognition/recognition technology引用本文复制引用
基金项目
河南省工程技术研究中心项目(豫科实20221号)
&&(2023YB007)
&&(郑工教[2022]5号)
出版年
2024
NETL
NSTL
万方数据