交通信息与安全2013，Vol.31Issue(2)：32-35,4.DOI:10.3963/j.issn1674-4861.2013.02.008
基于高维特征和RBF神经网络的湿滑道路图像判别方法
Classification of Slippery Road Images Based on High-dimensional Features and RBF Neural Network
摘要
Abstract
An image pattern recognition model of slippery roads classification was established. By reducing the dimension of complex multi-dimensional characteristics of images with J criterion, a number of testing conditions were set according to the characteristics and sample size. RBF neural network was used to classify eight different slippery roads, and to analyze the pros and cons, influencing factors and performance improvement methods of RBF neural network for image classification of the slippery road. The results show that, by reducing image features dimensionality, the classification accuracy rate of different slippery roads can reach to 78. 4% by RBF neural network method.关键词
湿滑道路图像/图像模式识别/图像特征/RBF神经网络Key words
slippery road images/ image pattern recognition/ image features/ RBF neural network分类
交通工程引用本文复制引用
万剑,赵恺,王维锋..基于高维特征和RBF神经网络的湿滑道路图像判别方法[J].交通信息与安全,2013,31(2):32-35,4.基金项目
国家自然科学基金项目(批准号:51208394)资助 （批准号:51208394）
