基于主元神经网络和K-均值的道路识别算法  

Road Recognition Algorithm Using Principal Component Neural Networks and K-Means

作　　者：程洪[1] 郑南宁[1] 高振海[1] 李青[1] 

机构地区：[1]西安交通大学电子与信息工程学院,西安710049

出　　处：《西安交通大学学报》2003年第8期812-815,共4页Journal of Xi'an Jiaotong University

基　　金：国家自然科学基金资助项目 (60 0 2 43 0 1 ).

摘　　要：为了提高道路识别算法的鲁棒性和自适应性,提出了基于局部统计特征和主元分析的道路识别算法.该方法用广义Hebb学习规则训练主元神经网络权值,然后将局部统计特征和图像像素值输入主元神经网络得到图像特征矢量,最后用K 均值分类器对该矢量进行分类,通过参考区域识别道路.仿真结果表明,该算法对于光照变化剧烈和阴影遮挡的道路图片均有较好的识别效果,以及较好的鲁棒性和自适应性.A new road recognition algorithm based on local statistical features and principal component analysis is introduced to improve its robustness and adaptiveness. The weights of the principal component neural networks are trained with the aid of the algorithm of generalized Hebbian learning rule, and the input vectors of the local spatial features and image pixels value are transformed into feature vectors which are once clustered by K-means classifier, the road surface and un-road surface can be distinguished by the reference area finally. The simulation results confirm the fine robustness and adaptiveness of the newly proposed algorithm, especially, the improved performance to recognize road images affected by illumination variations or shadows.

关 键 词：学习规则 主元神经网络 K-均值 道路识别 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置]