基于递增权函数的邻接矩阵与非负矩阵分解的图像分类方法  

Image classfication using increasing weighting function of adjacency matrix of graph and non-negetive matrix factorization

作　　者：蒋云志[1] 王年[1] 汪斌[1] 

机构地区：[1]安徽大学电子信息工程学院,安徽合肥230039

出　　处：《安徽大学学报（自然科学版）》2011年第5期63-67,共5页Journal of Anhui University(Natural Science Edition)

基　　金：国家自然科学基金资助项目(60772121);安徽省高校青年教师基金资助项目(2008JQ1023);安徽省教育厅自然科学研究基金重点资助项目(KJ2010A007)

摘　　要：将递增权函数的邻接矩阵和非负矩阵分解方法相结合,应用于图像分类.首先由图像中提取的特征点构造递增权函数的邻接矩阵,再对其进行非负矩阵分解,用分解后的特征向量作为PNN分类器的输入,实现对图像的分类.算法的可行性和准确性通过模拟图像和真实图像的多组实验得到了验证.In this paper,the adjacency matrix of graph based on the increasing weighting function combined with the method of non-negative matrix factorization was applied to the image classification.First,the character points could be distilled from different images.Then,these points were used to construct the adjacency matrix of the increasing weighting function,and the eigenvector of the image could be obtained by the non-negative factorization of the adjacency matrix.Finally,the eigenvector was put into PNN（Probabilistic Neural Network）classifier to accomplish the image classification.Several groups of experiments were presented between simulating images and real images.The results showed that the method presented in this paper was feasible and accurate.

关 键 词：递增权函数 邻接矩阵 非负矩阵 图像分类 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]