A Spectrum Based Algorithm for Image Classification

In this paper, a novel algorithm for image classification is presented which uses the projective value of adjacency spectrum as classified samples. Firstly, the eigenvalues of adjacency matrices constructed on the feature point-sets of images are obtained by singular value decomposition. Secondly, t...

Full description

Saved in:
Bibliographic Details
Published in:电子学报：英文版 2009-07, Vol.18 (3), p.427-430
Main Author: WANG Nian ZHANG Jiang TANG Jun FAN Yizheng LIANG Dong
Format: Article
Language:English
Subjects:
分类算法
协方差矩阵
图像分类
奇异值分解
射影分类
特征空间
矩阵特征值
神经网络分类器
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In this paper, a novel algorithm for image classification is presented which uses the projective value of adjacency spectrum as classified samples. Firstly, the eigenvalues of adjacency matrices constructed on the feature point-sets of images are obtained by singular value decomposition. Secondly, the eigenvalues are projected onto the eigenspace by means of the covariance matrix. Finally, image classification is performed by adopting RBF and PNN neural networks as classifiers respectively. Mean- while, some theoretical analyses are given to support the proposed method.
ISSN:1022-4653