桂林理工大学学报2013，Vol.33Issue(1)：80-85,6.DOI:10.3969/j.issn.1674-9057.2013.01.015
基于纹理特征的高分辨率遥感影像分类方法
Remote Sensing Image Classification Based on Texture Features
摘要
Abstract
Gray level co-occurrence matrix can reflect the space information of different pixel position,and wavelet transformation expresses the multi-scale image.This paper gives full play to the characteristics of GLCM and wavelet transformation,and extracts texture feature by combining them.The feature vector from the low-dimensional space is mapped into a high dimensional space and the optimal separating hyperplane is found in a high dimensional feature space through the support vector machine by solving the optimization problem so as to solve the classification problem of complex data.SVM parameter optimization using genetic algorithms can in avoid the excessive learning and less learning state.The classification of the high-resolution remote sensing image is established based on support vector machine classification model utilizing GLCM and wavelet transformation to extract texture features as the classification of feature vectors.Classification experiment shows the effectiveness of the method in this study.关键词
纹理特征/遥感影像分类/灰度共生矩阵/小波变换/支持向量机/遗传算法Key words
texture features / remote sensing/ gray level co-occurrence matrix (GLCM) / wavelet transform /support vector machine(SVM) / genetic algorithm分类
信息技术与安全科学引用本文复制引用
韦春桃,王宁,张利恒,原凯敏,邹瑄..基于纹理特征的高分辨率遥感影像分类方法[J].桂林理工大学学报,2013,33(1):80-85,6.基金项目
国家自然科学基金项目(41161073) （41161073）
广西青年科学基金项目(2012GXNSFBA053131) （2012GXNSFBA053131）
广西空间信息与测绘重点实验室研究基金项目(桂科能1103108-17 （）
1207115-03) （）
广西自然科学基金创新研究团队项目(2012GXNSFGA060001) （2012GXNSFGA060001）
