节点文献
纹理特征在多光谱遥感影像分类中的应用
APPLICATION OF TEXTURE FEATURE IN MULTI-SPECTRAL IMAGES CLASSIFICATION
【摘要】 提出了基于灰度共生矩阵的多光谱影像纹理分析的方法,实现了利用k-mean聚类算法对多光谱影像进行分类,比较了各种不同的分类结果。
【Abstract】 On the basis of the analysis of the methods based on single band image texture,this paper presents a method for texture analysis of multi-spectral images based on gray level co-occurrence matrix.We realize the extraction of texture feature,using spectral feature and texture feature. We use k-mean clustering algorithm to get the classification of the multi-spectral images.Finally,the results will be compared and discussed.
【关键词】 多光谱影像;
灰度共生矩阵;
聚类;
特征提取;
【Key words】 multi-spectral images; grey level co-occurrence; clustering; feature extraction;
【Key words】 multi-spectral images; grey level co-occurrence; clustering; feature extraction;
【基金】 国家自然科学基金资助项目(40371079);国家“973计划”资助项目(2006CB701303)
- 【文献出处】 测绘信息与工程 ,Journal of Geomatics , 编辑部邮箱 ,2006年03期
- 【分类号】TP79
- 【被引频次】44
- 【下载频次】696