摘要
提出了一种基于纹理和光谱信息的高分辨率遥感影像分类方法,阐述了其基本原理,并通过试验的对比和分析,证明了利用光谱特征与纹理特征相结合进行分类比单纯运用光谱特征进行分类效果更好。
A classification method for high resolution remotely sensed images based on texture and spectral information is proposed. Its theory and method are expatiated. Experimental results show that the performance of the presented algorithm is more efficient than the one of only using spectrum feature.
出处
《测绘信息与工程》
2009年第1期16-18,共3页
Journal of Geomatics
基金
国家NSFC创新研究群体科学基金资助项目(40721001)
关键词
影像分类
纹理信息
光谱信息
灰度共生矩阵
image classification
texture Information
spectral information
gray level co-oc-currence matrices
