遥感影像土地利用/覆盖分类方法研究综述

从六个方面总结了国内外出现的遥感影像土地利用／覆盖分类中针对传统计算机分类方法的改进：(1)从基于统计的分类向基于非线性并行处理的人工神经网络分类、基于模糊理论为分类、基于知识的分类以硬支撑向量机等分类技术发展；(2)分类从单一利用光谱信息到利用光谱、纹理、时相、角度等多砷信息；(3)从基于像元的逐点分类到基于图斑的分类；(4)从硬分类到亚像元分类；(5)从单源遥感影像分类到利用多源遥感影像融合的分类；(6)从单分类器向复合分类器发展。

Classification Methods of Land Use/Cover Based on Remote Sensing Technology

Remote sensing technology is characterized by it's ability to obtain terrestrial information that is macroscopical?real-time?dynamic and periodic. It is an efficient way to investigate land use/cover, in which remote sensing image classification is an important step. Since the first resource satellite was launched in 1972, the remote sensing community has witnessed the impressive progress in image classification methods, which is primarily driven by the advancement of remote sensing technology and computer technology. This article described and summaringed new methods that aimed at improving the conventional computer classifiers, which occurred in the land use/cover classification. It follows as this: (1) form statistic-based classification to artificial neural network classification, fuzzy classification, knowledge-based classification, support vector machine classification and so on; (2) from utilizing spectral signature to utilizing texture?phase?angle information; (3) Form per-pixel classification to per-field classification; (4) form crisp classification to sub-pixel classification; (5) form single image classification to image fusion classification; (6) from single to hybrid classifiers.