福州大学学报（自然科学版）2011，Vol.39Issue(1)：60-66,7.DOI:CNKI:35-1117/N.20110121.1724.013
一种基于高斯马尔可夫随机场模型的混合像元分解方法
A method of spectral mixture analysis based on Gaussian Markov random field model
摘要
Abstract
Traditional spectral mixture techniques only considered spectral information, and usually ignored the efficient utilization of spatial dependent information. In fact, the phenomenon of spatial dependence can be observed in beth remotely sensed images and unmixed abundance images. The spatial dependence was depicted by Ganssian Markov Random Field (GMRF) model in this paper. A hybrid model integrated image spectral with abundance spatial dependent information was established to improve the accuracy of spectral mixture analysis. Simulated and real remote sensing images were used to validate the present method, and experiments show that this method can significantly improve abundance estimation, especially in noise image.关键词
遥感/混合像元分解/GMRF模型/组分空间相关Key words
remote sensing/ spectral mixture analysis/ GMRF model/ abundance spatial dependence分类
信息技术与安全科学引用本文复制引用
詹锡兰,吴波..一种基于高斯马尔可夫随机场模型的混合像元分解方法[J].福州大学学报（自然科学版）,2011,39(1):60-66,7.基金项目
国家自然科学基金资助项目(4080118) （4080118）
福建省自然科学基金资助项目(2010J01251) （2010J01251）
