用混合像元线性模型提取中等植被覆盖区的粘土蚀变信息  

Extracting Clay Alteration Information of Medium Vegetation Covered Areas Based on Linear Model of Spectral Mixture Analysis

作　　者：刘成[1] 王丹丽[2] 李笑梅[3] 

机构地区：[1]沈阳航空工业学院信息经济系,辽宁沈阳110034 [2]东北大学资源与土木工程学院,辽宁沈阳110004 [3]沈阳航空工业学院中心,辽宁沈阳110034

出　　处：《遥感技术与应用》2003年第2期95-98,共4页Remote Sensing Technology and Application

摘　　要：基于混合像元线性分解模型,针对中等植被覆盖区提出了一种提取粘土蚀变信息的新方法。主要分3步实现:用混合像元线性分解模型提取植被覆盖丰度;对线性模型进行完善,并依此重构不含有植被信息的新的多波段图像;利用TM5/TM7增强粘土蚀变信息。经验证,提取的植被信息以及粘土类蚀变信息与实际吻合较好,与基于比值-主成分分析的方法相比有明显的优越性。A new method for extracting clay\|alteration information from ETM images was applied in three main steps. First, based on the linear unmixing model of MSA, this paper managed to get the vegetation abundance of the study area; second, making some supplements to the basic linear model of MSA, a new multi\|band data set was reconstructed containing no vegetation information; third, get the clay\|alteration information by making TM5/TM7 ratio image in which the shadow information was dramatically compressed. Field survey and other information have shown that the practice of this paper is right and can be adopted to extract thematic information for such areas as with medium vegetation cover. Compared with the methods mainly based on PCA^ratio combination, this method is more effective and offers a more universal frame to follow to extract clay\|alteration information in medium vegetation areas.

关 键 词：粘土蚀变信息 中等植被覆盖区 混合像元线性模型 遥感 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]