基于最大噪声分量变换(MNF)和矿物标识的植被区蚀变信息提取

Mapping Mineralogical Alteration Using MNF Transfrom and Mineral in Vegetated Areas

以往常规的遥感找矿的方法是主成分分析和比值方法,在植被发育的区域,利用这些常规方法难以识别矿物蚀变信息。本文提出了一种方法,将已知目标矿物转换为图像像元作为标识谱,利用植被在短波红外是暗目标和最大信噪比MNF变换可以有效的突出信息的前提,可以快速比对提取蚀变信息。以云南中甸地区的普朗-红山斑岩成矿带为例,进行绢英岩化和泥化蚀变信息提取。结果表明,该方法与主成分分析法和比值法结果具有一致性,但具有更好的降噪效果。该方法能很好的抑制植被影响,突出目标矿物,是一种快速有效的提取ASTER遥感影像矿物蚀变信息的方法。

Mapping of altered rocks in areas with heavy vegetation is difficult by using principal component analysis and band ratios to remote sensing data, which have been commonly applied to locate hydmthermal alteration zones related to metallic deposits. We propose a method to effectively map al- tered recks, which is based on that the reflectance of green vegetation is low in the shortwave infrared portion and maximum signal to noise ratio （MNF） transform can reduce the effect caused by the vetetation. This technique turns target mineral spectra into pixels of ASTER images as mineral indicators that provide a potential approach for fast and effective extracting mineral alteration from ASTER muhispectral data. In this study, the method was applied to ASTER bands converiong the SWIR with the objective of mapping phyllic and argillic hydrothermally altered rocks in the Prong-Hongshan copper belt in Yunnan Province. The mineral abundance images produced using this method are less affected by inerference compared with other methods such as the PCA and band ratio method. The result shows that this method can suppress well the influence of vegetation and highlight target ores, and extract alteration information of ores from ASTER remote sensing images fastly and efficiently.