遥感与化探数据融合技术在金矿预测中的应用

APPLICATION OF FUSION TECHNIQUES OF REMOTE SENSING AND GEOCHEMICAL DATA IN GOLD ORE EXPLORATION

作为一种先进的多元数据融合技术,遥感与化探数据融合技术的目的在于能从原始的遥感图像中获得较高质量的图像和提取实际需要的蚀变信息。文章首先研究了几个典型的遥感与化探融合的算法(如PCA、ISODATA、MLC等),然后提出了一个新的使用了专家系统的基于3个层次(像素级、特征级和决策级)的遥感与化探数据融合方法。该方法通过在招远金矿区应用示范,结果表明这种新的方法较之传统的遥感与化探融合方法是一种行之有效的遥感找矿技术方法。

As one of advanced multi -source remote sensing image fusion techniques, fusion techniques of remote sensing and geochemical data aim at obtaining information of greater quality and extract alteration information from the original images. Several classic algorithms, including Band Ratio, Principal Component Analysis, Iterative Self - Organizing Data Analysis Technique （ ISODATA）, and Maximum Likelihood Classifier （ MLC ） are studied. A novel remote sensing and geochemical data fusion method based on three levels （Pixel Level, Feature Level, and Decision Level） by using expert classifier is proposed. Experiment results in the Zhaoyuan gold district show that the method is remarkably better than the classical fusion methods.