Researches In Extractive_technique Of Lithologic Characters Based On Multispectral Images

The use of remote sensing technique in tectonics is a successful domain in geosciences applications. With the deeply and widely application of remote sensing and the increasingly demand on the image processing standard, the information extraction of image has therefore become a hot domain for research. This paper has priority Research on Lithologic classification technology and linear structures informational abstraction incorporation with Insight clearly the remote sensing image of Inner Mongolia region of interest.This paper was primarily conducted to probe into the method and approach which could improve classification accuracy of rocks in the research domain when using Landsat7 ETM+ and IRS-P6 images with remote sensing technique.The principles and technical methods of some popular image fusion models that include Color transform, PCA transform, HPF transform, wavelet transform are first described and analyzed, then on the basis of this some kinds of experiment fit to high resolution satellite remote sensing image(like ETM+, IRS-P6 images and so on) have been put forward. And then by the numbers induction, summarize know clearly mean, standard deviation, correlation coefficient, disagreement index, entropy, cross entropy, warp entropy, co-entropy, average gradient wait impersonality quantification criteria, combine they and with visual interpretation, design to several different kinds of fusion algorithmic effect proceed weigh and comparison.This paper introduce a processing flow suit for the region of interest and the characteristic of remote sensing image processing technics. As a result, it was found that the better classification method was produced by integrating unsupervised classification and supervised classification procedures before image interpretation which base on error matrix of precision appraise.