The Recognition And Extraction Of Granite Through Remote Sensing Images In Taibai Area,Shannxi

Remote sensing petrology is one of the most important branches in remote sensing technology. It is based on rock spectroscopy and aims at identify lithology through analyzing remote sensing images. Granite is of huge importance among igneous rocks because the formation and distribution of which has tight relationships with certain geological settings so that indicated information about regional tectonic and mineralization can be inferred. Remote sensing of identifying lithology of granite has long been focused and researched. Large amount of granite develops in Qinling Mountains-Taibai area which is located in the fault belts conjunction part, consequently it is the main tectonical mineralization area. Considering previous geological researches about this area have been done, the efficiency and veracity of extraction can be verified. Hence this area becomes an ideal region to process remote sensing information extraction of granite. Based on ASTER data, combing other remote sensing data, this paper aims to analyze typical granite spectral characteristics and summarize its regularity of absorbing and reflecting in order to find out a suitable remote sensing identify method for target granite. This method is going to offer basic data support for regional geological survey, as well as provide examples of remote sensing information extraction for similar regional granite. We get some conclusions as follows:(1) Granitoids rock samples collected from Taibai generally has not obvious spectral features in the VNIR-SWIR region,and the main spectral features come from the accessory minerals contained in the rocks.The main absorption bands could be found at Fe2+centered at 0.55?m,Fe3+centered at 0.87?m,H2O centered at 1.4?m and 1.9?m and hydroxy centered at 2.2?m and 2.35?m.Types of granites in different regions have different spectral characteristics because of the various kinds of mineral species and abundance,coupled with different background environment and weathering degree.(2) Compared with the 432 optimal bands combination based on the optimal index factor (OIF) calculation,the principal component analysis (PCA) processing of 345 components combination is more effective. The granite rock boundaries present brighter color and distinguished features in the false color composite images.Considering the enhancement of spectral characteristics,different band math component combinations are selected for different rock mass.The selected ratio of 2/1,2/4 and band 4 of OLI image are employed to identify lithological information of Taibai rock mass and ratio of 2/1,band 4 of OLI image and relative absorption-band depth(RBD)8 to Laojunshan rock mass.The granitoids information is obviously easy to identify through the enhancement of band math component combination.(3) Based on the band math component combination images and PCA 345 components combination image, an relatively accurate extraction information is obtained ultimately,which prove the feasibility of granitoids information extraction based on the ASTER and other auxiliary remote sensing images.