Proceedings Article | 14 May 2014
Xiao-peng Wang, Zhi-qiang Yang, Gao-feng Kang, Jun-feng Wang, Mou-shun Jin
KEYWORDS: Image enhancement, Minerals, Remote sensing, Iron, Principal component analysis, Satellites, Satellite imaging, Quartz, Image resolution, Copper
Studding on the remote sensing geological survey in Tashkurgan area of west Kunlun Metallogenic belt using the latest
Worldview 2 high resolution satellite image, using Optimum index factor (OIF) select the band combination suitable for
this area to do strata and Lithology interpretation is B8, B4 and B3, and test different image enhancement method for
mineralization alteration information such as band ratio, principal component analysis (PCA). Carried out lithology,
geological structure and mineralization belt and ore body interpretation on the basis of remote sensing data after
processing. The results show that the ratio band combination can identify multiple sets of diorite, marble, schist and the
lithological boundaries between them clearly; the principal component transform method can enhance the boundary
between black biotite-quartz schist and white granite, meanwhile it can clearly reflect the schist by a different hue and
brightness level due to contained different mineral such as quartz, mica, feldspar and others, iron mineralized belt is also
exposed very well. Spectrum measurement has been done for the rock and mineral in test area. Lithology inversion and
mineralization anomaly information extraction test have been carried out afterwards. The test result proved that the
single mineral composition rock such as marble is suitable for spectral inversion. The principal component transform of
bands B1, B4, B8, and B6 is used to extract iron alteration from worldView2 data, the result shows that PC3 is the main
component containing iron alteration abnormal information. Compared with the abnormalities extracted from
worldview2 and low resolution satellite image such as ETM , Aster, we found that they can only distinguish wide range
distributed mineralizing alteration information, their identification accuracy is not as good as Worldview2. WorldView2
data contained more abundant information and has higher resolution, it not only able to identify a wide range of
alteration abnormal, but also specifically identified to the ore body outcrop, abnormal extraction accuracy is very high
and the effect very well. The study shows that use WorldView2 data carry out geological and mineral remote sensing
survey, have advantages of both high ground resolution of optical characteristics and a certain spectral recognition
capability. The method has good result in structure and mineralized alteration zone recognition and worth promoting.
