The Application Of Remote Sensing Technique On The Extraction Of The Structural Information

The main focus of remote sensing technology is on the image processing, whichgrowth is relative to the development of computer science. With the deeply and widelyapplication of remote sensing and the increasingly demand on the image processingstandard, the information extraction has therefore become a hot area for research.The use of remote sensing technique in tectonics is a successful area in geosciencesapplications. At the beginning, remote sensing information is extracted mainly by themeans of visual interpretation. But in the recent twenty years, some scholars on remotesensing have developed a lot of methods, such as K-L transformation, principalcomponent, band ratio, artificial neutral net and classification by GIS software.In remote sensing images, the extract of linear structure and circular structureinformation is the main part of the extraction of the geological structural information, aswell as the stress of this thesis. The geological condition of Siziwang United Banner isvery comprehensive, and it holds a large area, thus the traditional filed survey wouldmake the field work much hard, and the visual interpretation will waste people and timefor the gigantic work. Through carefully comparing multi-principal component analysisand some other image processing methods and former experiences, this thesis adopts theways of eight direction code and texture analysis way based on gray-levelco-occurrence matrix to extract the lineaments and circular features in Siziwang UnitedBanner. The goal of this is to help reduce the burden of geological mapping on the scaleof 1:250000 in the Siziwang United Banner. It is proved that these two means can stressthe information on stratum, river and fracture, thus better reflect the lineament andcircular feature in the research area. And with carefully compared with the field surveyresults, we can find that accuracy of both these two means can reach over 85 percent.