Font Size: a A A

Study On Similar Mineral Identification And Uncertainty Of Mineral Mapping Of Imaging Spectrometry

Posted on:2007-08-25Degree:DoctorType:Dissertation
Country:ChinaCandidate:P K TangFull Text:PDF
GTID:1100360182482671Subject:Cartography and Geographic Information Engineering
Abstract/Summary:PDF Full Text Request
With continuously imaging and hyper-spectral resolution for targets, the RemoteSensing technology of Imaging Spectrometry plays an important role in geology,atmospheric science, ecology, hydrology and ocean. Especially, it has succeeded inminerals identification and minerals mapping of geology. However, it can not begotten a better application for minerals that the properties of physics, chemistry andspectral character are quite similar. In addition, it is uncertain because of beingaffected by a lot of factors. In course of its engineering application, the change ofspatial resolution and spectral resolution extremely affects and restricts execution ofengineering. Using the Australia's HyMap hyper-spectral data, the paper expounds theapplication analysis and discusses of identification of similar minerals, uncertaintyand sensitivity of mineral mapping. The main sections are as follows:In chapter one and chapter two, the paper's topic purporting and workingmethods and routes are summarily explained the and summarized, and presentresearch status and theories and methods of minerals identification of the imagingspectrometry remote sensing technology, spectral rebuilding and uncertainty ofaffecting factors are studied.Some questions are discussed in chapter three about how to identify similarminerals with emphasis. Because of similarly physical and chemical attributes andSpectral Characteristics, the similar minerals can not be identified in spectral levelwith single method. The chapter first analyzes precision evaluation via intrinsicminerals mapping in the research area in East Tianshan Mountains in XinjiangProvince. Via data analysis of some profiles in research area,spectral analysis of fieldspectrum of PIMA and rock-mineral identification indoor, using appropriate scale, itwas summarized that the linear relation between mapping intensity and mappingabundance of imaging spectra mineral mapping, and explores the relation between theprecision of mineral mapping and "detection limit" of mineral. Although mappingprecision of other minerals is better,chlorite and epidote, illite and muscovite areconfused and identified falsely. Analyzing some cases of similar minerals, the similar minerals were identifiedwith aimed methods according to imaging spectral data pattern. The contents include:first, there are many causes of mineral mapping confusion, but the major causes areisomorphism, polymorphism and mixture of minerals. Second, because imagingspectrometry data has the characteristics that the data is complex, the relativity ofinter-band is high, and huge memory, it is impossible to analyze it with some methods.Basing of summarizing the past research findings, the pager presents a viewpoint thatcarry on data sub-blocks with "separating" according to mineral spectroscopicCharacteristics, and decreases the relativity of inter-data and increases the workingspeed. After a method is applied to imaging, spectrometry data are separated bymasking it acceptant result, analyzing content is explicated and the affecting ofsurrounding data is removed. Third, aiming at the characteristics and identifyingfactor, minerals are differentiated with based ICA mineral mapping. (Aiming atmixture minerals and similar minerals), ratio analysis (Aiming at similar minerals)and similarity measures analysis methods through appropriate combined mode andapplicable sequence.Chapter four: the uncertainty and sensitivity of affecting factors of imagingspectra mineral mapping are analyzed. Spectrum rebuilding is an important part inimaging spectrometry minerals mapping. Firstly, the influence of atmosphericcorrection and spectral rebuilding methods was discussed. Then, the paper gaveinfluence of minerals mapping which results are got from images that their spatialresolution and spectral resolution are changed. Finally, Multi-linear Statistic Analysisand Discriminant Analysis are used to analyze sensitivity of spatial resolution andspectral resolution for minerals mapping quantificationally through a set of mappingresults of simulating images and a conclusion of effect digree was given.
Keywords/Search Tags:Remote Sensing Technology of Imaging Spectrometry, Similar Mineral Identification, Precision Evaluation, Uncertainty, Sensitivity
PDF Full Text Request
Related items