Font Size: a A A

Evaluation Of Simulated HyMap Data Quality And Mineral Mapping Based On The Simulated Hyperspectral Image

Posted on:2012-02-12Degree:MasterType:Thesis
Country:ChinaCandidate:H Y HuoFull Text:PDF
GTID:2120330332989256Subject:Cartography and Geographic Information Engineering
Abstract/Summary:PDF Full Text Request
there are several evaluation parameters whose principles are also different fromeach other, so their funtions are also different.Generally speaking,we usually use theparameters such as SNR,Histogram,Entropy and so on to evaluate the quality ofhyperspcetral image.This article focuses on analysis of the noise which is one of themost important factor that has great impact on the simulated hyperspectral data.Thenoise can be defined as a Noise in theory can be defined as unpredictable randomerror which can only be understand by statistical methods.So,The noise of thehyperspectral image can be seen as a multidimensional random process,and we candescribe the noise completely using random process and its function of probabilitydistribution and the function of probability density.However, in many cases, thisdescription method is very complicated or even impossible, and often unnecessary inpractice.Usually,we use the features and characters of noise to describe it during thepreprocess,namely,mean, variance and correlation function.The quality of simulated hyperspectral data with different payload indexes isevaluated according two aspects,namely,pre–processing and geoscience applicationsof the hyperspectral image data in this paper. In the aspect of simulated datapre–processing,three parameters were used to evaluate the quality of hyperspectralimage data,which are average variance abnormality,histogram abnormality andcorrelation abnormality. In the aspect of geoscience application of the simulatedhyspectral data,the simulated hyspectral image data were used to extract the alterationinformation of mineralization and mineral mapping,and the quality of simulatedhyperspectral data with different payload indexes,or with same payload indexes butdifferent scales,was analyzed and evaluated according to the number and kinds of theinformation extracted and the precision of mineral mapping.The results are very well.For the study of methodes of Mineral mapping in Hami Area which are based onsimulated HyMap image,this article will do it in three aspects and will focus onanalysis the method of mineral mapping based on SAI.First,for the study of method of mineral mapping based on MTMF and SAM,because of the height of orbit,the data collected by HyMap are influenced seriously by atmosphere.Also there is serious Swath Error and noise information,strong absorbing bands of water vapor and so on contended in Hyperion data.It is difficult to rebuild the spectral and extract the useful information.Using Fourier Transform,then defined the Filtering Operators which can be used for remove the Swath Error quickly and exactly.And then with the help of MNF transform,i extract the purified pixels information after the PPI process.At last,i use the method of spectral fitting,combining the method of spectral matching to do mineral mapping.Second,for the study of method of mineral mapping based on SAI,i use the SAI model (proposed by Wang Jinnian) which is defined as the reciprocal of the ratio of reflection intensity of the bottom and the corresponding position at the "Non-absorption of the baseline (two shoulder band connection) " .In this paper,we used Interactive Data Language to develop a software module,combining other features of altered minerals such as the Position of the absorption bottom,Depth,Width,Area of absorption,Symmetry and so on,to extract alteration information of minerals and do mineral mapping.According to the mineral mapping results,i analysis it and bring about some suggestions to improve the SAI software module.Third , for the study of mineral mapping based on Laboratory spectral library(measured by Portable Near Infrared Spectrometer for Mineral Analysis,PIMA ),I used the PIMA equipment to measure the spectrals of collected field samples.According to spectral characteristics of minerals,we used the PIMA to identify them and do semi-quantitative analysis.And then I established a spectral library using the measured spectrum,combining the SAM,to do mineral mapping.According to the geological data and the fieldwork,I compared and analized the above three methods of mineral mapping,summarized the advantages and disadvantages of them,brought about some advice to improve the SAI module.
Keywords/Search Tags:Simulated HyMap Data, Evaluation of Image Quality, Payload Indexes, Information Extraction, Mineral Mapping
PDF Full Text Request
Related items