Study On The Ground Modeling Of High-Resolution Spectral Remote Sensing

After decades of development, remote sensing has already changed greatly. It grows fromvisible light to the infrared, microwave light in the technology, develops progressively fromdetermining the nature to ration in theory, gets deeply from resource environment to economy ,society a great deal of fields in using, which shows wide prospect and enormous potentiality ofthe remote sensing technology and using. As spectrum range of the remote sensing device is being widened constantly, the resolutionof space, spectrum and time is raising constantly, Surface feature discerning and informationdraw technology are perfect constantly, Remote sensing application fields expand constantly andration take application level deepen constantly. Remote sensing data obtain from multispectral,hyperspectral to ultraspectral. It excavated information from "coarse" to "meticulous", remotesensing surface feature discern from indirectness to directness. Remote sensing application hasmore and more been determined by the study of the spectrum characteristic of surface feature ofquantitative analysis. The spectrum feature is the basis of remote sensing detection and the premise of remotesensing technology. The author has done two investigations of ground reflective spectrum inYunan province and collected 329 items of object spectrum in total, and there are 277 items inBeiya's gold diggings of Heqing county and 52 items in Pulang's copper diggings in Xianggelilacounty, and all the objects includes rock, vegetation, soil and water. Furthermore, the groundmodeling of high resolution spectrum remote sensing is studied, and the main contents andresults are as follows: 1. After the various influence factors of object spectrum are analyzed, the results show thatwe must use correct survey method, and select reasonable conditions of survey, and consideradequately the effect of these facts for gaining useful spectrum data in the survey of field objectspectrum. These data are the premise of information extraction. 2. For analyzing object spectrum and extracting information, functions of four kinds ofmodels, which includes spectrum distinguishing, spectrum classifying, influence factorsanalyzing and spectrum identifying, are used, and the information analysis system of the groundreflective spectrum's test is developed. The function menu of the system includes almostcomplete manipulation functions and is easily manipulated, and its interfaces are very clear andits result is easily analyzed. 3. The circumstances influencing the surround field object are complicated and there aremany kinds of object in the research area. In order to manage expediently these data, aspectrum-database is built and its management system is also done. The spectrum-databaseprovides enough data for building analysis system, and its management system not onlysystematizes and standardizes the management of the object spectrum data that is surveyed andnoted in field and lab, but also gives an entrance to the data in the internal and overseasopen-door spectrum-database. In the management system there are not only the usual input,modification, delete, find etc. functions, but also spectrum curves and data showing. The tables inthe spectrum-database are mostly built with Access, and its management system is developedwith ADO interface connect technology. 4. In research on the spectrum distinguishing models, in order to solve two problems: thefirst one is the difference of two groups of samples obvious, and the second one is which group anew sample belongs to? We use T test, MW test, Fisher distinguishing, Genetic Algorithm patterndistinguishing etc. methods, and these models' feasibility is analyzed. In the application of thesemodels with practical examples, the first problem is solved with the preceding two models, butthe second problem is solved with the latter two models. The veracity of these models is verified,so these models can be feasibly used in practic...