Study On The Hyperspectral And Polarized Characteristics Of The Vegetation And Soil Mixed Pixel

Remote sensing, perceiving an object indirectly, is a comprehensive applicationtechnology which is growing rapidly within the past forty to fifty years. It greatly enhancesdeveloping, mapping and monitoring capabilities of the human on the region and a globalscale. The history of remote sensing is not long, so it is still a new comprehensive explorationscience and technology. Hyperspectral remote sensing is even younger, which is based on themeasurement genealogy, which uses lots of very narrow band of electromagnetic waves tocollect the data from the target objects. Because it is built on modern physics, computerscience, mathematics theories and methods, and earth science law, hyperspectral remotesensing has been rapidly developed and widely utilized in geography, geology, environmentalscience, ecology, oceanography and atmospheric science and many other realms.Identification of surface types is mostly based on the feature spectrum. Because of thespatial resolution limitations of the sensor and the surface complex diversity, mixed pixels areprevalent in remote sensing image especially in the ground surface where the distribution offeatures are relatively complex. If you put this kind of mixed pixel in any pure kind simply, itwill surely bring the classification error, which may result in lower classification accuracy andinduce ineffective reflection of the true surface coverage. Mixed pixel classification of remotesensing has always been a hot and difficult point. Under the hyperspectral situation, sincethere is a significant increase in spectral resolution, the obtained emission values of pixelsconstitute nearly continuous spectral curves, which represent the target responsecharacteristics. Hyperspectral remote sensing contains the data of dozens or even hundreds ofbands, which provides a unique advantage for the deep study of spectral characteristics ofmixed pixel. Therefore, Hyperspectral remote sensing has been entirely possible to determinethe feature of mixed pixel in its composition and proportion.This paper considers the effects of detection zenith angle, vegetation characteristics andsoil properties on spectral characteristics of vegetation and soil mixed pixel. To quantitativelyreveal the influence of the factors on vegetation and soil mixed pixel size, and also to takeinto account the possible interaction effects, the author designed an orthogonal experimentalwith5different factors, each of which had2levels. Then the author used SPSS19.0statisticalsoftware to test the results and do the analysis. The following conclusion had been drawn:Under different experimental conditions, vegetation and soil mixed pixel spectral reflectancehas significant differences, but the differences vary in degrees. The soil type, vegetation typeand detection zenith angle have very significant effects. Under the same experimentalconditions, different bands are suitable to distinguish the vegetation and soil features.Interaction statistical analysis results showed that the impacts of interaction between thefactors are not apparent.