Study On Reflected And Hyperspectral Mixed-pixel Character Of Leaves And Soil

"Remote Sensing", that is, the "Remote sense". It is through the remote sensor "distant"acquisition target data, and through the analysis of the data to obtain relevant features targetor region or the phenomenon of information science and technology. But remote sensing as anew comprehensive detection technology of science, the development is less than forty years,and Hyperspectral remote sensing is more young, still less than twenty years of history. Butbecause it based on modern physics, electronic computer technology, mathematics methodand the Geology regularity, developing rapidly, has set up a file in the geography, geology,ecology, environmental science, atmospheric science and oceanography fields and widelyused in these fields. Hyperspectral remote sensing refers to the use of a lot of very narrowbands of electromagnetic waves from an interest object to obtain the relevant data, itcombines traditional remote sensing imaging technology and physical organic spectroscopy.However, because of the limited spatial resolution as well as the complexity and diversity ofthe target objects, the acquired remote sensing image in the prevalence of mixed pixels,especially in the more complex areas of the surface features distribution. Because of themixed pixel, it has direct impact on the feature identification and the interpretation and at thesame time affect the accuracy of disctricts classification. Therefore, the mixed pixel problemhas became an obctacles in quantitative remote sensing application, and effectively identifyand interpret mixed pixel is the key to the application of hyperspectral remote sensing. Atpresent, as to the mixed pixel the international has conducted long-term research, and alsoproposed a series of theories and methods to decompose mixed pixels, and these methodshave applied to different kinds of mixed pixels. Based on the current research about the mixedpixel, there are few about pixel mixed with leaves and soil. But this kind of research has greatsignificant in interpret and remote sensing images of forest and forest fire disaster mitigation.For these reasons this article from the different soil cover, different detection angles as well asdifferent types of dry leaves and soil mixed pixel three aspects reflective hyperspectral firstqualitative research and then design the positive cross-experiment, dead leaves and soil mixedpixel from the three aspects of the detection angle, polarization state, and leaves types ofqualitative analysis and research, and finally come to different influencing factors based onorthogonal experiment data analysis of dead leaves and the impacts of soil mixed pixel size.