Research For Remote Sensing Image Atmospheric Correction Method Based On Ground Surface Spectrum Vector Space

Because of the absorb and scatter cased by atmosphere during the electromagnetic wave(including visible and infrared band) transmitting through the atmosphere, the radiance that was obtained by the space-based sensors was distorted. In normal condition, the radiance got by satellite sensors contains the information of ground surface and atmosphere. The proportion of the radiance from ground surface is only 50% at 0.45μm, and at 0.85μm it is better about 80%. Thus it can be seen, in order to improve the quality of the remote sensing images and expand its application, it is imperative to carry out the atmospheric correction of remote sensing images to eliminate the effect of atmosphere.As an important step of the remote sensing images preprocess, the research of atmospheric correction lasted nearly forty years after it was raised. A lot of models of atmospheric correction were proposed, and there were some successes. But those models also more or less have some drawbacks. It could not completely meet the requirement of the precision to correct images by using those models in the further research and application of remote sensing image. Therefore, it is an urgent problem to be solved to find a method that has substantial physical theory and could adapt most kinds of complicate situations.The aim that is described in above paragraph is the end-result in this paper. The atmospheric correction method based on ground surface spectrum vector space (ACMGSSVS) was come up with in the paper. The full paper is composed of five single chapters. In the first chapter, the situation of civil and foreign research was introduced, and a briefly conclusion was made. Then the aim and significance of the research in the paper were also elaborated. In the second chapter, the methods which describe the reflection propertiy of ground surface and atmospheric radiative transfer equation were introduced, firstly. Then, the method called"atmospheric correction based on ground surface spectrum vector space"that was proposed in the paper was discussed. Its basic theory and solution method were also elaborated. In the third chapter, taken the content of the last two paragraphs, the experiment of remote sensing images was carried out by using the method ACGMSSVS come up with in this paper, with the ETM+(Enhanced Thematic Mapper Plus) images which were got by the sensor boarding the satellite called Landsat7. Compared with the result images corrected by ATCOR2, the images corrected by ACMGSSVS were more limpid, and they were better in vision, too. The relative error was less than 20% in ETM+ band 1, and was less than 10% in ETM+ band 2 and 3. The correction effect of the result images corrected by ACMGSSVS in the plain was the same as the mountainous area, and it was same between the unmixed and the mixed pixel, too. It can also overcome the inhomogeneity of the atmosphere, and obtain a better result. In the fourth chapter, a conclusion about the research carried out in this paper was made. The drawback of the research in this paper and the future of atmospheric correction of remote sensing image were also discussed.