Study On Simulating Near Natural Color Of Multi-Spectrum Remote Sensing Image

We discussed the method of simulating natural color on the base of ETM+ remote sensing image in this paper.In order to illuminate the applicability and limitation of evry wave band, we analysed the spectrum character of many objects on the earth and the spectrum character of ETM+ image'sbands at first. Because of the redundancy of evry wave band in containing information, wemust analyse the six wave bands in statistic method to eliminate information redundancy andfind out the best wave bands combination was TM5(R), 4(G), 3(B). Furthermore, this thesis demonstrated the principle of true color combination and the theory of false color combination, and explained the essential of simulating natural color. We also illuminate the feasibility of simulating natural color.By using many kinds of software such as Arc View, Arc/Info, ERDAS Imagine, ENVI to draw cartography and process remote sensing image, we could find a method which satisfys the theory request and our vision request. This study solved not only the matter of vision but also the matter of quantity of information that remote sensing image contains. Satisfying both of the two aspects in reason challenges us in this paper.To get our purpose, subseting and Geomitric Crecction were absolutely necessary. Geometric Correction could bring different spatial data into earth reference frame and elimate geometry aberrantion. We can rectify remote sensing image by this means. Because of high resolution precision of panchromatic image, we should resolution merge different type of remote sensing image to enhance the level of image interpretation. After being enhanced in many ways, ETM+ panchromatic band were merged with TM1, 2, 3, 5, 7 through three different fusion algorithms, they were Principal Component transform,Multiplicative transform, Brovey Transform fusion. And then, we estimated the fusion images in both subjective factors and objective factors to find out the best fusion image which was suitable to our study. The result of the estimation demonstrated that Multiplicative transform was the best of the three fusion algorithms and the fused images had higher spatial resolution while maintaining the basic spectral contents of the original TM images, the visal effect and the accuracy of interpretation had been greatly enhanced.On the basis of Data Preparation, Resolution Merge, and sevral image tranformations we did many experimentations of remote sensing image bands' combination to simulate natural color. And then, we found the best method that is to combine several wave bands which contain more information of vegetation properly to get a special wave band at the base of the best bands' combination ; the next step is to put green on this band, and compose this band whith TM5(R) , TM3(B). We got perfect remote sensing image by this means finaly.Taking practical forest data as reference to estimate the image of natural color simulating, it can be shown clearly that the method was feasible.Using the image of natural colorsimulating, we can make remote sensing image map of forestry. The technology of natural color simulating will have wide prospects in forestry remote sensing.