A Comparative Study On The Application Of The Interpretation Effects Of Current Land Use With CBERS And TM Image Data

Satellite remote sensing technology have been ipmrove rapidly with the development of human space technology and the of computer technology since the 1970s.Since the late 1990s,the multi temporal,multi-spectral remote sensing image data for the same area become more and more,Multi-sensor data fusion technology as one of the main branch of a data fusion technology have been importance attached by academics at home and abroad.CBERS image data as the images with independent intellectual property rights in China have continuously in-depth application in various field.So,it is inevitable related to the integration of data with other images.In land resource investigation field,the CBERS image data is commonly used with TM,SPOT and other image data.The CBERS multispectral image have the same performance parameters with TM image in visible band.If we do a multi-spectral image fusion with them,the interrelated spatial information will be well enhanced,and the spectral information of multi-spectral images wii be effectively kept.This paper use CBERS images and TM remote sensing image data fusion as an example,and analysis the traditional methods of Image fusion,and intergate the images with the principal component transform,IHS transform and other methods. This paper introducte a the improved methods by Zhangwei,then this paper put forward a new method based on the improved method and wavelat transform.And then this paper use the visual image interpretation and a series indicators of integration evaluation for the fusion results.Using the results,the authors do the supervisory classification,and show the practicality with the classification results.The Research work and main results are listed as follows:1.This paper introducte the level of Multi-source remote sensing method,the normal methods of image fusion on pixel level,and the evaluation Standard.2.This paper introducte the CBERS images and TM remote sensing image data,and introduce the pretreatment process before image fusion3.This paper introducte the normal methods and the fusion process of the CBERS images and TM remote sensing image data fusion,and copmare with the original data by visual interpretation.The definiton and the texture feature is more the the original data,but they have error of spectrums more or less.4.This paper put forward a Fusion method based on IHS transformation and the wavelet transform.The method combined the two transformation,and it has enhanced the spatial detail manifestations ability with preserving the Spectral properties of TM.The results show that the new method is effective.The combined image have all the information of CBERS images and TM image.The authors do the supervisory classification with the integrated data,and compare the results with the original classfication data.It show that the new method is close to the actual situation.So this paper holds that this method is available for the land use classfication with CBERS image data.