Remote Sensing Image Fusion Based On Multi-resolution Analysis Of Technical Studies

Remote sensing image fusion is an impotent research field in image fusion,which integrates image information from different remote sensors,and then fuses them by appropriate strategies.Finally it can get a new image which has more complete,objective cognition of the specified target and is adapt to further image processing.A comprehensive research on multi-sensor remote sensing image fusion based on wavelet transform is made in this paper,which mainly includes the following contents:First of all,we introduce the development of remote sensing image fusion and wavelet multiresolution analysis and the application of wavelet transform in remote sensing image processing. On the basis of these analyses,three algorithms are proposed,that is,a new regional remote image fusion method based-on wavelet transform,an improved image fusion method based on(?)trous wavelet transform,and an improved image fusion method based on curvelet transform.The first method:Firstly,the image is processed by wavelet transform.The matching degree and fusion coefficient can be gained by regional energy of decomposed high frequency sub-image.Two thresholds are given in order to construct different fusion coefficient according to various features. This method can better retain the spatial features.The second method:Firstly,multispectral image is transformed by IHS transform.Secondly, panchromatic(PAN) image is transformed by(?)trous wavelet transform.Then,features are extracted from the PAN image in order to integrate into the initial intensity component.In order to control the amount of PAN features to be integrated into multispectral(MS) images,an importance measure function is calculated by the ratio of high frequency coefficients and low frequency coefficients,for selecting the significant PAN features according to an adjustable threshold.The results of fusion replace intensity component of multispectral image.The third method:Firstly,the PAN image is enhanced in order to increase the edge information. Then,the enhanced image and MS image are fusioned based on curvelet transform and IHS transform. This method can better retain the spatial information and reduce the spectral distortion.At last,the methods for remote sensing image fusion proposed in this paper are tested respectively. Experimental results show that the performance of new schemes are better than other fusion methods in both of spatial and spectral domains.