Research On Remote Sensing Image Fusion Algorithm Based On NSST Transform

Image fusion refers to a new technology using the image data collected by multi-source channels on the same target through image processing to maximize the extraction of useful information in their respective channels,and finally produce the high-quality images.As an important type of image fusion,remote sensing image fusion has been applied to flood monitoring,topographic mapping,land resources survey,natural disaster monitoring and so on.At present,researchers have extensively studied and proposed many algorithms for the fusion of high spatial resolution PAN images and high spectral resolution MS images.In this dissertation,remote sensing image fusion algorithms based on NSST(non-subsampled shearlet transform)is mainly studied,which include the following three aspects:1.Combining with the HSV(Hue,Saturation,Value)transform can preserve the multispectral information and the NSST transform can better enhance the identification of boundary information and the presentation of texture details,a remote sensing image fusion algorithm based on HSV and NSST transform is proposed.2.The remote sensing image fusion algorithm based on NSCT(non-subsampled contourlet transform)has the problems of high computational complexity.Enlightened by respective advantages of multi-feature and adaptive PCNN(pulse coupled neural network),a multi-feature remote sensing image fusion algorithm based on NSST transform and adaptive PCNN is proposed.3.Combining with the salient features of PAN images can be extracted by the sparse matrix obtained by RPCA(robust principal component analysis)decomposition and NSST transform has high computational efficiency,and introducing PCNN model,a novel remote sensing image fusion algorithm is proposed.