Research On SAR Image And Multispectral Image Fusion Algorithm Based On NSST

Image fusion technology combines images acquired from multiple sensors with specific processing and fusion rules to obtain rich information which can not be obtained from a single sensor.The fusion image will achieve a more accurate and comprehensive expression of the target image.Compared with the source image,the fusion image will have higher resolution and higher extent of information gathering,which will provide well research material for the post-processing of the image,including image detection and target recognition.The fusion algorithm of Synthetic Aperture Radar(SAR)images and Multispectral(MS)images was studied.The main contents include:The fusion algorithm based on NonSubsampled Shearlet Transform(NSST)was studied.The Wavelet Transform(DWT),NonSubsampled Contourlet Transform(NSCT)and NSST are studied respectively.Since DWT and NSCT fails to inherit well the texture features of SAR images and the rich spectral information of multi-spectral images,in this paper a method that combines the independent features of HIS components with fine detail capture capabilities of the NSST was proposed.Experimental result indicates that,NSST in multi-scale transform methods has been got the best performance.The fusion algorithm based on sparse representation and NSST was studied.Although both sparse representation and NSST algorithms can represent the images sparsely,both sides have great differences in the fusion theory and algorithm implementation,which have both advantages and disadvantages and complementary.For this,a fusion algorithm based on sparse representation and NSST was proposed.In the algorithm,low frequency subband images after NSST are fused based on sparse representation.The fusion rule selects the adaptive fusion strategy based on coefficient energy,and the high frequency part adopts the criterion based on the correlation coefficient to fuse.The simulation results show,the algorithm can retain the advantages of a single algorithm,make up for the shortcomings,effectively improve the loss of image information,and improve the contrast,the sharpness and the system stability,it have practical value.