Image Fusion Method based on Non-Subsampled Contourlet Transform

摘要

Considering human visual system andcharacteristics of images, a novel image fusion strategy ispresented for panchromatic high resolution image andmultispectral image in non-subsampled contourlettransform (NSCT) domain. The NSCT can give anasymptotic optimal representation of edges and contours inimage by virtue of its characteristics of good multiresolution,shiftinvariance, and high directionality. An intensitycomponent addition strategy based on NMF algorithm isintroduced into NSCT domain to preserve spatial resolutionand color content. Experiments show that the proposedalgorithm can not only reduce computational complexities,but achieve better performances than other mentionedtechniques both in visual point and statistics compared withthe traditional principle component analysis (PCA) method,intensity-hue-saturation (IHS) transform technique, wavelettransform weighted fusion method, corresponding wavelettransform-based fusion method, and contourlet transformbasedfusion method.