A review upon fast discrete curvelet transforms in satellite image processing

Image Fusion based on the Fourier, M.IHS, PCA & wavelet transform methods displays rich multispectral details but they are poor when considering to the Spatial Details from the inputs we feed to them from the satellites. Although, Wavelets came more accurate in making the Linear Features of the Image but when coming to the nonlinear discontinuities in the image then the Curvelet Transforms have emerged up into the frame that overcame the problem in the feature representation. Moreover, Image fusion is increasingly its applications in different areas like Satellite Imaging, Remote Sensing, Multi focus Imaging. In our project we are proposing a novel fusion method with the Fast Discrete Curvelet Transforms (FDCT) that increases not only the efficiency of the image but also the feature representation of the Image both in the context of the Linear & Non-Linear Representations. For the experimental study we have considered Low Resolution(LR) Multispectral Image and as well as a High Resolution(HR) Panchromatic Image to generate a High Resolution(HR) Multispectral Image, which is quantitatively compared with Wavelet, PCA, HPF, M.IHS and Grams-Schmidt fusion Methods, which has outperformed spatially the other methods and retains rich multispectral details.