基于经验模态分解的高分辨率影像融合

摘要

High resolution image fusion is a significant focus in the field of image processing. A new image fusion model is presented based on the characteristic level of empirical mode decomposition (EMD). The intensity hue saturation (IHS) trans- form of the multi-spectral image first gives the intensity image. Thereafter, the 2D EMD in terms of row-column extension of the 1D EMD model is used to decompose the detailed scale image and coarse scale image from the high-resolution band image and the intensity image. Finally, a fused intensity image is obtained by reconstruction with high frequency of the high-resolution image and low frequency of the intensity image and IHS inverse transform result in the fused image. After pre- senting the EMD principle, a multi-scale decomposition and reconstruction algorithm of 2D EMD is defined and a fusion tech- nique scheme is advanced based on EMD. Panchromatic band and multi-spectral band 3,2,1 of Quickbird are used to assess the quality of the fusion algorithm. After selecting the appropriate intrinsic mode function (IMF) for the merger on the basis of EMD analysis on specific row (column) pixel gray value series, the fusion scheme gives a fused image, which is compared with generally used fusion algorithms (wavelet, IHS, Brovey). The objectives of image fusion include enhancing the visibility of the image and improving the spatial resolution and the spectral information of the original images. To assess quality of an image after fusion, information entropy and standard deviation are applied to assess spatial details of the fused images and correlation coefficient, bias index and warping degree for measuring distortion between the original image and fused image in terms of spectral information. For the proposed fusion algorithm, better results are obtained when EMD algorithm is used to perform the fusion experience.