A Comparative Analysis of Fusion Methods for MODIS 250m Images

摘要

The MODIS images have been useful for a wide range of land, ocean and atmosphere phenomena. However, lkm resolution would not be enough to detect land changes for small areas comprised of various land cover types. For this study, wavelet-based fusion and adaptive regression model were applied for improvement MODIS five 500m bands, and outputs are compared by some evaluation indicators with visual interpretation. Wavelet-based methods have vertical, horizontal and diagonal emphasized pixels which are created in the process of reconstruction. Fusion method using adaptive regression model including NDVI have a high spectral fidelity.