The evaluation of performances of fusion methods is a key problem in remote sensing image fusion. In this paper, four
representative fusion methods, PCA fusion, WT fusion, CT fusion and TLS-GIF-WC, are adopted to fuse two sets of
ALI images for comparison. The fusion products are applied to two remote sensing applications, vegetation index
extraction and image classification. The normalized difference vegetation index (NDVI), vegetation coverage and
classification accuracy indices are adopted to compare the fusion products. Experiments show that the GIF fusion
products are more adaptive for vegetation application, since the NDVI and vegetation coverage extracted from the fusion
product are consistent with that extracted from the initial image, and the ARSIS concept fusion and TLS-GIF-WC
products are more adaptive for image classification, because of the higher classification accuracy.