Remote sensing image fusion is an effective method in farm target identification. The purpose of this study is to investigate the optimum method of remote sensing image fusion for the vegetation information extraction of Nanjing. First, the HIS, PCA, Brovey and the wavelet transforms were applied to the merger of ETM+ PAN and the ETM+ multi-spectrum remote sensing image. Then the merged images were quantitatively evaluated according to their standard spectrum quality and spatial structure information based on the information content evaluation, the high resolution information integrate and the classified precision. The experimental results showed that the Brovey transforms is more effective than the other algorithms and could significantly improve the classification accuracy of the fused imagery in the exaction of vegetation information of Nanjing.