In order to make the fused multispectral (Ms) images keep the spectrum of the original multispectral images as much as possible and improve the spatial resolution effectively, a novel algorithm for remote sensing image fusion based on the nonsubsampled contourlet transform (NSCT) is proposed. Firstly, the panchromatic (Pan) image is decomposed by the NSCT to obtain the detail information of the Pan image. Secondly, the low frequency subband coefficients of the NSCT are substituted by the band to be fused of the resampled multispectral image and the bandpass directional coefficients of the NSCT are locally adjusted by employing an injection model, in which some physical characteristics of imaging systems are included, and then the NSCT coefficients of the fused multispectral image are attained. Lastly, the fused multispectral image with high spatial resolution is obtained by performing the inverse NSCT on the attained coefficients. Several sets of IKONOS images have been merged by the proposed algorithm. The experimental results demonstrate that the proposed approach outperforms the other traditional methods in the spectral information preservation and in the spatial resolution improvement.