To address the problems of poor inter-image correlation and obvious differences in fused brightness in high resolution remote sensing image fusion, this paper proposes a method for fusing domestic high-resolution panchromatic and multispectral remote sensing using the non-subsampled shearlet transform. Remote sensing images of GF-1, GF-2, and GF-7 are selected as the experimental data. The intensity-hue-saturation (IHS) algorithm is used to extract the luminance component of the multispectral image; the non-subsampled shearlet transform (NSST) algorithm is used to extract the high frequency and low frequency information from the luminance component and the panchromatic image; and the relationship between high frequency and low frequency is fully considered when designing an effective image fusion strategy. Finally, the fused image is obtained by the IHS and NSST algorithms. By comparing the proposed method to Brovey, Gram-Schmidt (GS)、Hue-saturation-value (HSV), and Co-occurrence filtering (COF) algorithms, it is determined that the proposed method is a feasible remote sensing image fusion method with the combination of subjective and objective evaluation for the fused images.