Aiming at the effect of noise on image resolution, a denoising algorithm based on two-dimensional variational modal decomposition (2D-VMD) and fractional integration is proposed. First, the image signal is decomposed into a number of intrinsic modal components (IMF) with different center frequencies by 2D-VMD, effective low-frequency IMF components are screened, thresholds are set according to the difference in image information, and fractional integral is adaptively selected. Convolution operation is performed on an effective component image, and the noise is filtered by the directional mask denoising operator according to the order of integration, and the image denoising is finally completed. The experimental results show that the objective evaluation parameter values have been improved, and this method can better maintain the details of the image such as contours or textures while filtering out the noise.