A fusion method for meteorological cloud images based on compressed sensing (CS) was presented. Aiming at the frequency aliasing of the original contourlet transform, the aliasing-free pyramidal filter banks (AFPFB) was combined with directional filer banks (DFB) to construct a new transform, called aliasing-free contourlet transform (AFCT). Then, AFCT was applied to the sparse representation stage of CS to decompose the cloud image into dense and sparse components. The dense components were fused using conventional approach while the sparse components were fused under the framework of CS via fusing a few linear measurements by solving the two-step iterative shrinkage/threshold reconstruction algorithm which uses two previous estimates to obtain a new one. The experiment results demonstrate that the proposed method outperforms the traditional methods in terms of visual quality and quantitative criterion, and the fusion results is propitious to reveal the comprehensive weather information.