A post-processing algorithm of still image compression was proposed. Iterative contourlet-based image interpolation was implemented to revive the highpass information, which was then fused with the highpass information of the decompressed image to further preserve the significant wavelet coefficients. The algorithm helped to reduce compression artifacts and improve regularity along edges of the decompressed image at low bitrates especially. The ability of the proposed method was demonstrated under JPEG2 000 and SPIHT compression systems. Only for once iteration, the PSNR (peak signal-to-noise ratio) of the post-processed image may improve 0.05~0.2 dB than that of the decompressed image. Ringing artifacts decrease with the increase of the number iteration.