Directional image decomposition using retargeting pyramid

Retargeting pyramid (RP) is an alternative method for multiscale image decomposition to the well-known Laplacian pyramid (LP). RP can be obtained by replacing the low-pass filtering process in LP with content-aware image resizing (a.k.a. retargeting), which is a developing technique for computer vision researches. Furthermore, we use RP for contourlet-based directional image decomposition. In experimental results, the proposed decomposition outperforms the LP-based contourlet transform for image denoising.