Volume 9 - Issue 23
Wavelet domain inpainting algorithm based on low-rank matrix recovery
Abstract
Wavelet domain inpainting (WDI) has attracted some attention, but most of WDI methods are based on total variation (TV), which are usually time-consuming because of wavelet transforms during iterations. This paper put forward a new WDI algorithm base on low-rank matrix recovery that directly restores corrupt wavelet coefficients by minimizing the ranks of the wavelet coefficients directly. It doesn't need wavelet transforms during iterations and so is much faster. Experimental results on different images are presented to show the efficiency and stability of the proposed algorithm.
Paper Details
PaperID: 84892844949
Author's Name: Zheng, C.
Volume: Volume 9
Issues: Issue 23
Keywords: Augmented lagrangian method, Low-rank matrix recovery, Rank minimization, Wavelet domain inpainting
Year: 2013
Month: December
Pages: 9465-9472