Multiscale Higher Order TV Operators for L1 Regularization
For researchers in signal/image denoising and reconstruction, this work addresses artifacts in ℓ1 regularization, but the improvements are incremental and domain-specific.
The authors show that ℓ1 regularization can produce artifacts inconsistent with ℓ0 sparsity, and propose a multiscale higher order total variation (MHOTV) method that improves over wavelets and classical HOTV in numerical experiments.
In the realm of signal and image denoising and reconstruction, $\ell_1$ regularization techniques have generated a great deal of attention with a multitude of variants. A key component for their success is that under certain assumptions, the solution of minimum $\ell_1$ norm is a good approximation to the solution of minimum $\ell_0$ norm. In this work, we demonstrate that this approximation can result in artifacts that are inconsistent with desired sparsity promoting $\ell_0$ properties, resulting in subpar results in {some} instances. With this as our motivation, we develop a multiscale higher order total variation (MHOTV) approach, which we show is related to the use of multiscale Daubechies wavelets. We also develop the tools necessary for MHOTV computations to be performed efficiently, via operator decomposition and alternatively converting the problem into Fourier space. The relationship of higher order regularization methods with wavelets, which we believe has generally gone unrecognized, is shown to hold in several numerical results, although notable improvements are seen with our approach over both wavelets and classical HOTV.