Current browse context:
eess.IV
Change to browse by:
References & Citations
Electrical Engineering and Systems Science > Image and Video Processing
Title: Enhanced total variation minimization for stable image reconstruction
(Submitted on 9 Jan 2022 (v1), last revised 19 Aug 2022 (this version, v2))
Abstract: The total variation (TV) regularization has phenomenally boosted various variational models for image processing tasks. We propose to combine the backward diffusion process in the earlier literature of image enhancement with the TV regularization, and show that the resulting enhanced TV minimization model is particularly effective for reducing the loss of contrast. The main purpose of this paper is to establish stable reconstruction guarantees for the enhanced TV model from noisy subsampled measurements with two sampling strategies, non-adaptive sampling for general linear measurements and variable-density sampling for Fourier measurements. In particular, under some weaker restricted isometry property conditions, the enhanced TV minimization model is shown to have tighter reconstruction error bounds than various TV-based models for the scenario where the level of noise is significant and the amount of measurements is limited. Advantages of the enhanced TV model are also numerically validated by preliminary experiments on the reconstruction of some synthetic, natural, and medical images.
Submission history
From: Hao-Ning Wu [view email][v1] Sun, 9 Jan 2022 10:24:02 GMT (1868kb)
[v2] Fri, 19 Aug 2022 04:26:10 GMT (1745kb)
Link back to: arXiv, form interface, contact.