Efficient Superimposition Recovering Algorithm

TL;DR

This paper introduces ESRA, an accelerated gradient-based algorithm for efficiently recovering high-quality transparent layers from superimposed images, with theoretical convergence guarantees.

Contribution

The paper presents a novel, fast superimposition recovery algorithm (ESRA) with a dual approach and parallel constrained TV, improving quality and convergence over existing methods.

Findings

Reconstructs high-quality transparent layers without color bias

Achieves rapid convergence with theoretical guarantees

Employs a dual approach and parallel algorithm for efficiency

Abstract

In this article, we address the issue of recovering latent transparent layers from superimposition images. Here, we assume we have the estimated transformations and extracted gradients of latent layers. To rapidly recover high-quality image layers, we propose an Efficient Superimposition Recovering Algorithm (ESRA) by extending the framework of accelerated gradient method. In addition, a key building block (in each iteration) in our proposed method is the proximal operator calculating. Here we propose to employ a dual approach and present our Parallel Algorithm with Constrained Total Variation (PACTV) method. Our recovering method not only reconstructs high-quality layers without color-bias problem, but also theoretically guarantees good convergence performance.