Cartoon-texture evolution for two-region image segmentation

TL;DR

This paper introduces a novel two-region image segmentation model that effectively handles images with oscillatory components and noise by utilizing cartoon-texture decomposition and ADMM optimization.

Contribution

The paper proposes a new segmentation model based on cartoon-texture decomposition, improving accuracy for textured and noisy images over previous smooth-image models.

Findings

Effective segmentation of noisy images

Accurate handling of textured images

Proven convergence of the optimization scheme

Abstract

Two-region image segmentation is the process of dividing an image into two regions of interest, i.e., the foreground and the background. To this aim, Chan et al. [Chan, Esedo\=glu, Nikolova, SIAM Journal on Applied Mathematics 66(5), 1632-1648, 2006] designed a model well suited for smooth images. One drawback of this model is that it may produce a bad segmentation when the image contains oscillatory components. Based on a cartoon-texture decomposition of the image to be segmented, we propose a new model that is able to produce an accurate segmentation of images also containing noise or oscillatory information like texture. The novel model leads to a non-smooth constrained optimization problem which we solve by means of the ADMM method. The convergence of the numerical scheme is also proved. Several experiments on smooth, noisy, and textural images show the effectiveness of the proposed model.