Image Segmentation Using Fuzzy Region Competition and Spatial/Frequency Information

Abstract

This paper presents a multiphase fuzzy region competition model that takes into account spatial and frequency information for image segmentation. In the proposed energy functional, each region is represented by a fuzzy membership function and a data fidelity term that measures the conformity of spatial and frequency data within each region to (generalized) gaussian densities whose parameters are determined jointly with the segmentation process. Compared with the classical region competition model, our approach gives soft segmentation results via the fuzzy membership functions, and moreover, the use of frequency data provides additional region information that can improve the overall segmentation result. To efficiently solve the minimization of the energy functional, we adopt an alternate minimization procedure and make use of Chambolle's fast duality projection algorithm. We apply the proposed method to synthetic and natural textures as well as real-world natural images. Experimental results show that our proposed method has very promising segmentation performance compared with the current state-of-the-art approaches.

A Comprehensive Texture Segmentation Framework for Segmentation of Capillary Non-Perfusion Regions in Fundus Fluorescein Angiograms

“…According to [24], [25], a general multiphase segmentation model for both texture and non-texture segmentation problems can be formulated to minimize the following energy,The first term concerns the smoothness of region boundaries. The second term is known as the region term which enforces the similarity measurement between different regions by and can vary depending on the problem domain and will be discussed in detail later this section.…”

“…Notably, other variants of region term definitions (i.e. [24]) can also be adopted by this model, reflecting the flexibility of the proposed multiphase model.…”

“…The former normally involve feature selection and/or extraction, along with acquisition of a training dataset to generate and train the classifier to perform segmentation on new unseen images [17]. Unsupervised segmentation can be achieved by different approaches [18], [19], [20], [21], [22], [23], [24], [25]. In spite of significant efforts unsupervised strategies have not yet been successfully employed for real world applications, since it is difficult to be application specific without considering domain knowledge.…”

Image segmentation via image decomposition and fuzzy region competition

“…The goal of image segmentation is to partition an image into a set of disjoint sub-regions with uniform and homogeneous attributes. Many approaches have been proposed in recent years, among which variational method [1][2][3][4][5][6][7][8] is one of the most successful methods. In the variational framework, the level set (LS) method [2,3] and the fuzzy region competition (FRC) method [4][5][6][7][8] are widely used.…”

“…Based on a nonconvex regularizer, Han et al [7] propose a variational model for the fuzzy multiphase image segmentation. Choy et al [8] present a multiphase fuzzy image segmentation model by taking into account both spatial and frequency information in data fidelity term. These works have shown very promising results.…”