A Bhattacharyya Coefficient-Based Framework for Noise Model-Aware Random Walker Image Segmentation

TL;DR

This paper introduces a noise model-aware framework for the random walker image segmentation algorithm that derives weight functions based on probabilistic modeling, eliminating the need for parameter tuning and improving performance across various noise types and datasets.

Contribution

The authors propose a novel probabilistic framework for deriving weight functions in random walker segmentation, adaptable to any noise model and removing the critical parameter dependence.

Findings

Superior performance on synthetic data

Effective segmentation of biomedical images

Framework applicable to other graph-based methods

Abstract

One well established method of interactive image segmentation is the random walker algorithm. Considerable research on this family of segmentation methods has been continuously conducted in recent years with numerous applications. These methods are common in using a simple Gaussian weight function which depends on a parameter that strongly influences the segmentation performance. In this work we propose a general framework of deriving weight functions based on probabilistic modeling. This framework can be concretized to cope with virtually any well-defined noise model. It eliminates the critical parameter and thus avoids time-consuming parameter search. We derive the specific weight functions for common noise types and show their superior performance on synthetic data as well as different biomedical image data (MRI images from the NYU fastMRI dataset, larvae images acquired with the FIM technique). Our framework can also be used in multiple other applications, e.g., the graph cut algorithm and its extensions.