An ICTM-RMSAV Framework for Bias-Field Aware Image Segmentation under Poisson and Multiplicative Noise

2025-11-12

AI Summary

To address the significant degradation in image segmentation performance caused by the superposition of strong multiplicative (Gamma) and Poisson noise with severe intensity inhomogeneity, this paper proposes a variational segmentation model that jointly performs denoising and bias field correction. Methodologically, we first integrate a relaxed modified scalar auxiliary variable (RMSAV) strategy into the iterative convolution thresholding method (ICTM) framework to achieve efficient optimization. We model the noise statistics using the I-divergence, incorporate a gray-level indicator-guided spatially adaptive total variation regularization, and embed a bias-field-aware mechanism to jointly estimate a smooth bias field. Experiments on both synthetic and real-world images demonstrate that the proposed method achieves substantially higher segmentation accuracy and robustness than state-of-the-art approaches under conditions of intense noise and severe illumination nonuniformity.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Calibration & Uncertainty Quantification
• Intelligent Robots: State Estimation

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

Image segmentation is a core task in image processing, yet many methods degrade when images are heavily corrupted by noise and exhibit intensity inhomogeneity. Within the iterative-convolution thresholding method (ICTM) framework, we propose a variational segmentation model that integrates denoising terms. Specifically, the denoising component consists of an I-divergence term and an adaptive total-variation (TV) regularizer, making the model well suited to images contaminated by Gamma--distributed multiplicative noise and Poisson noise. A spatially adaptive weight derived from a gray-level indicator guides diffusion differently across regions of varying intensity. To further address intensity inhomogeneity, we estimate a smoothly varying bias field, which improves segmentation accuracy. Regions are represented by characteristic functions, with contour length encoded accordingly. For efficient optimization, we couple ICTM with a relaxed modified scalar auxiliary variable (RMSAV) scheme. Extensive experiments on synthetic and real-world images with intensity inhomogeneity and diverse noise types show that the proposed model achieves superior accuracy and robustness compared with competing approaches.

Problem

• Segmenting images corrupted by multiplicative and Poisson noise
• Addressing intensity inhomogeneity through bias field estimation
• Developing robust segmentation for noisy images with varying intensities

Innovation

• Integrates denoising with I-divergence and adaptive TV
• Estimates smoothly varying bias field for inhomogeneity
• Combines ICTM with relaxed scalar auxiliary variable optimization