Polarization Denoising and Demosaicking: Dataset and Baseline Method

2025-09-12

AI Summary

Polarimetric imaging using division-of-focal-plane (DoFP) sensors suffers from coupled polarization demosaicing and denoising, yet lacks realistic benchmark datasets and physically grounded, reproducible methods. Method: We introduce the first real-scenario multi-noise-level polarimetric joint restoration dataset, comprising 40 diverse scenes with full-resolution paired noisy–clean polarization image patches. We propose a two-stage, reproducible framework—denoising first, then demosaicing—that integrates polarization-domain priors with classical signal processing components to ensure computational efficiency and physical consistency. Contribution/Results: Extensive experiments demonstrate that our method significantly outperforms existing single-task and end-to-end alternatives in PSNR and SSIM. It establishes the first standardized evaluation benchmark and practical baseline for polarimetric image joint restoration, enabling systematic assessment and advancement of physically informed polarimetric reconstruction.

Technology Category

• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Computer Vision: Multi-modal Vision
• Machine Learning: Multimodal Learning

Application Category

• Security and Privacy: Data transparency and provenance
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

A division-of-focal-plane (DoFP) polarimeter enables us to acquire images with multiple polarization orientations in one shot and thus it is valuable for many applications using polarimetric information. The image processing pipeline for a DoFP polarimeter entails two crucial tasks: denoising and demosaicking. While polarization demosaicking for a noise-free case has increasingly been studied, the research for the joint task of polarization denoising and demosaicking is scarce due to the lack of a suitable evaluation dataset and a solid baseline method. In this paper, we propose a novel dataset and method for polarization denoising and demosaicking. Our dataset contains 40 real-world scenes and three noise-level conditions, consisting of pairs of noisy mosaic inputs and noise-free full images. Our method takes a denoising-then-demosaicking approach based on well-accepted signal processing components to offer a reproducible method. Experimental results demonstrate that our method exhibits higher image reconstruction performance than other alternative methods, offering a solid baseline.

Problem

• Develops dataset for polarization denoising and demosaicking
• Addresses noise removal in division-of-focal-plane polarimeters
• Provides baseline method for joint denoising-demosaicking task

Innovation

• Real-world dataset with noise-level conditions
• Denoising-then-demosaicking signal processing approach
• Reproducible baseline method for joint tasks