High Dynamic Range and Super-Resolution from Raw Image Bursts

TL;DR

This paper presents a novel method for reconstructing high-resolution, high-dynamic-range color images from raw image bursts captured by handheld cameras, combining physical modeling, optimization, and learned representations for superior results.

Contribution

It introduces the first approach that integrates physical image formation models with learned alignment and priors for high-quality HDR and super-resolution from raw bursts.

Findings

Achieves super-resolution up to 4x on real photographs.

Demonstrates robustness to low-light, noise, and motion artifacts.

Fast and memory-efficient compared to existing learning-based methods.

Abstract

Photographs captured by smartphones and mid-range cameras have limited spatial resolution and dynamic range, with noisy response in underexposed regions and color artefacts in saturated areas. This paper introduces the first approach (to the best of our knowledge) to the reconstruction of high-resolution, high-dynamic range color images from raw photographic bursts captured by a handheld camera with exposure bracketing. This method uses a physically-accurate model of image formation to combine an iterative optimization algorithm for solving the corresponding inverse problem with a learned image representation for robust alignment and a learned natural image prior. The proposed algorithm is fast, with low memory requirements compared to state-of-the-art learning-based approaches to image restoration, and features that are learned end to end from synthetic yet realistic data. Extensive experiments demonstrate its excellent performance with super-resolution factors of up to $\times 4$ on real photographs taken in the wild with hand-held cameras, and high robustness to low-light conditions, noise, camera shake, and moderate object motion.