Dual Exposure Stereo for Extended Dynamic Range 3D Imaging

TL;DR

This paper presents a dual-exposure stereo imaging system that extends dynamic range for accurate 3D depth estimation in challenging lighting conditions, using automatic exposure control and a novel stereo network.

Contribution

It introduces an automatic dual-exposure control method and a motion-aware stereo network for extended dynamic range 3D imaging, improving robustness over existing methods.

Findings

Outperforms other exposure control techniques

Effective in diverse illumination conditions

Validated on real and synthetic datasets

Abstract

Achieving robust stereo 3D imaging under diverse illumination conditions is an important however challenging task, due to the limited dynamic ranges (DRs) of cameras, which are significantly smaller than real world DR. As a result, the accuracy of existing stereo depth estimation methods is often compromised by under- or over-exposed images. Here, we introduce dual-exposure stereo for extended dynamic range 3D imaging. We develop automatic dual-exposure control method that adjusts the dual exposures, diverging them when the scene DR exceeds the camera DR, thereby providing information about broader DR. From the captured dual-exposure stereo images, we estimate depth using motion-aware dual-exposure stereo network. To validate our method, we develop a robot-vision system, collect stereo video datasets, and generate a synthetic dataset. Our method outperforms other exposure control methods.