Saved in:
Bibliographic Details
Main Authors: Wang, Jason, Nguyen, Lucas, Eom, Hyunseung, Xu, Wei, Guo, Qi
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2606.00379
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918532596367360
author Wang, Jason
Nguyen, Lucas
Eom, Hyunseung
Xu, Wei
Guo, Qi
author_facet Wang, Jason
Nguyen, Lucas
Eom, Hyunseung
Xu, Wei
Guo, Qi
contents We present a non-learning stereo framework for disparity estimation from severely noisy images. Using the Field of Junctions (FoJ), it retains coarse visual features stable under severe noise for cost volume construction while discarding fine textures inseparable from photon noise. The resulting structural information guides boundary-aware Semi-Global Matching (SGM) that dynamically adapts smoothness penalties to preserve true disparity discontinuities. The output is a sparse disparity map more accurate than those of recent stereo algorithms over unmasked pixels on widely-used benchmark datasets.
format Preprint
id arxiv_https___arxiv_org_abs_2606_00379
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Non-Learning Low-Light Stereo Vision
Wang, Jason
Nguyen, Lucas
Eom, Hyunseung
Xu, Wei
Guo, Qi
Computer Vision and Pattern Recognition
We present a non-learning stereo framework for disparity estimation from severely noisy images. Using the Field of Junctions (FoJ), it retains coarse visual features stable under severe noise for cost volume construction while discarding fine textures inseparable from photon noise. The resulting structural information guides boundary-aware Semi-Global Matching (SGM) that dynamically adapts smoothness penalties to preserve true disparity discontinuities. The output is a sparse disparity map more accurate than those of recent stereo algorithms over unmasked pixels on widely-used benchmark datasets.
title Non-Learning Low-Light Stereo Vision
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2606.00379