UnModNet: Learning to Unwrap a Modulo Image for High Dynamic Range Imaging
Abstract
A conventional camera often suffers from over- or under-exposure when recording a real-world scene with a very high dynamic range (HDR). In contrast, a modulo camera with a Markov random field (MRF) based unwrapping algorithm can theoretically accomplish unbounded dynamic range but shows degenerate performances when there are modulus-intensity ambiguity, strong local contrast, and color misalignment. In this paper, we reformulate the modulo image unwrapping problem into a series of binary labeling problems and propose a modulo edge-aware model, named as UnModNet, to iteratively estimate the binary rollover masks of the modulo image for unwrapping. Experimental results show that our approach can generate 12-bit HDR images from 8-bit modulo images reliably, and runs much faster than the previous MRF-based algorithm thanks to the GPU acceleration.
Cite
Text
Zhou et al. "UnModNet: Learning to Unwrap a Modulo Image for High Dynamic Range Imaging." Neural Information Processing Systems, 2020.Markdown
[Zhou et al. "UnModNet: Learning to Unwrap a Modulo Image for High Dynamic Range Imaging." Neural Information Processing Systems, 2020.](https://mlanthology.org/neurips/2020/zhou2020neurips-unmodnet/)BibTeX
@inproceedings{zhou2020neurips-unmodnet,
title = {{UnModNet: Learning to Unwrap a Modulo Image for High Dynamic Range Imaging}},
author = {Zhou, Chu and Zhao, Hang and Han, Jin and Xu, Chang and Xu, Chao and Huang, Tiejun and Shi, Boxin},
booktitle = {Neural Information Processing Systems},
year = {2020},
url = {https://mlanthology.org/neurips/2020/zhou2020neurips-unmodnet/}
}