Robust photometric stereo in a scattering medium via low-rank matrix completion and recovery

Hui Yu, Hao Fan, Yisong Luo, Junyu Dong

Research output: Chapter in Book/Report/Conference proceedingConference contribution

252 Downloads (Pure)

Abstract

Photometric stereo is a popular method for its better detail recovery. However, when it is used in a scattering medium such as lake and ocean, the recovery result will be impacted negatively by the absorption, scattering and the impurities in water. In this paper, we present a new method to solve the problem of better 3D reconstruction via Low- Rank Matrix Completion and Recovery . First, we use the dark points like shadows and darkness in water to fit the scattering effect distribution and then remove the scattering from the image. Next, we use robust principal component analysis method (RPCA) to recover the image by removing the sparse noise including shadows, impurities and some corrupted points caused by backscatter compensation. Finally, we combine the RPCA results and least-squares (LS) results to get the normal and accomplish the 3D reconstruction. Extension experimental results demonstrate that our method achieves accurate estimates of surface normal and 3D reconstruction than previous techniques.
Original languageEnglish
Title of host publicationProceedings of the 9th International Conference on Human System Interactions (HSI)
PublisherIEEE
ISBN (Electronic)978-1509017294
ISBN (Print)978-1509017300
DOIs
Publication statusPublished - 4 Aug 2016
Event9th International Conference on Human System Interactions: HSI 2016 - University of Portsmouth, Portsmouth, United Kingdom
Duration: 6 Jul 20168 Jul 2016

Conference

Conference9th International Conference on Human System Interactions
Abbreviated titleHSI 2016
Country/TerritoryUnited Kingdom
CityPortsmouth
Period6/07/168/07/16

Fingerprint

Dive into the research topics of 'Robust photometric stereo in a scattering medium via low-rank matrix completion and recovery'. Together they form a unique fingerprint.

Cite this