Augmenting depth estimation from deep convolutional neural network using multi-spectral photometric stereo

Yisong Luo, Hengchao Jiao, Lin Qi, Junyu Dong, Shu Zhang, Hui Yu

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

100 Downloads (Pure)

Abstract

Multi-Spectral Photometric Stereo can recover surface normals from a single image, but requires an initial estimate of the normals due to the tangle of the illumination, reflectance and camera responses on each of the RGB channels. Instead of employing a depth sensor or binocular stereo device, in this paper, we propose a method to estimate fine-scale geometry structures with the popular Deep Convolutional Neural Networks (CNNs). We train the network with rendered images of synthetic 3D objects, and apply the trained model with real world data. The CNN is used to estimate a rough prediction of the depth, then the normals from Multi-Spectral Photometric Stereo are progressively refined accordingly. Experiments demonstrate the competitive results of our method for improving the depth estimation.
Original languageEnglish
Title of host publication2017 IEEE SmartWorld, Ubiquitous Intelligence & Computing, Advanced & Trusted Computed, Scalable Computing & Communications, Cloud & Big Data Computing, Internet of People and Smart City Innovation (SmartWorld/SCALCOM/UIC/ATC/CBDCom/IOP/SCI)
PublisherIEEE
ISBN (Electronic)978-1-5386-0435-9, 978-1-5386-0434-2
ISBN (Print)978-1-5386-1591-1
DOIs
Publication statusPublished - 28 Jun 2018
Event14th IEEE International Conference on Ubiquitous Intelligence and Computing - San Francisco, United States
Duration: 4 Aug 20178 Aug 2017
http://ieee-smartworld.org/2017/uic/

Conference

Conference14th IEEE International Conference on Ubiquitous Intelligence and Computing
Abbreviated titleUIC 2017
CountryUnited States
CitySan Francisco
Period4/08/178/08/17
Internet address

Fingerprint

Dive into the research topics of 'Augmenting depth estimation from deep convolutional neural network using multi-spectral photometric stereo'. Together they form a unique fingerprint.

Cite this