Epipolar plane images based light-field angular super-resolution network

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Light field (LF) imaging has proven to be a promising technique in computer vision field. However, there is a tradeoff between spatial and angular resolution of LF images, which limits the application of LF cameras. Super-resolution (SR) of the angular domain is proposed to improve the angular resolution of LF images. However, most of the SR frameworks cannot adapt to LF datasets with multi-size disparities, especially the large disparities. In this paper, we proposed a learning-based SR framework named EASRnet. The EASRnet consists of three parts-Disparity adaptation, Feature extraction, and Feature restoration parts, and achieves angular SR tasks by using residual blocks and a structure with branches to reconstruct high-frequency details of up-sampled epipolar plane images (EPI). It employs an additional blur layer to accommodate LF datasets with different disparities. The experimental results show that the proposed approach can reconstruct novel view images with satisfactory accuracy.

Original languageEnglish
Title of host publicationSeventh Asia Pacific Conference on Optics Manufacture, APCOM 2021
EditorsJiubin Tan, Xiangang Luo, Ming Huang, Lingbao Kong, Dawei Zhang
PublisherSPIE
ISBN (Electronic)9781510652088
DOIs
StatePublished - 2022
Event7th Asia Pacific Conference on Optics Manufacture, APCOM 2021 - Shanghai, China
Duration: 28 Oct 202131 Oct 2021

Publication series

NameProceedings of SPIE - The International Society for Optical Engineering
Volume12166
ISSN (Print)0277-786X
ISSN (Electronic)1996-756X

Conference

Conference7th Asia Pacific Conference on Optics Manufacture, APCOM 2021
Country/TerritoryChina
CityShanghai
Period28/10/2131/10/21

Keywords

  • Light field
  • Network
  • Super-resolution

Fingerprint

Dive into the research topics of 'Epipolar plane images based light-field angular super-resolution network'. Together they form a unique fingerprint.

Cite this