4.7 Article

Efficient Light Field Reconstruction via Spatio-Angular Dense Network

Journal

Publisher

IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TIM.2021.3100326

Keywords

Convolutional neural network; deep learning; image processing; light field (LF) imaging; light field; reconstruction

Funding

  1. Research Foundation for Advanced Talents of Beijing Technology and Business University [19008021181]
  2. National Natural Science Foundation of China [61877002]
  3. Beijing Natural Science Foundation
  4. Fengtai Rail Transit Frontier Research Joint Fund [L191009]
  5. Scientific Research Program of Beijing Municipal Education Commission [KZ202110011017]

Ask authors/readers for more resources

The article introduces an end-to-end spatio-angular dense network (SADenseNet) for light field reconstruction, incorporating correlation blocks and spatio-angular dense skip connections to address domain asymmetry and efficient information flow issues. Through extensive experiments on real-world and synthetic datasets, it is demonstrated that SADenseNet achieves state-of-the-art performance with significantly reduced memory and computation costs, resulting in sharp and detailed reconstructed light field images that can improve the accuracy of measurement applications.
As an image sensing instrument, light field images can supply extra angular information compared with monocular images and have facilitated a wide range of measurement applications. Light field image capturing devices usually suffer from the inherent tradeoff between the angular and spatial resolutions. To tackle this problem, several methods, such as light field reconstruction and light field super-resolution, have been proposed but leaving two problems unaddressed, namely domain asymmetry and efficient information flow. In this article, we propose an end-to-end spatio-angular dense network (SADenseNet) for light field reconstruction with two novel components, namely correlation blocks and spatio-angular dense skip connections to address them. The former performs effective modeling of the correlation information in a way that conforms with the domain asymmetry. Also, the latter consists of three kinds of connections enhancing the information flow within two domains. Extensive experiments on both real-world and synthetic datasets have been conducted to demonstrate that the proposed SADenseNet's state-of-the-art performance at significantly reduced costs in memory and computation. The qualitative results show that the reconstructed light field images are sharp with correct details and can serve as preprocessing to improve the accuracy of related measurement applications.

Authors

I am an author on this paper
Click your name to claim this paper and add it to your profile.

Reviews

Primary Rating

4.7
Not enough ratings

Secondary Ratings

Novelty
-
Significance
-
Scientific rigor
-
Rate this paper

Recommended

No Data Available
No Data Available