Advanced search
Start date
(Reference retrieved automatically from Web of Science through information on FAPESP grant and its corresponding number as mentioned in the publication by the authors.)

Light Field Image Quality Enhancement by a Lightweight Deformable Deep Learning Framework for Intelligent Transportation Systems

Full text
Ribeiro, David Augusto [1] ; Silva, Juan Casavilca [2] ; Lopes Rosa, Renata [1] ; Saadi, Muhammad [3] ; Mumtaz, Shahid [4] ; Wuttisittikulkij, Lunchakorn [5] ; Zegarra Rodriguez, Demostenes [1] ; Al Otaibi, Sattam [6]
Total Authors: 8
[1] Univ Fed Lavras, Dept Comp Sci, BR-37200000 Lavras, MG - Brazil
[2] Pontifical Catholic Univ Peru, Dept Sci, Lima 1801 - Peru
[3] Univ Cent Punjab, Fac Engn, Dept Elect Engn, Lahore 54000 - Pakistan
[4] Inst Telecomunicaces, P-3750011 Aveiro - Portugal
[5] Chulalongkorn Univ, Dept Elect Engn, Wireless Commun Ecosyst Res Unit, Bangkok 10330 - Thailand
[6] Taif Univ, Coll Engn, Dept Elect Engn, At Taif 21944 - Saudi Arabia
Total Affiliations: 6
Document type: Journal article
Source: ELECTRONICS; v. 10, n. 10 MAY 2021.
Web of Science Citations: 0

Light field (LF) imaging has multi-view properties that help to create many applications that include auto-refocusing, depth estimation and 3D reconstruction of images, which are required particularly for intelligent transportation systems (ITSs). However, cameras can present a limited angular resolution, becoming a bottleneck in vision applications. Thus, there is a challenge to incorporate angular data due to disparities in the LF images. In recent years, different machine learning algorithms have been applied to both image processing and ITS research areas for different purposes. In this work, a Lightweight Deformable Deep Learning Framework is implemented, in which the problem of disparity into LF images is treated. To this end, an angular alignment module and a soft activation function into the Convolutional Neural Network (CNN) are implemented. For performance assessment, the proposed solution is compared with recent state-of-the-art methods using different LF datasets, each one with specific characteristics. Experimental results demonstrated that the proposed solution achieved a better performance than the other methods. The image quality results obtained outperform state-of-the-art LF image reconstruction methods. Furthermore, our model presents a lower computational complexity, decreasing the execution time. (AU)

FAPESP's process: 18/26455-8 - Audio-Visual Speech Processing by Machine Learning
Grantee:Miguel Arjona Ramírez
Support type: Regular Research Grants