Real-Time Light Field Video Focusing and GPU Accelerated Streaming

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26230%2F23%3APU149403" target="_blank" >RIV/00216305:26230/23:PU149403 - isvavai.cz</a>

Výsledek na webu

<a href="https://link.springer.com/article/10.1007/s11265-023-01874-8" target="_blank" >https://link.springer.com/article/10.1007/s11265-023-01874-8</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1007/s11265-023-01874-8" target="_blank" >10.1007/s11265-023-01874-8</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Real-Time Light Field Video Focusing and GPU Accelerated Streaming

Popis výsledku v původním jazyce

This paper proposes a novel solution of real-time depth range and correct focusing estimation in light field videos represented by arrays of video sequences. This solution, compared to previous approaches, offers a novel way to render high-quality synthetic views from light field videos on contemporary hardware in real-time. Only the video frames containing color information with no other attributes, such as captured depth, are needed. The drawbacks of the previous proposals such as block artifacts in the defocused parts of the scene or manual setting of the depth range are also solved in this paper. This paper describes a complete solution that solves the main memory and performance issues of light field rendering on contemporary personal computers. The whole integration of high-quality light field videos supersedes the approaches in previous works and the paper also provides measurements and experimental results. While reaching the same quality as slower state-of-the-art approaches, this method can still be used in real-time which makes it suitable for industry and real-life scenarios as an alternative to standard 3D rendering approaches.

Název v anglickém jazyce

Real-Time Light Field Video Focusing and GPU Accelerated Streaming

Popis výsledku anglicky

This paper proposes a novel solution of real-time depth range and correct focusing estimation in light field videos represented by arrays of video sequences. This solution, compared to previous approaches, offers a novel way to render high-quality synthetic views from light field videos on contemporary hardware in real-time. Only the video frames containing color information with no other attributes, such as captured depth, are needed. The drawbacks of the previous proposals such as block artifacts in the defocused parts of the scene or manual setting of the depth range are also solved in this paper. This paper describes a complete solution that solves the main memory and performance issues of light field rendering on contemporary personal computers. The whole integration of high-quality light field videos supersedes the approaches in previous works and the paper also provides measurements and experimental results. While reaching the same quality as slower state-of-the-art approaches, this method can still be used in real-time which makes it suitable for industry and real-life scenarios as an alternative to standard 3D rendering approaches.

Druh

J_imp - Článek v periodiku v databázi Web of Science

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

<a href="/cs/project/8A21015" target="_blank" >8A21015: AI-augmented automation for efficient DevOps, a model-based framework for continuous development At RunTime in cyber-physical systems</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)<br>S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2023

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název periodika

Journal of Signal Processing Systems for Signal Image and Video Technology

ISSN

1939-8018

e-ISSN

1939-8115

Svazek periodika

95

Číslo periodika v rámci svazku

6

Stát vydavatele periodika

US - Spojené státy americké

Počet stran výsledku

17

Strana od-do

703-719

Kód UT WoS článku

000991328300001

EID výsledku v databázi Scopus

2-s2.0-85160112835