Skip to Main content Skip to Navigation
Conference papers

Génération d'images omnidirectionnelles à partir d'un environnement virtuel

Ahmed Rida Sekkat 1, 2 Yohan Dupuis 3 Pascal Vasseur 1 Paul Honeine 2
1 STI - LITIS - Equipe Systèmes de Transport Intelligent
LITIS - Laboratoire d'Informatique, de Traitement de l'Information et des Systèmes
2 DocApp - LITIS - Equipe Apprentissage
LITIS - Laboratoire d'Informatique, de Traitement de l'Information et des Systèmes
Abstract : This paper describes a method for generating omnidirectional images using cubemap images and corresponding depth maps that can be acquired from a virtual environment. For this purpose, we use the video game Grand Theft Auto V (GTA V). GTA V has been used as a data source in many research projects, due to the fact that it is a hyperrealist open-world game that simulates a real city. We take advantage of developments made in reverse engineering this game, in order to extract realistic images and corresponding depth maps using virtual cameras with 6DoF. By combining the extracted information with an omnidirectional camera model, we generate Fish-eye images intended for instance to machine learning based applications.
Complete list of metadatas

Cited literature [30 references]  Display  Hide  Download

https://hal-normandie-univ.archives-ouvertes.fr/hal-02183033
Contributor : Paul Honeine <>
Submitted on : Sunday, July 14, 2019 - 6:25:20 PM
Last modification on : Thursday, March 5, 2020 - 3:31:00 PM

File

19.gretsi.fisheye.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-02183033, version 1

Citation

Ahmed Rida Sekkat, Yohan Dupuis, Pascal Vasseur, Paul Honeine. Génération d'images omnidirectionnelles à partir d'un environnement virtuel. 27-ème Colloque GRETSI sur le Traitement du Signal et des Images, Aug 2019, Lille, France. ⟨hal-02183033⟩

Share

Metrics

Record views

201

Files downloads

134