Designing objective quality metrics for panoramic videos based on human perception

Sandra Nabil 1 Raffaella Balzarini 2, 3 Frédéric Devernay 4 James L. Crowley 3
1 IMAGINE - Intuitive Modeling and Animation for Interactive Graphics & Narrative Environments
Inria Grenoble - Rhône-Alpes, LJK - Laboratoire Jean Kuntzmann, INPG - Institut National Polytechnique de Grenoble
3 PERVASIVE - Interaction située avec les objets et environnements intelligents
Inria Grenoble - Rhône-Alpes, LIG - Laboratoire d'Informatique de Grenoble, Grenoble INP - Institut polytechnique de Grenoble - Grenoble Institute of Technology, UGA - Université Grenoble Alpes
Abstract : The creation of high quality panoramic videos for immersive VR content is commonly done using a rig with multiple cameras covering the required scene. Unfortunately, this setup introduces both spatial and temporal artifacts due to the difference in optical centers as well as imperfect synchronization of cameras. Traditional image quality metrics are inadequate for describing geometric distortions in panoramic videos. In this paper, we propose an objective quality assessment approach for calculating these distortions based on optical flow coupled with an existing salience detection model. Our approach is validated with a human-centered study using error annotations and eye-tracking. Preliminary results indicate a good correlation between errors detected by the algorithm and human perception of errors.
Complete list of metadatas

Cited literature [10 references]  Display  Hide  Download

https://hal.inria.fr/hal-01849043
Contributor : Sandra Nabil <>
Submitted on : Wednesday, July 25, 2018 - 2:17:18 PM
Last modification on : Thursday, May 2, 2019 - 3:30:31 PM
Long-term archiving on : Friday, October 26, 2018 - 2:36:16 PM

File

imvip2018_camera_ready_paper31...
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01849043, version 1

Citation

Sandra Nabil, Raffaella Balzarini, Frédéric Devernay, James L. Crowley. Designing objective quality metrics for panoramic videos based on human perception. IMVIP 2018 - Irish Machine Vision and Image Processing Conference, Aug 2018, Belfast, United Kingdom. pp.1-4. ⟨hal-01849043⟩

Share

Metrics

Record views

913

Files downloads

179