Graph-based Transforms for Predictive Light Field Compression based on Super-Pixels - Archive ouverte HAL Access content directly
Conference Papers Year :

Graph-based Transforms for Predictive Light Field Compression based on Super-Pixels

(1) , (1) , (1) , (1)
1

Abstract

In this paper, we explore the use of graph-based transforms to capture correlation in light fields. We consider a scheme in which view synthesis is used as a first step to exploit interview correlation. Local graph-based transforms (GT) are then considered for energy compaction of the residue signals. The structure of the local graphs is derived from a coherent super-pixel over-segmentation of the different views. The GT is computed and applied in a separable manner with a first spatial unweighted transform followed by an interview GT. For the interview GT, both unweighted and weighted GT have been considered. The use of separable instead of non separable transforms allows us to limit the complexity inherent to the computation of the basis functions. A dedicated simple coding scheme is then described for the proposed GT based light field decomposition. Experimental results show a significant improvement with our method compared to the CNN view synthesis method and to the HEVC direct coding of the light field views.
Fichier principal
Vignette du fichier
ICASSP18_GTLF.pdf (643.01 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-01737332 , version 1 (19-03-2018)

Identifiers

Cite

Mira Rizkallah, Xin Su, Thomas Maugey, Christine Guillemot. Graph-based Transforms for Predictive Light Field Compression based on Super-Pixels. ICASSP 2018 - IEEE International Conference on Acoustics, Speech and Signal Processing, Apr 2018, Calgary, Canada. pp.1718-1722, ⟨10.1109/ICASSP.2018.8462288⟩. ⟨hal-01737332⟩
305 View
443 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More