Light-Field Demultiplexing and Disparity Estimation
Résumé
In this paper we study the post-processing pipeline to recover the views (light-field) from the raw data of a plenoptic camera such as Lytro. First, the microlens centers are estimated and then the raw image is demultiplexed without demosaicing it beforehand. This avoids image artifacts due to view cross-talk. Furthermore, we present a new blockmatching algorithm to estimate disparities for plenoptic views that have not been demosaiced. Our algorithm enforces the coherence through the views thanks to the view configuration given by the plenoptic camera: (i) the views are horizontally and vertically rectified and have the same baseline, and therefore (ii) at each point, the vertical and horizontal disparities are the same. Finally, we show that disparity estimation is more accurate when the raw image is demultiplexed without demosaicing the raw image. In particular, we show that our algorithm outperforms the disparity estimation method in [17].
Domaines
Traitement des images [eess.IV]
Origine : Fichiers produits par l'(les) auteur(s)
Loading...