par Fachada, Sarah 
;Bonatto, Daniele ;Lafruit, Gauthier ;Teratani, Mehrdad
Référence EUVIP(2024: September 8-11, 2024: Geneve, Suisse), 12th EUROPEAN WORKSHOP ON VISUAL INFORMATION PROCESSING
Publication Publié, 2024-09-09
;Bonatto, Daniele ;Lafruit, Gauthier ;Teratani, Mehrdad Référence EUVIP(2024: September 8-11, 2024: Geneve, Suisse), 12th EUROPEAN WORKSHOP ON VISUAL INFORMATION PROCESSING
Publication Publié, 2024-09-09
Publication dans des actes
| Résumé : | This paper presents a new calibration method for the focused plenoptic cameras that do not rely on subaperture view images and on calibration patterns. Given a single view captured with a plenoptic camera, and the key distances of the scene with a few millimeters incertitude, we retrieve the principal internal camera parameters of the camera, using micro-disparities across its micro-images. Such a method allows calibration of datasets that are provided without calibration values or calibration images and, by bypassing the step of subaperture view extraction, avoids errors due to reconstruction artifacts. We compare our method to a subaperture-free pattern-based method and a subaperture-based pattern-free method on synthetic and natural datasets, showing reliable results. |
