A learning-based view extrapolation method for axial super-resolution

Zhaolin Xiao (1) , Jinglei Shi (2) , Xiaoran Jiang (2) , Christine Guillemot (2)

1 XUT - Xi'an University of Technology
2 Sirocco - Analysis representation, compression and communication of visual data

Axial light field resolution refers to the ability to distinguish features at different depths by refocusing. The axial refocusing precision corresponds to the minimum distance in the axial direction between two distinguishable refocusing planes. High refocusing precision can be essential for some light field applications like microscopy. In this paper, we propose a learningbased method to extrapolate novel views from axial volumes of sheared epipolar plane images (EPIs). As extended numerical aperture (NA) in classical imaging, the extrapolated light field gives re-focused images with a shallower depth of field (DOF), leading to more accurate refocusing results. Most importantly, the proposed approach does not need accurate depth estimation. Experimental results with both synthetic and real light fields show that the method not only works well for light fields with small baselines as those captured by plenoptic cameras (especially for the plenoptic 1.0 cameras), but also applies to light fields with larger baselines.

Mots clés: Light field; refocus precision; view extrapolation; convolutional network; axial resolution

Domaines: Traitement du signal et de l'image [eess.SP]

Zhaolin Xiao, Jinglei Shi, Xiaoran Jiang, Christine Guillemot. A learning-based view extrapolation method for axial super-resolution. Neurocomputing, 2021, 455, pp.229-241. ⟨10.1016/j.neucom.2021.05.056⟩. ⟨hal-03230321⟩