Abstract | ||
---|---|---|
A plenoptic camera consists of microlenses that enable the measuring of directional information. A common image formation model of the plenoptic camera is based on geometrical optics and provides an object reconstruction with a resolution equal to the number of microlenses. We use a more accurate model that takes into account the wave properties of light. A discrete formulation of this model is done in order to develop the inverse imaging technique which is efficient for super resolution object reconstruction even from noisy data. The proposed algorithm is based on a special modification of the IDD-BM3D deblurring algorithm. The achieved resolution highly exceeds the number of microlenses which is what we call super resolution imaging. Simulations show that the proposed strategy essentially outperforms the previously suggested method based on the wavefield modeling. |
Year | DOI | Venue |
---|---|---|
2014 | 10.1109/3DTV.2014.6874735 | 3DTV-Conference: The True Vision - Capture, Transmission and Display of 3D Video |
Keywords | Field | DocType |
cameras,computerised instrumentation,geometrical optics,image resolution,image restoration,microlenses,IDD-BM3D deblurring algorithm,common image formation model,directional information measurement,geometrical optics,light wave property,microlens,object reconstruction,plenoptic camera,super resolution inverse imaging,wavefield modeling,Fourier optics,inverse imaging,plenoptic camera,super resolution,wavefield modeling | Computer vision,Inverse,Noisy data,Deblurring,Computer graphics (images),Computer science,Image formation,Artificial intelligence,Geometrical optics,Call super,Superresolution | Conference |
ISSN | Citations | PageRank |
2161-2021 | 0 | 0.34 |
References | Authors | |
2 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Petri Helin | 1 | 16 | 3.06 |
Vladimir Katkovnik | 2 | 3414 | 178.00 |
Atanas P. Gotchev | 3 | 223 | 38.55 |
Jaakko Astola | 4 | 1515 | 230.41 |