[Coregroup-3d-coding] Book Chapter – Proposed Table of Contents

[Sjöström Mårten]

Caroline, all,

I think it is important to mention all these names to show that we are aware of their existence. It would also be an advantage to try explain their differences.

I would prefer the term Displacement Intra Prediction, as you suggested, because self-similarity is usually meant a similarity in different scales. Prediction, on the other hand, is about correlated information at different locations or times, i.e. in the same image or in different frames.


Best regards,
Mårten Sjöströmon
www.miun.se/en/personnel/MartenSjostrom<http://www.miun.se/en/personnel/MartenSjostrom>

From: conti.carol at gmail.com [mailto:conti.carol at gmail.com] On Behalf Of Caroline Conti
Sent: den 4 juni 2015 16:19
To: Kovács Péter Tamás; coregroup-3d-coding at lx.it.pt; Sjöström Mårten
Subject: Re: [Coregroup-3d-coding] Book Chapter – Proposed Table of Contents

Dear Mårten, Peter, and All,


·         The difference between Super multiview, Plenoptics, Lightfield, Holoscopic and Full Parallax Imaging should be made clear, at least in the context of this book chapter. As a consequence, the structure of the chapters may be preferred to be changed. I think this is crucial!
I second this opinion. I'd prefer to contribute to the light-field part, but now I'm listed for super-multiview, as that's the term used in MPEG. We should definitely (try to) clarify.


I agree it would be an asset to the chapter to make clear all the different nomenclatures that have been used for these technologies.
I had tried not to impose any particular name in this draft, since each of us would prefer to continue using our own terms.

An alternative would be to use a more general term (e.g.: Light Field), and to divide into two groups to make clear the greatest difference between the technologies. Then, simply to list all the names used to the same technology. For instance:

  *   Light field with horizontal parallax: Holografika’s Light-Field, Super-Multiview (adopted by MPEG FTV)
  *   Light field with full parallax: Integral Photography (adopted by MPEG FTV), Light-Field Photography (adopted by Lytro, Raytrix), Plenoptic (adopted by Adobe, JPEG), Holoscopic (adopted by 3D-VIVANT, Arri)
(The groups would also be: Light field Imaging with Camera Array, Light field Imaging with Plenoptic Camera).

Please let me know your opinion and suggestions.


•         “Multiple Hypothesis in Displacement Intra Prediction for Plenoptic Content” has the basis on Displacement Intra Prediction that you once proposed. I think you called it self-similarity prediction. Therefore, I think we should try to merge the contents or at least streamline the presentation and the nomenclature.
We could adopt the term Displacement Intra Prediction throughout the chapter, and include a reference to the term Self-Similarity at the first occurrence (e.g., firstly proposed as Self-Similarity Prediction).

•         There might me more authors to my contributions, but that is an easy fix.
Please feel free to add more authors.


Best Regards,
Caroline.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://cascais.lx.it.pt/pipermail/coregroup-3d-coding/attachments/20150604/f15c155e/attachment.html