Monday, April 29, 2013

Compressive Light Field Photography Using Overcomplete Dictionaries And Optimized Projections



After watching Saturday's video on the work at Duke on Hyperspectral video imaging, I was not expecting the item Kshitij Marwah just sent me:

Dear Igor, 
Have been a great fan of your blog. There is some recent work I did on a new practical light field camera architecture using compressive sensing that, thanks to its great resolution, is very competitive to LYTRO. Though still a research prototype I wanted to draw your attention to it. This is the first practical and working light field camera based on CS techniques.This technique allows conversion of today's DSLRs into light field cameras as well so that people do not have to go and buy a new camera for refocusing and 3D.
Let me know what you think. Thanks.
regards,

Wow!  thanks Kshitij. Here is the paper and accompanying videos from the project page:


Light field photography has gained a significant research interest in the last two decades; today, commercial light field cameras are widely available. Nevertheless, most existing acquisition approaches either multiplex a low-resolution light field into a single 2D sensor image or require multiple photographs to be taken for acquiring a high-resolution light field. We propose a compressive light field camera architecture that allows for higher-resolution light fields to be recovered than previously possible from a single image. The proposed architecture comprises three key components:light field atoms as a sparse representation of natural light fields, an optical design that allows for capturing optimized 2D light field projections, and robust sparse reconstruction methods to recover a 4D light field from a single coded 2D projection. In addition, we demonstrate a variety of other applications for light field atoms and sparse coding techniques, including 4D light field compression and denoising.




No comments:

Printfriendly