Compressed imaging by sparse random convolution

Diego just sent me the following:

Hello,

I've followed NuitBlanche for a couple years now and it would be a honour for me if you would consider a recent publication of ours to appear in your CS hardware list.

We only point out that Gaussian measurement matrices are not feasible when measuring incoherent light and propose to use sparse non-negative random measurement matrices instead. The article can be found here.

Thanks a lot for giving us your time through your blog!

Diego Marcos

Sure Diego !

Compressed imaging by sparse random convolution by Diego Marcos, Theo Lasser, Antonio López, and Aurélien Bourquard

The theory of compressed sensing (CS) shows that signals can be acquired at sub-Nyquist rates if they are sufficiently sparse or compressible. Since many images bear this property, several acquisition models have been proposed for optical CS. An interesting approach is random convolution (RC). In contrast with single-pixel CS approaches, RC allows for the parallel capture of visual information on a sensor array as in conventional imaging approaches. Unfortunately, the RC strategy is difficult to implement as is in practical settings due to important contrast-to-noise-ratio (CNR) limitations. In this paper, we introduce a modified RC model circumventing such difficulties by considering measurement matrices involving sparse non-negative entries. We then implement this model based on a slightly modified microscopy setup using incoherent light. Our experiments demonstrate the suitability of this approach for dealing with distinct CS scenarii, including 1-bit CS.

 
 

Join the CompressiveSensing subreddit or the Google+ Community or the Facebook page and post there !
Liked this entry ? subscribe to Nuit Blanche's feed, there's more where that came from. You can also subscribe to Nuit Blanche by Email, explore the Big Picture in Compressive Sensing or the Matrix Factorization Jungle and join the conversations on compressive sensing, advanced matrix factorization and calibration issues on Linkedin.