Research Project

Compressed Sensing Theory

Compressed sensing in ultrahigh speed phonic imaging.

We are using the theory of compressed sensing in ultrahigh speed phonic imaging. Using broad band laser pulses and fiber optic dispersion processing, we can temporally encode a laser pulse with a random sequence, using this to spatially sample an object.

Compressed sensing theory says real signals are sparse and compressible, so they can be captured with much less information than previously thought required by Nyquist sampling. This allows us to use the projection of the randomly encoded laser pulse on an object to reconstruct it with many fewer measurements than a typical imaging system.  We have successfully demonstrated imaging speeds with line rates in the tens of MHz, becoming one of the fastest imaging systems in the world.

 

Back to top