Cossairt & Katsaggelos Receive Office of Naval Research & NU-ACCESS Awards

The grants are awarded for a three years from the Office of Naval Research & one year with NU-ACCESS.

Prof. Oliver (Ollie) S. Cossairt & Prof. Aggelos K. Katsaggelos

Prof. Oliver (Ollie) S. Cossairt, Lisa Wissner-Slivka and Benjamin Slivka Junior Professor of Computer Science and Prof. Aggelos K. Katsaggelos, AT&T Professor have received a three year award from the Office of Naval Research (ONR) entitled "Theoretical Bounds on Imaging through Scattering Media" (Cossairt PI) & a one year award with NU-ACCESS entitled "Approaches to Fast Widefield X-ray Imaging" (Katsagellos PI).

The OAR project is a collaboration with Prof. Ashok Veeraraghavan at Rice University, who is a sub-contractor, while the grant is termed from 9/1/2015-8/31-2018, for a total of $750k.

Project Description: In this proposal we develop theoretical bounds on imaging through scattering media using active illumination. Our analysis will provide quantitative bounds on the maximum performance that can be achieved when imaging at large distances (>10km) at high resolution (<10cm) through dense scattering media. Optimal performance will be achieved by applying a set of complimentary constraints (temporal, geometric, spatial-frequency, polarization and spectral) that can be combined together to create maximum rejection of scattered photons reaching the detector.   

The NU-ACCESS project is a collaboration with Marc Walton, senior scientist with NU-ACCESS, while the grant is termed from 9/1/2015-8/31/2015.

Project Description: In this proposal we will develop new approaches to capturing high-resolution X-Ray Fluorescence (XRF) scans extremely quickly yet without sacrificing counting statistics nor sharpness/fidelity of the resulting image. Instead of relying on a hardware approach to this problem (e.g., increasing the source intensity, or relying on more efficient detectors), we will use the existing NU-ACCESS scanning XRF instrument (XGlab Elio, Rh tube XRF) to test new point-scanning methodologies coupled to ‘smart’ image reconstruction methods. Towards the reduction of acquisition times we will focus on three distinct approaches adopted from the signal processing community , namely in-painting, super resolution, and noise removal.