| Paper ID | B-1-2.6 |
| Paper Title |
A Match Pursuit Based Method Adapted to Overcomplete Dictionary for Compressive Spectral Imaging |
| Authors |
Jianchen Zhu, Shengjie Zhao, Rongqing Zhang, Tongji University, China |
| Session |
B-1-2: Adaptive and Intelligent Signal Processing |
| Time | Tuesday, 08 December, 15:30 - 17:00 |
| Presentation Time: | Tuesday, 08 December, 16:45 - 17:00 Check your Time Zone |
|
All times are in New Zealand Time (UTC +13) |
| Topic |
Signal and Information Processing Theory and Methods (SIPTM): |
| Abstract |
This paper proposes to introduce greedy pursuit, a conceptually simple hard thresholding pursuit algorithm called Signal Space Subspace Pursuit (SSSP) for calculating spare signal representations with overcomplete dictionaries whenever the sensing matrix (sampling operator) satisfies the Restricted Isometry Property adapted to Dictionary (D-RIP). Signal space greedy method has the ability to optimally compute (near) best projections that allow one to identify the most related a small number of dictionary atoms of an arbitrary signal in this setting. More practically, standard CS recovery algorithms are applicable to such projections while maintaining accurate signal recovery. Simulation results with a typical hyperspectral data set demonstrate the superiority of the proposed approach. |