Problems Associated with Compressed Sensing in Processing Signals on Sphere

Yibeltal Alem (ANU)

APPLIED SIGNAL PROCESSING SERIES

DATE: 2012-08-02
TIME: 10:00:00 - 11:00:00
LOCATION: RSISE Seminar Room, ground floor, building 115, cnr. North and Daley Roads, ANU
CONTACT: JavaScript must be enabled to display this email address.

ABSTRACT:
We investigate the sensitivity problems associated with compressed sensing (CS) when it is applied to the reconstruction of spherical harmonic (SH) signals. A signal on the sphere can be expressed with SH coefficients, and we only need to calculate the coefficients to represent the signal with SH basis functions. CS allows us to tremendously reduce the number of measurements required to calculate the coefficients. However, it is difficult to avoid contamination of measurements by noise and sampling jitter on spherical points during the process of acquiring CS samples. The contaminating noise and the sampling jitter can cause additive and multiplicative perturbations, respectively. These perturbations have severe adverse effects on calculating the SH coefficients with CS. In this paper, we show the problems of calculating the SH coefficients with CS under the two main perturbations. Furthermore, we introduce SH calibration and SH diversity techniques to reduce the effects of the aforementioned perturbations. We also support our claims by numerical results to show that the proposed methods can reduce the effect of noise significantly.

BIO:
Yibeltal is a PhD student in the ASP group.