Translator Disclaimer
27 August 2010 Information theoretic characterizations of compressive-sensing-based space object identification
Author Affiliations +
The solar-reflected brightness distribution of a man-made space object has regions of spatially uniform brightness and spectral content that are interrupted only by boundaries separating one material region from another. The relatively simple structure of this distribution permits, as we demonstrate here, spectral-correlation-based strategies to extract information about the boundaries and material constituents of the segments of the object surface. Still simpler compressive-sensing (CS) based approaches that require no specific spectral analysis can also efficiently perform such information extraction, which is a critical task of any space-object identification (SOI) system. We analyze here these latter approaches by means of statistical information theory (IT) in the context of a highly idealized satellite model with rectilinear material boundaries and quasi-one-dimensional (1D) brightness distribution. Our analysis includes spectrally dependent diffractive blur as well as detector noise against which we optimize, via our IT calculations, the choice of the CS mask set, the bandwidth of the spectral measurements, and the minimum number of measurements needed for extracting information about the boundary locations and material identities.
© (2010) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Sudhakar Prasad and Douglas Hope "Information theoretic characterizations of compressive-sensing-based space object identification", Proc. SPIE 7800, Image Reconstruction from Incomplete Data VI, 78000B (27 August 2010);

Back to Top