1 January 2004 Enhanced iterative processing algorithms for restoration and superresolution of tactical sensor imagery
Author Affiliations +
Optical Engineering, 43(1), (2004). doi:10.1117/1.1626665
Abstract
Imagery data acquired in practice to support tactical surveillance and tracking missions in hostile environments typically suffer from a variety of degradations making it essential to subject the data to digital postprocessing aimed at restoration and superresolution before they can be used for any image exploitation tasks (visualization, target detection and characterization, etc.). A number of novel iterative techniques for resolution enhancement are presently being developed, with statistical optimization and set-theoretic estimation offering two popular approaches for algorithm design. The challenges posed by the processing needs of tactical imagery data often require greater capabilities than what the existing algorithms can offer, however, and typically require more enhanced procedures to achieve satisfactory restoration and superresolution. We outline three such enhancements: parallel projection implementation with adaptive relaxation, use of scene-derived information for constraint set design, and a hybrid statistical and set-theoretic estimation procedure. The restoration and superresolution performance of an iterative algorithm that incorporates these enhancements is illustrated by application to tactical imagery data [images acquired from state-of-the-art synthetic aperture radar (SAR) and passive millimeter-wave (PMMW) sensors].
Malur K. Sundareshan, Supratik Bhattacharjee, "Enhanced iterative processing algorithms for restoration and superresolution of tactical sensor imagery," Optical Engineering 43(1), (1 January 2004). http://dx.doi.org/10.1117/1.1626665
JOURNAL ARTICLE
10 PAGES


SHARE
KEYWORDS
Image processing

Image restoration

Super resolution

Algorithm development

Sensors

Image enhancement

Expectation maximization algorithms

RELATED CONTENT


Back to Top