학술논문
Hierarchical ship classifier for airborne synthetic aperture radar (SAR) images
Document Type
Conference
Author
Source
Conference Record of the Thirty-Third Asilomar Conference on Signals, Systems, and Computers (Cat. No.CH37020) Signals, systems and computers Signals, Systems, and Computers, 1999. Conference Record of the Thirty-Third Asilomar Conference on. 2:1230-1234 vol.2 1999
Subject
Language
ISSN
1058-6393
Abstract
Lockheed Martin Canada has developed an agent-based adaptable data fusion testbed (ADFT) within the knowledge based system (KBS) architecture which is currently made of a multi-sensor data fusion (MSDF) module and an image support module (ISM). The MSDF module fuses the information provided by nonimaging (2D-radar, ESM) sensors and the various propositions provided by the ISM when processing a synthetic aperture radar (SAR) image. Currently, the ISM processes, simulated and/or real images of ships through a four-step hierarchical classifier that can extract attributes such as ship length, ship category, ship type and ship class. The SAR classifier can distinguish between merchant and combatant categories and can select amongst 5 combatant types. Tests on simulated and real SAR images show a good recognition rate up to the ship type for merchant and line ships.