The hybrid optoelectronic neural object recognition system (HONORS), underw
ay at the jet propulsion laboratory (JPL), consists of two major building b
locks: (1) an advanced grayscale optical correlator (OC) and (2) a massivel
y parallel, neural 3-D processor (N3DP). The OC, with its inherent advantag
es in parallel processing and shift invariance, will be used for target of
interest (TOI) detection and segmentation. The N3DP,with its robust neural
learning capability, will be used for target classification and identificat
ion. HONORS, with its powerful combination of optical processing and neural
networks, will enable real-time, large frame, automatic target recognition
(ATR). This paper presents the system architecture and processing algorith
ms. We provide the results from simulations and experiments, including the
detection, classification, and tracking of tanks and aircraft. We will also
discuss the potential application of using HONORS for real-time digital pa
lomar observatory survey system (DPOSS). (C) 1999 Published by Elsevier Sci
ence B.V. All rights reserved.