NTRS - NASA Technical Reports Server

Back to Results
optical implementation of a feature-based neural network with application to automatic target recognitionAn optical neural network based on the neocognitron paradigm is introduced. A novel aspect of the architecture design is shift-invariant multichannel Fourier optical correlation within each processing layer. Multilayer processing is achieved by feeding back the ouput of the feature correlator interatively to the input spatial light modulator and by updating the Fourier filters. By training the neural net with characteristic features extracted from the target images, successful pattern recognition with intraclass fault tolerance and interclass discrimination is achieved. A detailed system description is provided. Experimental demonstrations of a two-layer neural network for space-object discrimination is also presented.
Document ID
Document Type
Reprint (Version printed in journal)
Chao, Tien-Hsin
(JPL Pasadena, CA, United States)
Stoner, William W.
(Science Applications International Corp. Billerica, MA, United States)
Date Acquired
August 16, 2013
Publication Date
March 10, 1993
Publication Information
Publication: Applied Optics
Volume: 32
Issue: 8
ISSN: 0003-6935
Subject Category
Funding Number(s)
CONTRACT_GRANT: N00014-86-C-0601
Distribution Limits