Expand this Topic clickable element to expand a topic
Skip to content
Optica Publishing Group

N-dimensional correlation filters for pattern recognition

Not Accessible

Your library or personal account may give you access

Abstract

N-dimensional filters generalize the notion of a filter as a vector in a signal space to a filter acting as a N-dimensional subspace in a signal space.1,2 This new approach to filtering is easily implemented in an optical correlator by time-integrating the response of a series of filters on the output detector array. The resulting response is then threshold detected to determine if a target is present. This improves target discrimination for distortion-invariant recognition and helps alleviate the disparity between the operating speed of the filter spatial light modulator and the output detector array.2 An algorithm for designing distortion-invariant N-dimensional filters is presented, and simulation results will be given.

© 1990 Optical Society of America

PDF Article
More Like This
Optimization of the amplitude-modulated inverse filter for pattern-recognition applications

S. H. Zheng, M. A. Kraim, and M. L. Gao
ThTT1 OSA Annual Meeting (FIO) 1990

Phase-extraction pattern recognition

Tuvia Kotzer, Joseph Rosen, and Joseph Shamir
ThTT5 OSA Annual Meeting (FIO) 1990

Optical pattern recognition and the genetic code

William A. Christens-Barry, Larry D. Brasher, and James C. Martin
ThW5 OSA Annual Meeting (FIO) 1990

Select as filters


Select Topics Cancel
© Copyright 2024 | Optica Publishing Group. All rights reserved, including rights for text and data mining and training of artificial technologies or similar technologies.