Share Email Print

Proceedings Paper

Enhancement of optical correlation system performance utilizing a neural-network-based preprocessor filter
Author(s): Steve T. Kacenjar; H. Chen; D. Tong; T. Rimlinger; J. Blike
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

Map annotations (such as lettering) normally degrade automatic map-to-image optical correlation system performance which when removed, improves both the SNR and accuracy of such systems when generating conjugate data point pairs between the two optical formats. This paper describes improvements to the map-to-image correlation that results when an annotation removal preprocessor filter is applied first to map data. Specifically, the paper describes the impact of implementing a neural network annotation filter on the performance of map-to-image optical correlation systems. This new filter is capable of automatically identifying and then removing annotations before performing the optical correlation. As shown, this removal process impacts the correlation SNR and phase-only filtering systems. Greatest improvement in system performance is achieved when the annotation filter is applied first to map data before implementing a binary, phase-only filtering process.

Paper Details

Date Published: 1 July 1990
PDF: 10 pages
Proc. SPIE 1246, Parallel Architectures for Image Processing, (1 July 1990); doi: 10.1117/12.19591
Show Author Affiliations
Steve T. Kacenjar, General Electric Co. (United States)
H. Chen, General Electric Co. (United States)
D. Tong, General Electric Co. (United States)
T. Rimlinger, General Electric Co. (United States)
J. Blike, General Electric Co. (United States)

Published in SPIE Proceedings Vol. 1246:
Parallel Architectures for Image Processing
Joydeep Ghosh; Colin G. Harrison, Editor(s)

© SPIE. Terms of Use
Back to Top