Abstract

There is a growing awareness in the scientific community that as methods of analysis become more efficient, a massive data explosion is ensuing. Reducing data to meaningful trends has become a major aspect of research because of the bulk of information involved and because scientific interest is focusing on more and more subtle phenomena. One aid to data reduction and trend detection is pattern recognition, which is a collection of computer-based data manipulation techniques that have proved useful in a number of problems. This paper is intended as an illustrative introduction to these techniques, with the aim of giving the reader an intuitive feel for the process of a real application of the methods. A problem in archaeology, solved by neutron activation analysis, is discussed.

Recognizing Patterns in Photographs

C. K. Rushforth
Appl. Opt. 4(4) 379-381 (1965)

Parallel optical syntactic pattern recognizers

George Eichmann and Sanghamitra Basu
Appl. Opt. 26(10) 1859-1865 (1987)

Comment on “Recognizing Patterns in Photographs”

Thomas J. Harley and Laveen N. Kanal
Appl. Opt. 4(10) 1351-1352 (1965)

Masking fields: a massively parallel neural architecture for learning, recognizing, and predicting multiple groupings of patterned data

Michael A. Cohen and Stephen Grossberg
Appl. Opt. 26(10) 1866-1891 (1987)

Complex phase tracing method for fringe pattern analysis

Janusz Kozłowski and Giovanni Serra
Appl. Opt. 38(11) 2256-2262 (1999)