Experimental results with optimal filters for reliable recognition of characters

J. Campos, F. Turon, M. J. Yzuel, L. P. Yaroslavsky

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

To evaluate the filter performance in optical pattern recognition different criteria may be used. Each of them takes into account some features of the recognition process. Many efforts in this field are devoted to the design of filters which optimize some criteria: Classical Matched Filter optimizes the robustness against noise; Phase Only Filter, the light efficiency; Minimum Average Correlation Filter, minimizes the energy of the correlation plane, etc. The synthesis of a filter with maximal discrimination capability was described and is called Optimal Filter. In this work we present different ways to design the Optimal Filter depending on the priori knowledge of the scene and the objects to be rejected. Some of these filters are applied to colour pattern recognition using a multi-channel correlator.

Original languageEnglish
Title of host publicationOptics as a Key to High Technology
EditorsGy. Akos, T. Lippenyi, G. Lupkovics, A. Podmaniczky
PublisherPubl by Society of Photo-Optical Instrumentation Engineers
Pages376-377
Number of pages2
Editionpt 1
ISBN (Print)0819412309
StatePublished - 1993
Externally publishedYes
EventProceedings of the 16th Congress of the International Commission for Optics - Budapest, Hung
Duration: 9 Aug 199313 Aug 1993

Publication series

NameProceedings of SPIE - The International Society for Optical Engineering
Numberpt 1
Volume1983
ISSN (Print)0277-786X

Conference

ConferenceProceedings of the 16th Congress of the International Commission for Optics
CityBudapest, Hung
Period9/08/9313/08/93

Fingerprint

Dive into the research topics of 'Experimental results with optimal filters for reliable recognition of characters'. Together they form a unique fingerprint.

Cite this