Digit classification using an edge based hierarchical neural representation

Autor:	E. Seidenberg, H. Wiemers, M. Busemann, K. O. Kräuter, Georg Hartmann
Rok vydání:	1992
Předmět:	Generalization Computer science business.industry Representation (systemics) Edge based Computer vision Artificial intelligence Enhanced Data Rates for GSM Evolution Content-addressable memory business Numerical digit
DOI:	10.1016/b978-0-444-89488-5.50160-3
Popis:	A pattern recognition system is presented. Edge information is used in order to generate an abstract representation of the camera image. Generalization of the information takes place mainly in special preprocessing steps. An appropiate associative memory divides the inputs in different groups or creates new groups if necessary. In this article we demonstrate the efficiency of our system for handwritten digits.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::151c9da87b1a0634a3fede385b100d81 https://doi.org/10.1016/b978-0-444-89488-5.50160-3 Zobrazit plný text záznamu