Application of a Proposed Efficient Smoothing Algorithm on Directional Vectors for Isolated Handwritten Characters

Autor:	Otman M. Razib, Zafar M. Faisal, M. Dzulkifli, Abdul Rab
Rok vydání:	2008
Předmět:	Neural gas Artificial neural network business.industry Time delay neural network Computer science Feature extraction Trajectory Pattern recognition Artificial intelligence business Algorithm Character recognition Smoothing
Zdroj:	Wireless Networks, Information Processing and Systems ISBN: 9783540898528 IMTIC
DOI:	10.1007/978-3-540-89853-5_8
Popis:	This paper describes an online isolated character recognition system using advanced techniques of pattern smoothing and direction feature extraction. The composition of direction elements and their smoothing are directly performed on the online trajectory, and therefore are computationally efficient. We compare recognition performance when direction features are formulated using smoothed and unsmoothed direction vectors. In experiments, direction features from the original pattern yielded inferior performance, while primitive sub-character direction features, using smoothed direction encoded vectors, made a large difference. The average recognition rates obtained were about 75% using original direction vectors, and about 86% using the smoothed direction vectors.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::4e07c2cc65a54d162304d7b1c7090d3f https://doi.org/10.1007/978-3-540-89853-5_8 Zobrazit plný text záznamu