Speed, accuracy, and flexibility are crucial to the practical use of on-lin
e handwriting recognition. Besides, extensibility is also an important conc
ern as we move from one domain to another which requires the character set
to be extended. In this paper, we will propose a simple yet robust structur
al approach for recognizing on-line handwriting. Our approach is designed t
o achieve reasonable speed, fairly high accuracy and sufficient tolerance t
o variations. At the same time, it maintains a high degree of reusability a
nd hence facilitates extensibility. Experimental results show that the reco
gnition rates are 98.60% for digits, 98.49% for uppercase letters, 97.44% f
or lowercase letters, and 97.40% for the combined set. When the rejected ca
ses are excluded from the calculation, the rates can be increased to 99.93%
, 99.53%, 98.55% and 98.07%, respectively. On the average, the recognition
speed is;about 7.5 characters per second running in Prolog on a Sun SPARC 1
0 Unix workstation and the memory requirement is reasonably low. With this
simple yet robust structural approach, we already have an effective and eff
icient on-line character recognition module. This module will be used as pa
rt of a larger system, a pen-based mathematical equation editor, which is b
eing developed by the authors using a syntactical pattern recognition appro
ach. (C) 1999 Pattern Recognition Society. Published by Elsevier Science Lt
d. All rights reserved.