In this paper, an invariant recognition system using a fuzzy neural ne
twork to recognize handwritten Chinese characters on maps is proposed;
characters can be in arbitrary location, scale and orientation. A nor
malization process is first used to normalize characters such that the
y are invariant to translation and scale. Simple rotation-invariant fe
ature vectors called ring-data vectors are then extracted from thinned
or non-thinned characters. Finally, a fuzzy min-max neural network is
employed to classify the ring-data vectors by means of its strong abi
lity of discriminating heavy-overlapped and ill-defined character clas
ses. Several experiments with two kinds of character sets are carried
out to analyze the influence factors of the proposed approach. The per
formances of the ring-data features and the fuzzy min-max neural netwo
rk are compared with those of moment invariants and two traditional st
atistical classifiers, respectively. The ring-data features are found
to be superior to the moment invariants, and also the fuzzy min-max ne
ural network is found to be superior to the two classifiers. However,
from the experimental results, we also see that the proposed approach
is suitable to handle the translation, scale and rotation problem, but
cannot solve the high-shape-variation problem of handwritten Chinese
characters. (C) 1997 Elsevier Science B.V.