Improvement of identification algorithms in the freehand curve identifier FSCI

摘要

As a freehand figure recognition engine, we have already proposed a freehand curve identifier FSCI based on the fuzzy theory. We have also studied methods to improve recognition rates of FSCI by optimizing FSCI's parameters using some learning algorithms. This paper newly proposes two methods to improve FSCI's identification algorithms; one is 'improvement of formulation of hypothetical fuzzy models" and the other is "introduction of necessity into fuzzy inference rules." Experimental results show the proposed methods go a long way toward solving such problems that the past parameter optimization methods do not answer to.