作者: Dong Hwa Yang , Dae-Jong Lee , Myung-Geun Chun
DOI: 10.5391/JKIIS.2005.15.1.012
关键词:
摘要: The segment matching method shows better performance than the global and points-based methods to compare reference signature with an input signature. However, segment-to-segment has problem of decreasing recognition rate according variation partitioning points. This paper proposes a fusion model based on HMM construct more reliable authentic system. First, classifier is designed by conventional technique calculate values lot dynamic information signatures. And also, novel constructed using principal component analysis for static Finally, SVM adopted effectively combine two independent classifiers. From various experiments, we find that proposed method.