论文部分内容阅读
针对离线手写体文本依存的笔迹鉴定研究,提出了基于改进多数规则的笔迹鉴定信息融合算法,将多种笔迹鉴定算法结果进行融合,并提出在判定的同时给出两份笔迹的相似程度的思想,给出了相似程度的计算方法。融合的结果和笔迹的相似程度共同输出,辅助文检专家进行判定。文检专家可根据实际应用的要求情况和对系统的信任度做出最后判定。实验结果表明,运用本文的融合算法,在融合后笔迹鉴定准确率有大幅的提高。
Aiming at the research of handwriting recognition based on offline handwritten texts, this paper proposes a handwriting recognition information fusion algorithm based on the improved majority rule, combines the results of various handwriting recognition algorithms, and presents the idea of similarity of two handwriting at the same time, Gives a similar degree of calculation method. The results of the fusion and the similarity of handwriting output together to assist the inspection specialist to determine. Literary experts according to the actual application requirements and trust in the system to make the final decision. The experimental results show that the accuracy of handwriting identification after fusion is greatly improved by using the fusion algorithm in this paper.