Handwriting Verification Based on Fusion of Feature Extracting Algorithms

摘要

The handwriting verification information fusion algorithm based on advanced majority rule was put forward.Several methods of the off-line,text-dependent writer verification were combined with that fusion algorithm,and the ideal of giving the similarity degree of two handwritings while judging was proposed.We also give calculation method of the similarity degree.The result of combination and the similarity degree of two handwriting images can assist document examiners in making determination.Document examiners can make the final determination according to the request of specific application and the confidence degree of the system.Experimental results indicate that compared with current methods,verification accuracy is greatly raised with this approach.