Writer Recognition by Using New Searching Algorithm in New Local Arc Method

Abstract

In the previous studies, authors proposed a new local arc method with new similarity evaluation function to the off-line writer recognition, and are obtaining high recognition ratios. However, to calculate the similarity value of the author from the product of the similarity values of three or four kinds of characters, the difference of author similarity value becomes sometimes quite small. Then, it may not be stable recognition method. In this study, the new local arc method was modified to give more information of writers. And by examining the characteristic of the feature of new local arc method, it was found that the feature of each writer was possible to be extracted. Especially, even when the difference between author similarity values is small, stable writer recognition becomes possible.