A Preliminary Study on Handwriting Classification Using Pen Pressure Patterns

Writing with a finger or a stylus pen on a tablet, especially in the case of signing, has been more and more popular in our ordinary life. The necessity of the analysis of handwriting data acquired online for forensic purposes has also been more and more important. There has been, however, a very small number of researches conducted on online handwriting data analysis from the standpoint of forensic document examination. Writer classification experiments using the dataset collected both offline and online for forensic purposes were conducted in this paper. Japanese handwriting samples of four categories of Kanji characters, which were acquired both online and offline, were combined to be a “signature” of a Japanese person, analyzed and compared the classification results between online and offline. Online data used for the experiments were pen tip trajectory and pen pressure whose pattern was expressed on the trajectory as a map. The classification was done using LeNet. The results showed over 95% accuracy in both online and offline conditions, which were fairly good. The results suggested the availability of online data for forensic handwriting classification.