Development of a handwriting expert system

OData support
Dr. Kővári Bence András
Department of Automation and Applied Informatics

Nowadays, one of the simplest physical identification method, where the subjective doesn’t even have to be present at the time of its identification, is the signature. Because of its features, it is also used by formal organizations to identify people. With our signature, we can confirm that we have seen and agreed with the signed document. This may involve several traps. To sign something, from the outside, seems very simple, so more people may think that, for example, if they sign an acknowledgment of receipt on behalf of a family member, which is received from the post office, is an innocent act, since they just want to help out a family member. What they do not consider is that it is a crime. In addition, a sample of writing contains more information than we may think.

Signing the receipt is just one example of many. There are plenty of stories like this and similar in the media. In case of an accusation, the case itself will get to the writing experts, to establish the authenticity of the signatures. They are able to evaluate the samples according to a well-developed set of criteria, from which they can make an expert evidence at the end. For a long time certifications and evaluations were written on paper, but the possibilities of today offer another solution, since paper-based information storage is now considered obsolete, and it is also difficult to handle, store, and forward.

It is my job, to provide something like an alternative solution to the writing experts instead of paper-based data management. I will implement an application that will allow me to replace all of the obsolete methods completely. I will provide opportunities for case creation, and for uploading, qualifying and saving signatures.

In this document, I summarize what technologies I used and how did I solve this problem. I will present the user interfaces, the use of possible functions, and the structure of my application.


Please sign in to download the files of this thesis.