DigiSign: A Mobile Application for Comparing Digitized Signature Using Feature Detection and Matching Algorithms

A signature is a formal declaration of a person’s legal identity but can be copied and forged by others. Fake signatures in online transactions, including shipping, purchasing, and banking, can result in financial losses. This study examines how digitized signatures comparing systems work on touchscreen-based mobile devices. DigiSign is a mobile application designed to detect and match digitized signatures to determine whether they are similar or not. The application employs an image processing algorithm to transform digitized signatures into images, enabling the detection of keypoints through feature detection and matching to assess their similarities. By utilizing feature detection and feature matching to recognize keypoints, DigiSign ensures secure data storage in image memory or RAM, avoiding the need for storing sensitive information in databases. DigiSign utilized the BRISK algorithm and the hamming distance algorithm, resulting in a 38% acceptance rate for similarity. The application’s performance was evaluated according to ISO/IEC 25010 standards, achieving a mean score of 4.26 out of 5.0. The study confirms its effectiveness in accurately interpreting the resemblance of digital signatures, as indicated by the strongly agreed responses.