A Comprehensive Survey of Pattern Recognition, Deep Learning, and Wearable Technology Approaches for Secure Signature Verification Systems

Abstract

Automatic signature verification is a complex biometric authentication problem because signatures involve intent, distinct writing styles, visual texture, and motor dynamics. This review consolidates signature verification work involving offline, online, mobile, contactless, and in-air verification. The reviewed works show a clear shift from traditional classifiers and handcrafted descriptors toward deep metric learning, convolutional feature extractors, capsule models, and recurrent temporal models. Stronger image-level representation does not mean deployment-ready security. There are still many modern cyberattacks and challenges related to skilled forgery, replay attacks, presentation attacks, sensor variability, demographic and cross-linguistic transfer, data scarcity, and privacy-preserving learning. Thus, the paper analyzes those challenges and presents an original, practical taxonomy for over 50 selected works and standards. It also outlines available datasets and measures of performance and suggests a set of best-case solutions in the context of a given operational constraint, an expanded wearable-technology analysis, and a trustworthiness framework for explainable and auditable biometric decision-making. The future works for secure biometric services are the integration of advanced computer vision, calibrated and reliable security thresholds, a range of methods for detection of presentation attacks, assurance of cancellability, continuous post-deployment assessment, and privacy are some key components. Received: 24 May 2026 | Revised: 23 July 2026 | Accepted: 11 August 2026 Conflicts of Interest The authors declare that they have no conflicts of interest to this work. Data Availability Statement Data sharing is not applicable to this article as no new data were created or analyzed in this study. Author Contribution Statement Amal Hameed Khaleel: Conceptualization, Methodology, Validation, Formal analysis, Investigation, Resources, Data curation, Writing – original draft, Writing – review & editing, Visualization, Supervision, Project administration. Suhad Muhajer Kareem: Conceptualization, Methodology, Validation, Resources, Writing – review & editing.