Securing Identity from Birth: Biometric Fingerprint Algorithms For Robust Childbirth Registration in Ghana

Abstract

Ghana faces persistent challenges in achieving universal birth registration, especially in rural and underserved communities. Traditional paper-based systems remain prone to loss, fraud, and inefficiencies, leaving many children without legal identity and limiting access to critical services such as healthcare and education. This study presents a biometric fingerprint-based childbirth registration system tailored for infants and mothers, designed to integrate with Ghana's national identity framework (Ghana Card). Using a convolutional neural network (CNN)-based fingerprint matching algorithm, our system achieved an identification accuracy of 86.7% for maternal-infant linking during controlled field testing in a selected Chps zone in Aburi, Akwapim South Municipality in the Eastern region in Ghana. The findings demonstrate that early-stage biometric data collection is feasible and reliable within low-resource settings. Ethical consent, data protection, and system misuse were addressed through community engagement protocols and adherence to Ghana's Data Protection Act. The results indicate that implementing a secure biometric registration system can significantly strengthen identity management in Ghana. The study’s primary contribution lies in the development and testing of a context-sensitive biometric algorithm that addresses both technological and infrastructural limitations, offering a scalable and secure model to help Ghana meet Sustainable Development Goal 16.9: ensuring legal identity for all, including birth registration.