Neonatal Biometrics
Information Processing Research Group
Federal University of Technology - Paraná
Information Processing Research Group
Federal University of Technology - Paraná
Currently, there are over 663 million children between 0 and 5 years old worldwide, with approximately 385,000 births daily—most occurring in impoverished regions with limited access to official identification documents. The absence of reliable biometric identification from birth exposes millions of children to various forms of vulnerability. Among the primary issues are infant switching in maternity wards, child abduction, and trafficking—a crime affecting more than 40 million people globally—as well as exclusion from essential systems such as healthcare, vaccination, education, and public safety. Without a verifiable ID, it becomes impossible to ensure vaccination tracking, fair food distribution in critical regions, and proper monitoring of child development, further worsening social inequalities and the risk of basic rights violations.
Although documents such as identity cards (RG) and individual taxpayer numbers (CPF) are widely used, they fail to meet the identification needs of newborns and infants, as they rely on variable physical characteristics like photos and fingerprints, which change significantly in the early years of life. Fingerprints, in particular, have been the primary method of civil identification since the early 20th century, supported by consolidated databases in many countries. However, applying them to neonates presents technical and biological challenges: fingerprints are extremely small, and their patterns change with growth, making simple periodic re-collection unviable. Current facial and iris recognition technologies also lack high accuracy for this age group. This research seeks to determine whether newborn fingerprint patterns persist over time and can be correlated with those of adults—a finding that would enable a continuous biometric identification system from the first days of life, requiring high-resolution sensors and advanced image processing techniques.
The project is developed by the Federal University of Technology – Paraná (UTFPR) at the Pato Branco campus. The project team is primarily affiliated with the Department of Informatics (DAINF), comprising over 15 professors and more than 30 high school, undergraduate, master’s, doctoral, and postdoctoral students. The project's main corporate partner is Natosafe/Infant ID. Additionally, data collection is being conducted at São Lucas Hospital with support from the Municipal Secretariat of Science, Technology, and Innovation of Pato Branco.
On the academic front, the group maintains partnerships with the Western Paraná State University (Unioeste), the UNIDEP University Center, and the Center for Artificial Intelligence at the University of São Paulo (USP). Internationally, the project collaborates with the Indian Institute of Technology Jodhpur and the National Institute of Technology Goa, which provided infant fingerprint images collected in India. Our students are part of the Graduate Program in Electrical and Computer Engineering (PPGEEC), Electrical Engineering and Industrial Informatics (CPGEI), and Computer Science (PPGCC).
Our funding sources include UTFPR itself and the company InfantID. Additionally, there is public funding from the federal government through the Ministry of Science, Technology, and Innovation, specifically from the National Council for Scientific and Technological Development (CNPq). Furthermore, funding is provided by the State Government of Paraná through the Secretariat of Science, Technology, and Higher Education (SETI) via the Regional, Sustainable Development, and Innovation Agency (AGEUNI).
Through a systematic literature review encompassing more than 5,000 studies from 1899 to 2025, the research group identified only 45 relevant to the infant biometrics scenario. Based on these studies, over 30 research directions were highlighted as promising for refining newborn biometric solutions.
The research group meets weekly with its partners to advance research centered on several key fronts:
Infant fingerprint collection is currently conducted at São Lucas Hospital in Pato Branco and in the homes of participating families.
Collected fingerprints are analyzed by the team to evaluate their quality. A new method for assessing infant fingerprint quality is currently being proposed.
Multiple research streams focus on enhancing collected fingerprints. This enhancement includes image composition of the same finger, resolution super-resolution, noise removal, and segmentation, among others.
Feature extraction from the collected fingerprints is another primary focus of the research group.
In the biometric system's most crucial stage—matching, where the infant is actually recognized—innovative and milestone research is underway, including the use of graph-based fingerprint representations.
A low-cost prototype leveraging a standard smartphone for data collection has been developed and is currently undergoing testing.
An Android mobile application is under development to streamline the collection process.
Solutions for effective fingerprint storage and retrieval are also actively being developed.
Cutting-edge research aimed at synthetically aging fingerprints is in the development and implementation stage.
Public policy contributions are also being made, including proposals for ISO protocol updates, fingerprint collection tutorial demonstrations, and recommendations for policy changes in health and public safety.