MSUToday
Published: June 17, 2019

Compact, low-cost fingerprint reader could help reduce infant mortality around the world

Contact(s): Pat Mroczek , Anil Jain College of Engineering office: (517) 355-9282 jain@cse.msu.edu

A team of Michigan State University researchers have created Infant-Prints – a low-cost, high-resolution and portable solution to accurately identify infants in an effort to help reduce infant mortality around the world.

University Distinguished Professor of Computer Science and Engineering Anil Jain is leading a group of scientists from MSU and Dayalbagh, India, that will introduce Infant-Prints at two international conferences. (See Infant-Print video.)

They presented at the 2019 Computer Vision for Global Challenges, sponsored by Facebook AI and in conjunction with the IEEE Conference on Computer Vision and Pattern Recognition, in Long Beach, California.

Additionally, Jain will chair an academic session on June 19 on Infant Biometrics at ID4Africa in Johannesburg, South Africa. ID4Africa is part of Africa’s ID-4-All movement to develop digital identity ecosystems for improved humanitarian action and security.

Jain said Infant-Prints is a complete infant fingerprint recognition system with a low-cost, custom, high-resolution fingerprint reader, a new high-resolution, texture-based infant fingerprint matching algorithm and an Android application for verifying an infant’s identity in real-time. 

Children born in developing countries are subject to a lack of proper record keeping that prevents urgently needed health care and nutrition supplements to reduce infant mortality. By the age of five, more than 5 million children each year lose their lives to vaccine-preventable diseases alone, Jain said.

The lack of any official identification documentation makes it exceedingly difficult to identify and track at-risk infants. The complete system of a high-resolution (1900 ppi), low-cost ($85) fingerprint reader and matcher running on a mobile application allows for accurate record keeping and proper delivery of vaccination, health care and nutrition to infants. A biometric-based infant ID system is of immense interest to all the NGOs, international agencies and governments in least developed and developing countries around the globe, Jain said. 

“Our goal is to transfer the Infant-Prints prototype to an organization for extensive evaluation and deployment in least developed and developing countries,” he said. “Accurate and reliable recognition of infants will provide for effective delivery of critical vaccinations and nutritional supplements in time.”

This next-phase work improves on earlier iterations by focusing on enrollment fingerprints of infants ages 0 to 3 months and then recognizing them after up to a year later.

Jain’s research team includes MSU doctoral students Joshua Engelsma and Debayan Deb. Anjoo Bhatnagar, at Saran Ashram Hospital, and Prem Sudhish, at Dayalbagh Educational Institute (Agra, India), also contributed to this work.

For more on Infant-Prints, read the abstract.