John Tait

John TaitJohn Tait

RMIT University


I am currently a 3rd year student in the Department of mathematics and geospatial sciences at RMIT, studying a Bachelor of Science, majoring in Mathematics.  I have won the CSIRO first year mathematics prize as well as the Citywest water work integrated learning award for project work completed in my first year.  I am interested in applied mathematics, and in particular networks and optimization.

Ballprint identification of infants: Is it possible?

The “Happy feet” project at RMIT aims to create reliable technology that will positively identify children from birth to two years of age through their ballprint, with the aim being to improve vaccination and immunization programs in developing countries.Registration and identification of infants is an enormous global problem. In December 2013 UNICEF estimated that nearly 230 million children under the age of 5 worldwide were not formally recognised at birth, and consequently may be denied healthcare, education and political and economic rights.Discovering a reliable biometric for infants is part of the Grand Challenge in Global Health issued by the Bill and Melinda Gates Foundation to improve uptake and coverage of childhood vaccinations. The lack of a reliable biometric has been identified as a roadblock for improving vaccination programs in developing countries since the vaccination history of a child is often unknown, and paper-based identification systems are unreliable.In 2012-13 a database of footprints and ballprints was collected from 54 babies at birth, at 2 months of age and at 6 months of age. A 4th collection is currently underway now that the babies are about 2 years old.  Positively identifying an infant or child is essential for looking up the child’s vaccination history. If doctors and other health care workers can identify the immune status of a child it would help determine if the vaccine is actually working for the target population.The fast paced growth of very young children presents a major challenge to matching biometrics captured at different ages. This project is about extracting a noise-tolerant, scale-invariant graph from ballprints that is easily matched using graph-matching and graph dissimilarity techniques with the graphs extracted from prints of the same ballprint at older ages.Children are uncooperative subjects and so ballprints suffer from distortion and noise in a similar way to latent prints taken from a crime scene. This project will work on reducing the impact of that noise on the reliability of edges in the extracted graphs; manually finding corresponding features in ballprints taken at 2 months of age and 2 years of age; and determining how the point pattern of features has changed over time.

Contact Us

We're not around right now. But you can send us an email and we'll get back to you, asap.

Not readable? Change text.