University of South Australia (UniSA) scientists have developed computer vision-based baby detector software that uses a digital camera to automatically detect a baby's face in a hospital bed and remotely monitor its health.
UniSA's Javaan Chahl said tubes and other equipment can hinder computers from recognizing infants, and the system was trained on videos of babies in the neonatal intensive care unit to reliably identify their skin tone and faces.
High-resolution cameras recorded the infants, while advanced signal processing techniques extracted vital physiological data.
UniSA's Kim Gibson said using neural networks to detect babies' faces is a critical achievement for non-contact monitoring.
From University of South Australia
View Full Article
Abstracts Copyright © 2021 SmithBucklin, Washington, DC, USA