Mean-Shift Algorithm-based Human Frontal Face Tracking System Using Thermal Imaging

This paper presents the application of mean-shift algorithm that has been used to develop an automated frontal facial detection and tracking system. A thermal camera was used and mounted on a tracking system to automatically trace and capture the video frames of the frontal facial. The system is planned to be implemented and applied in the study of identification of affective states on children with Autism Spectrum Disorder (ASD). As to current method of detecting affective state of children with Autism Spectrum Disorder poses challenge due to the invasive and intrusive method that needs sensor patch on the skin. The study proposes to use thermal imaging as a non-invasive technique to detect affective states of the autistic children from the frontal facial thermal patterns. While many systems rely on face color information in visual images, the paper presents a methodology leveraging thermal images. The frontal face image is tracked based on the tracking mean-shift algorithm developed in LabVIEW® vision module environment.