Processing of thermal images oriented to the automatic analysis of hand thermoregulation

Abstract

In this work we propose a methodology for segmenting and labeling regions of interest (ROI) corresponding to the palm and the five from infrared thermal images of the hand, acquired from a immersion test. The segmentation relies on measuring distances between the pixels in the contour of the hand and its center of mass, and involves the consideration of very few free parameters. The results are promising and suggest that the proposed methodology could serve as a component of an automatic scheme for the analysis of the hand thermoregulation.