Automatic eye corners detection and tracking algorithm in sequence of thermal medical images

The article presents automatic eye corners detection algorithms in thermal images. Its target application is to perform quick and unnoticed measurement of human body temperature. It is proved that the temperature of eyes’ corners is the most reliable and stable temperature considering infrared imaging. That measurements were done manually so far. Our approach is to do this automatically and create complete system for measurement of core human body temperature in crowded places where it is impossible to do this in another way (for example on the airport, railway station). Such system could prevent people for spreading off the epidemic. Two proposed algorithms are presented: first based on morphological operations and geometric features of human face, second based on the cross-correlation and idea of pattern tracking. The selection of appropriate ROI size for reliable temperature extraction was tested according to the distance to person under observation.