Missing or injured persons are often searched for from helicopters during rescue missions using thermal imaging cameras. In densely forested areas, however, it is almost impossible to find them because such cameras generate images from the difference between body heat and ambient temperature. And either the vegetation obscures the underground too much or the sunlit trees have a similar temperature to the missing person.

Researchers at the University of Linz have now developed a drone that combines multiple thermal images into an integral image in real time - and the AI evaluates whether the detected object is actually a person. Since all calculations are performed in real time on the drone, it is thus able to fly fully autonomously and decide on its own flight path.

A wide range of possibilities

The development also holds significant potential for other applications - such as detecting people in rubble after earthquakes or explosions, or monitoring wildlife in forests.