Rhino and Human Detection in Overlapping RGB and LWIR Images

Abstract: The poaching of rhinoceros has increased dramatically the last few years andthe park rangers are often helpless against the militarised poachers. LinköpingUniversity is running several projects with the goal to aid the park rangers intheir work.This master thesis was produced at CybAero AB, which builds Remotely PilotedAircraft System (RPAS). With their helicopters, high end cameras with a rangesufficient to cover the whole area can be flown over the parks.The aim of this thesis is to investigate different methods to automatically findrhinos and humans, using airborne cameras. The system uses two cameras, onecolour camera and one thermal camera. The latter is used to find interestingobjects which are then extracted in the colour image. The object is then classifiedas either rhino, human or other. Several methods for classification have beenevaluated.The results show that classifying solely on the thermal image gives nearly as highaccuracy as classifying only in combination with the colour image. This enablesthe system to be used in dusk and dawn or in bad light conditions. This is animportant factor since most poaching occurs at dusk or dawn. As a conclusion asystem capable of running on low performance hardware and placeable on boardthe aircraft is presented.