Unmanned Aerial Vehicles (UAVs) provide an opportunity to make 24/7 surveillance possible in large areas of strategic importance such as industrial plants and borders. In this paper we present an algorithm for detecting people in surveillance areas in low visibility conditions. The efficiency of the current solution is based on using an onboard thermal camera as a sensor, which generates a heat map even at nighttime and in bad weather. UAV flight details and technical characteristics are also discussed.