Automatic Detection of Civilian and Military Personnel in Reconnaissance Missions using a UAV

In the last few years, Unmanned Aerial Vehicles (UAVs) have proven to be an invaluable tool in Intelligence, Surveillance, and Reconnaissance (ISR) missions, changing paradigms regarding military operations in various theaters of conflict. These unmanned vehicles bring significant advantages to these operations, from direct battlefield support to increased terrain and enemy situational awareness. This work shows how it is possible to automatically detect civilian and military personnel on the ground in real-time, using a standard RGB camera integrated into a UAV and a neural network for detection. To train this network, a dataset was collected and annotated to be able to perform supervised learning. The proposed approach was experimentally validated, with the final model showing promising results and a great potential to be considered a support tool for ISR missions.