Unmanned Aerial Vehicles have shown potential to improve time-critical first aid deliveries in both military and civil applications. But finding a precise final delivery location in unprepared sites with broken-down infrastructure, still poses a significant challenge. During the cargo-handling phase at low altitude, moving obstacles and smaller sensor footprints can quickly lead to hazardous situations. To address this, a novel hybrid sensing approach without required infrastructure on the receiving end is proposed. It relies on lidar-based delivery site evaluation and recognition of hand signals by human recipients on-site. Two delivery position candidates are obtained and compared via the multi-attribute decision making method TOPSIS. To validate the resulting decision-making, a first-aid delivery task into a post-earthquake settlement is simulated with AirSim in Unreal Engine. The goal is to use the synergy of human-drone interaction and automatic ground topology assessments to increase the number of successful deliveries.