UAV Based Search and Rescue with Honeybee Flight Behavior in Forest

In this paper, we present a search-and-rescue maneuver algorithm for UAV using bio-inspired honeybee natural behavior. The system is designed for condition when the target is located in unreachable and dangerous area such as forest. The proposed system consists of two main features. First is distinguishing targets at different distances by monitoring the size of bounding box which produced by You Only Looked Once (YOLO). The second point is the approach movement of UAV towards detected target. The expansion size of bounding box is become the input system to generate speed control of the vehicle. As the output system, the UAV's linear and angular speed direct the vehicle close to the target location. On the other hand, YOLO is applied to recognize the number of targets which have been successfully detected in the mission. The feasibility of the proposed system is demonstrated using unmanned aerial vehicle in the real environment. The experiment results show that the system has performed with approaching time 34 seconds and 83% accuracy.