Design of a Gesture-Controlled Multi-Blimp System

Recent advances in human-swarm interaction (HSI) have highlighted the critical need for safe, intuitive control of aerial robot collectives in human-populated environments. This paper presents a monocular vision-based control framework for a human-multiple-blimp collaboration system. Our approach eliminates the need for wearable devices or external tracking systems by integrating three key components: (1) real-time hand recognition; (2) leader's occlusion-aware control based on Virtual Force Fields (VFF); and (3) distributed Artificial Potential Field (APF) optimization. The system enables intuitive control via natural gestures while maintaining formation stability in visual occlusions between blimps. We implement and validate the entire system in the physical environment.