Abstract: One of the keys to the generalization of UAVs is the development of a natural and intuitive interaction method. Gesture, as one of the most common communication methods in our daily life, has become the focus of research on UAV human-computer interaction. Focusing on wearable gesture sensing and recognition methods, this review analyzed the main gesture data acquisition methods such as electromyography, stress-strain, motion, ultrasound, and optoelectronic sensing, and proposed data processing and recognition algorithms for dynamic and static gestures. In addition, this review discussed the application of gesture recognition technology in real-time obstacle avoidance, path planning and trajectory tracking for UAVs. Finally, the key issues facing current gesture recognition technologies, including pervasiveness, robustness and realtime performance, were summarized, and future directions for wearable gesture recognition technologies were discussed. The close integration with biosensing technology, edge computing, cloud computing, reinforcement learning, adaptive learning, and multimodal data fusion, can drive gesture recognition technology toward higher accuracy, more natural interaction, and a wider range of application areas.