One of the devices used to realize cyber-physical systems is a drone, such as an unmanned aerial vehicle (UAV). Drones are increasingly being applied to infrastructure inspections over wide areas and in challenged communication environments, as well as to equipment monitoring for rapid fault detection and prediction. Drone flight control has generally been performed by a human operator or based on predefined settings. However, research on autonomous flight control is progressing as image recognition technology improves. From a practical standpoint, hybrid control - i.e., fine control based on autonomous decisions while following simple instructions via gestures - is necessary. While it is becoming possible to recognize a single gesture, it remains difficult to accurately identify multiple consecutive gestures. In this paper, we propose a method for controlling drone flight by recognizing multiple consecutive gestures, and present the results of its implementation.