Gesture control of small unmanned aerial vehicle flight

The problem of constructing gesture commands for controlling a small unmanned aerial vehicle, such as a quadcopter, is considered. Commands coming from a video camera are identified by a classifier based on a convolutional neural network, and the multimodal control interface equipped with an intelligent solver converts them into control commands for the quadcopter. Neural networks from the Ultralytics neural network library allow selecting targets in a frame in real-time. The commands are sent to a specialized program on a smartphone, developed on the basis of DJI SDK flight simulators, which then sends commands via the remote control channel.The quality of recognition of developed gesture commands for DJI Phantom 3 standard edition quadcopters is investigated, and a brief guide in the form of operator work scenarios with unmanned vehicles is provided. The prospects of gesture control of several vehicles in extreme conditions have been revealed, considering the complex safety challenges of joint flight and interaction of aircraft in confined space.

Ultralytics, unmanned aerial vehicle, control, gestures, convolutional neural network, Ultralytics, intelligent interface, recognition