A Gesture Recognition and Drone Control System based on Residual Neural Network
Rui Wang, Chengjiao Sun · 2023
To achieve visual communication between Drone and operator, a static gesture recognition and Drone control system based on residual neural network is designed. The system segments static gestures through YCrCb colour space and elliptical skin colour detection model, extracts gesture features using the residual neural network ResNet34 and uses deep learning techniques to realise the combination of gesture recognition and UAV control. The system uses the gesture images taken from the UAV viewpoint as the dataset and will send the corresponding control signals to the UAV side according to the calibrated control commands to achieve stable control of the UAV, and the classification accuracy is used as an evaluation index to assess the performance of the model of this system in gesture recognition. Finally, the control performance of the system was verified in a real flight in a Tello vehicle platform, and the experimental results showed that the average accuracy rate could reach 90.43%.