The Research Of The Key Technologies About The Drone's Self Navigation And Human Machine Interaction In City Area

Urban flying drone is an important part of the smart cities in the future.The traditional Global Positioning System(GPS)and inertial navigation are easy to receive interference from urban constructions.The fifth generation(5G)network provides rich communication resources for autonomous navigation of drone.Moreover,to build a friendly urban flying drone system,all climate aerial interaction is important in collaboration and interaction between human and drone.The paper introduced the common transmission types of electromagnetic signal and the advanced 5G technologies.We focused on the millimeter wave and the massive Multiple-input Multiple-output(MIMO)antenna system.To achieve urban flying drone's cooperative communication and autonomous navigation,this paper presented an omnidirectional cylindrical drone's mounted antenna.Experiments proved that the antenna system can effectively receive the beamforming signal from the base station.The paper further discussed the state of art human machine interaction methods.We used solid state multipoint Time of Flight(TOF)sensors to collect gesture data for hand segmentation in complicated background.Then,we presented a gesture classification method based on deep learning.Simulation experiment shown that the gesture segmentation and classification based on the depth information was effective.We will use multipoint TOF sensors to increase the data density,to build a more stable and reliable dynamic gesture recognition system.