A Two-branch Convolutional Neural Network-based Approach to Flight Action Recognition
Flight action recognition is a key technology in the fields of flight training evaluation and flight accident investiga-tion.To address the problems of weak generality,complex model and long recognition time of existing recognition methods,a flight motion recognition method based on two-branch convolutional neural network is proposed.Using the idea of motion decomposition,the aircraft motion is decomposed into horizontal and vertical motions on two branches,and the flight feature parameters on the branches are extracted,and the convolutional neural network model is trained and recognized using the real flight parameters,and then the motions on these two branches are synthesized to realize the flight motion recognition.The final experimental results show that the use of this method simplifies the recognition process,the recognition model is more general,and the recognition method has a high accuracy rate.
万方数据
维普
