Research On The Training Method Of Locomotive Operation Gesture Based On Somatosensory Technique

With the development of China's railway transportation,more locomotive operation's training is required,among which the gesture training for the relevant signal is particularly important.In accordance with the current gesture training development situation,gesture training still depends on the traditional teaching mode.In the gesture signal training,trainers cannot get instant feedback.Based on the somatosensory technique,the standard templates were established by collecting the data,and template matching algorithm were utilized to evaluate the match result in real-time,then constructed a well interactive gesture signal training system.According to the fectures of locomotive operations,gestures were classified as two types:hand signals and driving staff gestures,the signal training system was designed.the somatosensory devices were applied to obtain the skeleton vectors as the hand signal's feature and finger joint angles and angles between the vectors of right arm as the driving staff gesture's fecture.The data obtained was processed and the standard template libraries were set up respectively.The real-time data collected by Somatosensory equipment was matched with the standard template by adopting the template matching algorithm.according to the fectures of gesture signal,different grading algorithms were chosed to evaluate the gesture training results in real time,and the scores were calculated to obtain the threshold.Finally,a locomotive gesture signal training system was designed by combining Kinect and LeapMotion somatosensory equipment with Unity3 D platform.The applicability and robustness of this gesture signal training system were verified by the experimental and shows it could be applied in practice.