Research on Gesture Recognition and Interaction of Virtual Collaborative Disassembly Training

Abstract

At present, most of the virtual disassembly training systems are single person training systems, and the interaction is relatively simple, which can not reflect the real natural interactive operation, so it can not meet the needs of multi-person collaborative disassembly training of complex equipment. In order to solve this problem, research on gesture recognition and interaction technology. Based on the demand analysis of disassembly training, 5 basic gestures and 3 auxiliary gestures are designed; A gesture feature extraction method suitable for immersive virtual disassembly training environment is proposed; The one-handed interaction object model and two-handed interaction object model are proposed, and the interactive hardware experimental platform is designed to realize the multi-person cooperative operation. The research content is applied to the virtual disassembly of the manual gearbox of a car, and the operation evaluation of the virtual disassembly training is carried out. The results show that: the gesture designed meets the requirements of common virtual disassembly and assembly operations, and the gesture and interaction intention are identified accurately, which can realize the interactive operation required by disassembly and assembly, and restore the real natural interactive operation to a certain extent.