A natural bare-hand interface-enabled interactive AR assembly guidance
Wei Fang,Zhan Teng,Qiankun Zhang,Zhaowei Wu
DOI: https://doi.org/10.1007/s00170-024-13922-z
IF: 3.563
2024-06-13
The International Journal of Advanced Manufacturing Technology
Abstract:Augmented reality (AR) can benefit manual assembly operations by superimposing intuitive visual guidance on the workbench directly, alleviating the mental load to determine the next operation while improving assembly performance. Where the bare-hand gesture interface is considered a promising manner to manipulate the virtual assembly model, nevertheless, it usually suffers from unreliable recognition results and poor adaptability due to the unfamiliarity with gesture libraries in actual AR assembly tasks. In this paper, we try to establish a robust and easy-to-use natural bare-hand interactive AR assembly system, enabling the operator to interact against the ongoing virtual assembly models intuitively and resulting in a human-centric AR assembly performance. Firstly, a normalized feature information extraction for real-time bare-hand gestures is presented, which can establish a robust gesture model expression when encountering different sensing distances and varying operators. Followed by the definition of some typical bare-hand interaction functions related to manual assembly operations, and thus reliable interactions between hand gestures and virtual components are available. Then, a bare-hand-enabled interactive AR assembly system is established, where operators can explore and manipulate a large amount of assembly-related information with the help of real-time bare-hand interaction. Finally, extensive experiments are carried out to evaluate the proposed bare-hand interactive AR assembly system, and results demonstrate that it can alleviate the cognitive load for the operators while providing a human-centric assembly performance.
engineering, manufacturing,automation & control systems