Visual Encoding Method for MR Interface Operation Process Prompts Supporting Blind Area Assembly

Recently, research on the accuracy of visual cues in human-machine interface prompts within industrial settings remains insufficient, particularly when dealing with manual assembly tasks that face multiple physical obstructions and limited visibility. Centering on the two dimensions of visual element layout and visual spatial depth, the blind zone visual information coding model is constructed to reveal the real-time occlusion relationship between manual operational gestures and the objectives of tasks within blind zones. Leveraging the principle of “3D image geometric transformation”, this model adeptly translates intricate three-dimensional information onto a more manageable two-dimensional plane to reveal the real-time occlusion relationship between manual operational gestures and the objectives of tasks within blind zones. Through a case study experiment, the proposed visual encoding method was shown to effectively reduce the dispersion of operators' attention during assembly tasks in blind zones, compared to traditional assembly setups and visual encoding groups with digital mirror scenes. It can improve efficiency and accuracy in operators' assembly tasks in blind zones and reduce the cognitive load in understanding the intent of these tasks. Furthermore, in the aspect of enhancing user performance, the alignment of dynamic guiding cues with the intuitiveness of cognitive operations and the effectiveness of the cognitive framework of the MR interface presented within the operator's field of view are identified as two pivotal controllable factors for elevating user performance. The findings hold significant implications for guiding research in this direction.