Abstract: To realize the accurate description for natural hand gestures and its intention,and to achieve the natural and efficient factory layout design under large space augmented reality,a novel interactive semantic model based on data glove and magnetic trackers was proposed.The ideas of Turing machine was used for reference in model design,and the model was divided into three layers:physical layer,lexical layer and syntax layer.Physical layer was used to get hand information (included hand gesture,gesture sequence and hand motion trajectory) collected by data glove and magnetic trackers.Lexical layer was composed by basic hand gestures'information.Syntax layer was a logical reasoning layer used to mapping hand gesture information to specific operations.The proposed model laid a foundation for implementing operations (included add virtual model,move model,rotate model,delete model,et al) used in layout design based on real workshop environment.A prototype system was developed to verify the effectiveness of the proposed model.

Key words: augmented reality, factory layout, human-computer interaction, hand gesture

摘要： 为了准确描述人手自然动作及其背后隐含的操作意图,达到在大空间增强现实环境下高效、自然地进行车间布局设计的目的,提出一种基于数据手套、磁跟踪器的人机交互语义模型。该模型借鉴图灵机的思想,分成物理层、词法层和语法层三部分。其中,物理层的作用是从数据手套、磁跟踪器中获得操作者手部信息(手势、手势变化和手部运动轨迹);词法层由可组合成各种不同操作意图的基本手势信息组成;语法层则是根据手部信息映射到具体操作意图的逻辑推理层。该模型的提出可为实现在真实车间环境中进行车间布局设计的各项操作奠定基础。通过原型系统的实际运行情况验证了所提交互语义模型的有效性。

关键词: 增强现实, 车间布局, 人机交互, 手势识别