基于大数据的虚拟实验室手势交互感知仿真

Virtual Laboratory Gesture Interaction Perception Simulation Based on Big Data

鉴于人机交互手势感知的重要意义和实际价值及当前相关研究成果存在的感知假阳性率较高等问题,提出基于大数据的虚拟实验室手势交互感知方法.虚拟实验室手势,形成交互感知模型,同时引入包含各种类型人体模型和姿势的数据集CAESAR,以此构建手势大数据集合.分析目标手势基本特征,将手势的图像颜色直方图转换成颜色概率分布形式,构造能够描述手势特征的图像颜色直方图,完成课题研究.实验结果表明,所提方法手势感知连续性能强,在比较短的移动距离内就能够进行手势感知,且失效率要远低于文献成果,具有可靠性和实际应用价值.

Due to the importance and practical value of human-computer interaction gesture perception and the high false positive perception rate of current research results, this article puts forward a method of gesture interaction perception in virtual laboratory based on big data. Firstly, the virtual laboratory gesture was used to build the interactive perception model. Meanwhile, the data set CAESAR was introduced to construct the big data set of gesture, including various human models and gestures. Then, the basic features of target gesture were analyzed and the image color histogram of gesture was transformed into the form of color probability distribution. Finally, the color histogram describing the gesture features was constructed, and thus to complete this research. Simulation results show that the proposed method has a strong continuity of gesture perception, which can be used for gesture perception within short moving distance. Meanwhile, the failure rate is much lower than that of the literature result. It has better reliability and application value.