基于微惯性传感器的AR设备手势交互实现  

Implementation of MEMS-inertial-sensor-based gesture interaction function used on AR devices

作　　者：张龑 王子源 张昕倚 费骏豪 韩华[1] ZHANG Yan;WANG Ziyuan;ZHANG Xinyi;FEI Junhao;HAN Hua(School of Electronic and Electrical Engineering,Shanghai University of Engineering Science,Shanghai 201620,China)

机构地区：[1]上海工程技术大学电子电气工程学院,上海201620

出　　处：《智能计算机与应用》2024年第10期107-113,共7页Intelligent Computer and Applications

基　　金：国家自然科学基金(62103257);上海市自然科学基金(22ZR1426200)。

摘　　要：当今AR技术在用户与设备交互体验上还存在着诸如设备体积过大,设备生产成本较高,交互实时性较差的的短板,故本文创新性地提出了AR设备手势交互实现方案。该设备应用2个微惯性传感器佩戴在用户食指与拇指之上,主动测量出两手指标志点的各个运动信息,然后使用这些数据,解算出相应的检测点在某一相对固定的归一化坐标系下的空间位置;并结合基于本文提出的2套检测机制所构建的多种可识别标志点运动模式,以实现手势的识别工作。最后,相应的识别结果在用户的AR显示器中予以即刻反馈,进而实现手势交互功能。性能测试实验证明,本文所提出的硬件实例方案上的AR设备手势交互实现方案在手势识别正确率、设备质量以及功耗上均具有一定优势。The current augmented reality technology has certain limitations in terms of user-device interaction experience,such as large device size,high device production costs and poor real-time interaction.This paper proposes an innovative AR device gesture interaction implementation scheme.This scheme applies two MEMS inertial sensors wearing on user′s index finger and thumb to actively measure the motion information of two reference points,and then uses these data to calculate the 3D position of the corresponding detection point in a relatively fixed geodetic coordinate system.By combining multiple recognizable gesture marker motion patterns built on two sets of software detection mechanisms based on innovative design,gesture recognition can be achieved.Finally,the corresponding recognition results will be immediately fed back in the user's AR display,thus achieving gesture interaction.Performance testing shows that compared with existing theories and technologies,the AR device gesture interaction implementation scheme proposed in this paper has certain advantages in terms of gesture recognition accuracy,device quality,and power consumption on the proposed hardware instance.

关 键 词：手势交互 微惯性传感器 增强现实技术 特征提取 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]