基于VR设备中IMU的无手交互方法研究

Hands Free Interaction Algorithm based on an IMU in Visual Reality Device

摘要:

随着头戴式显示设备的发展,在基于虚拟现实(Virtual Reality,VR)的教育培训中,存在用户与设备间进行交互的场景。针对用户与VR视频中对象的模拟靠近与躲避问题,提出了姿态感知与步态识别相结合的方法。通过姿态感知算法解算用户头部姿态,通过步态识别算法识别出人体的静止与行走状态,进而在行走状态时,将计算所得航向角和固定步长代入三角函数公式进行位置更新,在静止状态时,保持位置不变。实验证明,提出的姿态感知算法可以有效的计算出使用者头部的姿态,与商用惯性测量单元提供的姿态角相比具有1.1×10_^-2的平均姿态偏差；提出的步态识别算法可以有效的识别出人体的静止与行走状态；所提出的两者结合的交互方法,可以有效的实现虚拟的靠近与躲避。

Abstract:

With the development of head-mounted display devices, education and training based on virtual reality (Virtual Reality, VR) have become popular. For education and training based on VR devices, there are interaction behaviors between users and devices. For the issue that users need to approach and avoid objects in the VR video, an interactive method combining attitude sensing and gait recognition is proposed. The head attitude of the user is calculated by the attitude sensing algorithm, while the motions of the human including stillness and walking are recognized through the gait recognition algorithm. In the walking state, the calculated heading angle and the fixed step length are substituted into the trigonometric function for position updating. In the still state, the position stays unchanged. Several experiments show that the proposed attitude algorithm can effectively calculate the pose of the user"s head, and has an overall average deviation of 1.1 × 10^-2 degrees from the attitude angle computed by a commercial IMU. The proposed gait recognition algorithm can effectively recognize the stillness and walking states. Moreover, the proposed interactive method can effectively achieve the simulated approach and avoidance.