Camera-based calibration and testing of inertial sensors in 3D joint kinematics estimation

基于摄像机的运动捕捉系统已被广泛用于量化肢体关节运动学 在实验室环境中的不同活动期间。然而，它们受到其空间要求的限制，而且 成本。近年来，惯性传感器和磁传感器的组合已被引入作为替代方案， 在真实生活环境中进行步态分析的目标。然而，没有统一的协议用于 处理惯性传感器数据，这使得系统/协议之间的比较变得困难。 Has-Motion的主要产品Visual3D是首屈一指的3D生物力学建模分析工具包， 分析和报告功能。目前，Visual3D能够处理基于相机的 运动捕捉系统，但不是惯性传感器数据。Has-Motion Inc.已确定需要增强 Visual3D能够分析由基于惯性传感器的系统收集的3D运动学数据。上一次 通过Engage Grant进行的协作已取得实质性进展，并验证了惯性的有用性 传感器使用功能校准协议估计3D关节运动学。 此Engage Plus项目是Engage项目的直接延伸。在该项目中，一个独立的简单 将在Visual3DServer中开发基于摄像机的解剖定标协议来处理惯性传感器 数据并获得身体各部分的解剖姿势。结果将通过基于相机的运动进行验证 捕获系统。另一项任务是进一步验证在Engage Grant中开发的方法 跑步任务中的仪表式跑步机。 该项目将进一步加强Has-Motion和Queen‘s之间的合作，以开发一个独立的 Visual3D中的惯性传感器系统。因此，这将直接有助于增加Has-Motion的销售 Visual3D为用户提供了基于惯性传感器的运动分析，并直接惠及加拿大经济。

Camera-based motion capture systems have been widely used for quantifying lower limb joint kinematics during different activities in a laboratory setting. However they are limited by its space requirement and high cost. In recent years, combinations of inertial and magnetic sensors have been introduced as an alternative, with the objective of performing gait analysis in a real life environment. However, there is no unified protocol for processing inertial sensor data which makes the comparison between systems/protocols difficult. HAS-Motion's primary product, Visual3D, is the premier 3D analysis toolkit for 3D biomechanics modeling, analysis, and reporting functions. Currently, Visual3D is capable to process data collected by camera-based motion capture systems, but not the inertial sensor data. HAS-Motion Inc has identified a need to enhance Visual3D's ability in analysing 3D kinematic data collected by inertial sensor-based system. The previous collaboration through an Engage Grant has made substantial progress and validated the usefulness of inertial sensors in estimating 3D joint kinematics using functional calibration protocols. This Engage Plus project is a direct extension of the Engage project. In the project, a stand-alone simple camera-based anatomical calibration protocol in Visual3DServer will be developed to process inertial sensor data and obtain anatomical poses for body segments. The results will be validated with a camera-based motion capture system. Another task is to further validate the methods developed in the Engage Grant with an instrumented treadmill during running tasks. This project will further enhance the collaboration between HAS-Motion and Queen's to develop a stand-alone inertial sensor system in Visual3D. Consequently, it will directly help to increase HAS-Motion's sale of Visual3D to the inertial sensor-based motion analysis users and directly benefit Canadian economy.