摘要随着虚拟现实技术的发展,运动姿态捕捉逐渐被应用于动画制作、康复医学、电影特技等领域。论文研究了三种基于MEMS惯性/地磁测量组件的人体运动跟踪与捕捉方法,即四元数算法、分解四元数算法(FQA)和将前两种算法结合的姿态解算方法。四元数算法是用加速计和陀螺仪测量值解算姿态信息,由于陀螺仪的误差较大,四元数算法会产生误差累积的现象,因此四元数法不适于长时间使用。FQA算法是利用三轴加速度计的测量值解算物体的橫滚角和俯仰角,然后再利用三轴磁传感器计算航向角,该算法最大的优点就是计算量较传统的四元数算法简便很多,但当载体剧烈运动时解算结果会有明显误差,因此,只适于用载体静止或缓慢移动时的姿态解算,。将FQA算法与四元数算法结合,即设定阈值来判断载体运动的剧烈程度,使得在载体静止或缓慢移动时使用FQA算法,当物体做剧烈运动时应用四元数算法。从而充分发挥两种算法各自的解算优势、避免各自的缺点,在一定程度上降低解算误差,获得更为准确的解算结果,使得解算结果比单独使用FQA算法或四元数算法都理想。6607
关键词:惯性测量单元; 磁传感器; 姿态解算; 四元数算法; 分解四元数算法;
毕业设计说明书(论文)英文摘要
Title The human body motion tracking and capturing method based on MEMS inertial/geomagnetic measurements
Abstract
Along with the development of the virtual reality technology,the human body motion tracking and capture method is gradually applied to capture animation production, rehabilitation medicine, movie stunt, etc. There human body motion tracking and capture method based on MEMS inertia/geomagnetic measurements components was proposed. There are Quaternions estimation algorithm , the factored quaternion algorithm (FQA) and the first two algorithm combining the method for the attitude.
Quaternions algorithm uses accelerometer and gyroscope measured values for calculating attitude information. Because the gyroscope's much error, Quaternions algorithm can produce error accumulation phenomenon, so quaternions method is not suitable for long time use. FQA algorithm using the three accelerometer measurement value to determination the objects of the roll and pitch Angle, then reuse three axis magnetic sensor calculation yaw Angle. The biggest advantage of this algorithm is that small amount of calculation. But it is only suitable for use carrier static or moving slow position of the solution. Combining The FQA algorithm and the quaternions algorithm which setting threshold value to judge carrier acuteness degree of motion method. It can make up for the shortcomings of each algorithm, and reduce the error.
Keywords IMU magnetic sensors attitude calculation Quaternions algorithm FQA
目 录
一 绪论 1
1.1 论文研究背景及意义 1
1.2 国内外研究现状 1
1.2.1 机械式运动捕捉 2
1.2.2 光学式运动捕捉 2
1.2.3 声学式运动捕捉 2
1.2.4 电磁式运动捕捉 3
1.2.5 本论文研究方法现状 3
1.3 本论文主要内容 4
二 基于惯性测量组件的人体姿态捕捉方法 5
2.1 惯性测量单元简介 5
2.2 基于四元数算法的姿态捕捉方法 5
2.2.1 四元数的基本概念 6
2.2.2 人体姿态捕捉中常用坐标系 8
2.2.3 四元数姿态解算方法 9
2.2.4 四元数姿态算法中初始姿态的确定 15