IMU-Based Motion Capture Data for Various Walking Tasks

This dataset contains motion capture data collected from 11 healthy subjects performing various walking tasks using IMU-based sensors. Each subject performed 8 different tasks under the following conditions:1. Normal walk2. Fast walk3. Normal walk while holding a 1 kg weight with the dominant hand4. Fast walk while holding a 1 kg weight with the dominant hand5. Normal walk with a knee brace on one leg6. Fast walk with a knee brace on one leg7. Normal walk with a knee brace while holding a 1 kg weight (a combination of Task 3 and Task 5)8. Fast walking with a knee brace while holding a 1 kg weight (a combination of Task 4 and Task 6)<br><b>Data Collection</b>The data was collected using a commercial IMU-based motion capture system, with 10 modules worn on the following body parts:- Left foot- Right foot- Left shank- Right shank- Left thigh- Right thigh- Left arm- Right arm- Trunk- Pelvis<br>Each module recorded the following data along X, Y, and Z axes:- Accelerometer data- Gyroscope data- Magnetometer data<b>Sampling Rate:</b>- The data sampling rate is 4 ms for all subjects except sub_01 and sub_03, where the sampling rate is 6 ms.- In certain rows of the files, there are irregularities in the recorded time. This occurs when the time value reaches 65,535 or multiples of it (e.g., 131,070 and 196,605). This problem is associated with the way time is displayed and does not impact the sample rate.<b>Data Structure</b>The dataset is organized into the following folders for each subject.<br>Subject folders:- sub_01- sub_02- sub_03- sub_04- sub_05- sub_06- sub_07- sub_08- sub_09- sub_10 (Note: Task 2 is missing from sub_10 folder)- sub_11<br>Task folders within each subject’s folder include:- 1_walking normal- 2_walking fast- 3_weight normal- 4_weight fast- 5_brace normal- 6_brace fast- 7_brace weight normal- 8_brace weight fast<br>Each trial folder contains four CSV files, named according to the trial condition. For example, for the first trial, the files are:- walking normal_Raw.csv- walking normal_Processed.csv- walking normal_Euler.csv- walking normal_JointsKinematics.csv<b>CSV File Descriptions</b>1. Raw: Contains raw sensor data.- Time (ms)- Accelerometer data (X, Y, Z)- Gyroscope data (X, Y, Z)- Magnetometer data (X, Y, Z)<br>2. Processed: Contains preprocessed data.- Time (ms)- Quaternion components (Q0, Q1, Q2, Q3)- Acceleration in IMU coordinate system (X, Y, Z)- Linear acceleration without gravity (X, Y, Z)- Acceleration in the global coordinate system (X, Y, Z)<br>3. Euler: Contains Euler angles.- Time (ms)- Roll, Pitch, and Yaw angles<br>4. Joints Kinematics: Contains joint angle data.- Time (ms)- Abduction-Adduction angle- Internal-External Rotation angle- Flexion-Extension angle<br><b>Column Labels</b><br>Raw Data File:- Time_LeftFoot, AccX_LeftFoot, AccY_LeftFoot, AccZ_LeftFoot, GyroX_LeftFoot, GyroY_LeftFoot, GyroZ_LeftFoot, MagX_LeftFoot, MagY_LeftFoot, MagZ_LeftFoot, *<br>Processed Data File:- Time_LeftFoot, Q0_LeftFoot, Q1_LeftFoot, Q2_LeftFoot, Q3_LeftFoot, Acc_X_LeftFoot, Acc_Y_LeftFoot, Acc_Z_LeftFoot, Acc_linX_LeftFoot, Acc_linY_LeftFoot, Acc_linZ_LeftFoot, Acc_GlinX_LeftFoot, Acc_GlinY_LeftFoot, Acc_GlinZ_LeftFoot, *<br>Euler Data File:- Time_LeftFoot, Roll_LeftFoot, Pitch_LeftFoot, Yaw_LeftFoot, *<br>Joints Kinematics Data File:- Time_LeftAnkle, Abduction-Adduction_LeftAnkle, Internal-External Rotat_LeftAnkle, Flexion-Extension_LeftAnkle, **<br><b>Additional Notes</b>- This dataset can be used for research in biomechanics, rehabilitation, and human motion analysis.* (Similar pattern for RightFoot, LeftShank, RightShank, LeftThigh, RightThigh, LeftHumerus, RightHumerus, Pelvic, Trunk)** (Similar pattern for RightAnkle, LeftKnee, RightKnee, LeftHip, RightHip, LeftShoulder, RightShoulder, Pelvic, Trunk2Ground)

本数据集包含来自11名健康受试者的运动捕捉数据，这些受试者使用基于惯性测量单元（IMU）的传感器完成各类行走任务。每名受试者需在以下8种不同工况下完成行走任务：
1. 正常行走
2. 快速行走
3. 优势手持1kg重物的正常行走
4. 优势手持1kg重物的快速行走
5. 单腿佩戴护膝的正常行走
6. 单腿佩戴护膝的快速行走
7. 佩戴护膝且优势手持1kg重物的正常行走（任务3与任务5的组合）
8. 佩戴护膝且优势手持1kg重物的快速行走（任务4与任务6的组合）
<b>数据采集</b>
本数据集采用商用基于IMU的运动捕捉系统采集，共10个传感器模块佩戴于以下身体部位：
- 左脚
- 右脚
- 左小腿
- 右小腿
- 左大腿
- 右大腿
- 左臂
- 右臂
- 躯干
- 骨盆
每个模块沿X、Y、Z三个轴向记录以下数据：
- 加速度计数据
- 陀螺仪数据
- 磁力计数据
<b>采样率</b>
除sub_01与sub_03的采样率为6ms外，其余所有受试者的数据采样率均为4ms。
在部分文件行中，记录的时间会出现异常：当时间值达到65535或其整数倍（如131070、196605）时会触发该问题。该现象仅与时间显示逻辑有关，不会影响实际采样率。
<b>数据结构</b>
本数据集按受试者划分至对应文件夹，具体如下：
受试者文件夹：
- sub_01
- sub_02
- sub_03
- sub_04
- sub_05
- sub_06
- sub_07
- sub_08
- sub_09
- sub_10（注：sub_10文件夹中缺失任务2的数据）
- sub_11
每个受试者文件夹下包含任务文件夹，具体为：
- 1_walking normal → 1_正常行走
- 2_walking fast → 2_快速行走
- 3_weight normal → 3_负重正常行走
- 4_weight fast → 4_负重快速行走
- 5_brace normal →5_护膝正常行走
- 6_brace fast →6_护膝快速行走
-7_brace weight normal →7_护膝负重正常行走
-8_brace weight fast →8_护膝负重快速行走
每个任务试验文件夹内包含4个CSV格式文件，文件名对应试验工况。例如，首个试验的文件名为：
- walking normal_Raw.csv → 正常行走_原始数据.csv
- walking normal_Processed.csv → 正常行走_预处理数据.csv
- walking normal_Euler.csv → 正常行走_欧拉角数据.csv
- walking normal_JointsKinematics.csv → 正常行走_关节运动学数据.csv
<b>CSV文件说明</b>
1. 原始数据（Raw）：包含传感器原始采集数据
- 时间（ms）
- 加速度计数据（X、Y、Z轴）
- 陀螺仪数据（X、Y、Z轴）
- 磁力计数据（X、Y、Z轴）
2. 预处理数据（Processed）：包含经过预处理的传感器数据
- 时间（ms）
- 四元数分量（Q0、Q1、Q2、Q3）
- 传感器坐标系下的加速度（X、Y、Z轴）
- 无重力影响的线加速度（X、Y、Z轴）
- 全局坐标系下的加速度（X、Y、Z轴）
3. 欧拉角数据（Euler）：包含欧拉角数据
- 时间（ms）
- 滚转角、俯仰角与偏航角
4. 关节运动学数据（Joints Kinematics）：包含关节角度数据
- 时间（ms）
- 外展-内收角度
- 内-外旋转角度
- 屈曲-伸展角度
<b>列标签说明</b>
原始数据文件：
- Time_LeftFoot、AccX_LeftFoot、AccY_LeftFoot、AccZ_LeftFoot、GyroX_LeftFoot、GyroY_LeftFoot、GyroZ_LeftFoot、MagX_LeftFoot、MagY_LeftFoot、MagZ_LeftFoot，其余列遵循相同命名规则（对应右脚、左小腿、右小腿、左大腿、右大腿、左臂、右臂、骨盆、躯干）
预处理数据文件：
- Time_LeftFoot、Q0_LeftFoot、Q1_LeftFoot、Q2_LeftFoot、Q3_LeftFoot、Acc_X_LeftFoot、Acc_Y_LeftFoot、Acc_Z_LeftFoot、Acc_linX_LeftFoot、Acc_linY_LeftFoot、Acc_linZ_LeftFoot、Acc_GlinX_LeftFoot、Acc_GlinY_LeftFoot、Acc_GlinZ_LeftFoot，其余列遵循相同命名规则
欧拉角数据文件：
- Time_LeftFoot、Roll_LeftFoot、Pitch_LeftFoot、Yaw_LeftFoot，其余列遵循相同命名规则
关节运动学数据文件：
- Time_LeftAnkle、Abduction-Adduction_LeftAnkle、Internal-External Rotat_LeftAnkle、Flexion-Extension_LeftAnkle，其余列遵循相同命名规则（对应右脚踝、左膝、右膝、左髋、右髋、左肩、右肩、骨盆、躯干-地面）
<b>补充说明</b>
本数据集可用于生物力学、康复医学以及人体运动分析相关研究。