IMU

A single 6-axis IMU (3-axis gyroscope plus 3-axis accelerometer, no magnetometer) sits inside the back-of-hand module, with two subscription APIs:

API	Data	`orientation`
`glove.imu_palm()`	Raw data: angular velocity + linear acceleration	Unavailable (`orientation_covariance[0]` = -1, see REP-145)
`glove.imu_data_palm()`	Fused data: raw data + complementary-filter orientation	Available

# Fused data: with orientation
sub = glove.imu_data_palm().subscribe()
imu = await sub.recv_async()
print(f"Orientation: {imu.orientation}")        # quaternion (x, y, z, w)
print(f"Angular velocity: {imu.angular_velocity}")    # rad/s
print(f"Linear acceleration: {imu.linear_acceleration}") # m/s²

# Raw data: angular velocity and linear acceleration only
sub = glove.imu_palm().subscribe()
raw = await sub.recv_async()
print(f"Angular velocity: {raw.angular_velocity}, linear acceleration: {raw.linear_acceleration}")

The fused orientation comes from an SDK complementary filter that starts when the first subscriber opens the stream. Roll and pitch initialize from the first accelerometer frame against gravity and stay stable long term, while yaw has no absolute reference — it zeroes at stream start and drifts slowly over time. Yaw resets only after every subscriber, including indirect ones such as tf, has unsubscribed and the stream restarts.

IMU data is compatible with ROS sensor_msgs/Imu and can be plugged directly into a ROS integration. Both streams can be recorded to MCAP files with TopicRecorder. See the SDK Data Recording page.

ImuData

IMU sensor data, compatible with ROS sensor_msgs/Imu.

Field	Type	Description
`header`	FrameHeader	Frame header
`orientation`	Quaternion	Orientation quaternion. Fused value on `imu_data_palm()` only (`orientation_covariance[0]` = 0) — on `imu_palm()` it is the identity quaternion with `orientation_covariance[0]` = -1 marking it unavailable (REP-145)
`orientation_covariance`	`List[float]`	Orientation covariance matrix (3×3, row-major)
`angular_velocity`	Vector3F64	Angular velocity (rad/s)
`angular_velocity_covariance`	`List[float]`	Angular velocity covariance matrix
`linear_acceleration`	Vector3F64	Linear acceleration (m/s²)
`linear_acceleration_covariance`	`List[float]`	Linear acceleration covariance matrix

Sample raw frame from imu_palm():

{
  "header": {
    "seq": 500,
    "timestamp_us": 1709876543210,
    "frame_id": ""
  },
  "orientation": {
    "x": 0.0,
    "y": 0.0,
    "z": 0.0,
    "w": 1.0
  },
  "orientation_covariance": [-1, 0, 0, 0, 0, 0, 0, 0, 0],
  "angular_velocity": {
    "x": 0.01,
    "y": -0.02,
    "z": 0.005
  },
  "angular_velocity_covariance": [0, 0, 0, 0, 0, 0, 0, 0, 0],
  "linear_acceleration": {
    "x": 0.1,
    "y": -0.05,
    "z": 9.8
  },
  "linear_acceleration_covariance": [0, 0, 0, 0, 0, 0, 0, 0, 0]
}

Python
C

Python: See the Python API examples earlier on this page.

The C SDK provides typed subscriptions for palm raw IMU and fused IMU data.

wuji_glove_subscribe_imu_palm(dev, on_imu, user_data, &imu_sub);
wuji_glove_subscribe_imu_data_palm(dev, on_imu_data, user_data, &imu_data_sub);

Frame pointers are valid only during the callback. Call wuji_sub_close when the subscriptions are no longer needed.

Overview

Hardware

Software

More

ImuData

​ImuData

ImuData