XQ20 stereo headset

Demo and values

Unit AUnit B

Head-worn stereo fisheye camera with a 6-axis IMU, recording to a memory card. Borme, a brand of Shenzhen Feikasi Industrial Co., Ltd., sells it as the XQ20. Two units were calibrated.

The 3D scene MVIS recovered from unit B.

Reprojection median0.15 px

cam0–cam1 baseline60.3 mm

cam0 td+5.77 ms

cam0 lever arm36.2 mm

Reprojection is how far, in pixels, the detected corners lie from where the solve puts them. td is the camera–IMU time offset, and the lever arm is the camera–IMU position offset.

No independent reference. These values show what MVIS solved, not how close they are to the truth.

Calibration process

The path of the rig around the board, as MVIS solved it. To explore it:

• Press Play or drag the bar to follow the path.
• Use the view buttons to change the view.
• Pick a camera card to put that camera on the frustum, the pyramid that shows its view.

Each card draws the board from the solved pose, with the corners MVIS detected. We publish no camera images.

Solved rig

The solved rig in 3D, to scale. Tap a sensor to see its values. In full screen you can rotate and zoom. Grid squares are 20 mm.

3DFrontTopSide+−

To scale, from above and the front

From above

From the front

Camera	Lever x, y, z (mm)	Distance (mm)	log(R) (rad)	td (ms)	Field of view
cam0	−29.8, −9.8, +18.0	36.2	[−0.007, −0.003, +0.003]	+5.77	122° × 91°
cam1	+30.5, −9.7, +17.6	36.6	[+0.002, +0.003, +0.003]	+5.75	122° × 91°

Solved by MVIS. Lever arms (camera–IMU position offsets) are in the IMU frame. Rotation is log(R), camera in the IMU frame.

Specifications

From the supplier's sheet
Cameras	2 × 1920×1200, H.265 MP4
Frame rate	30 or 60 fps on the sheet; about 30 Hz in these recordings
Field of view	123.6° on the sheet; axis not stated
IMU	6-axis, 300 or 600 Hz on the sheet; 300.5 Hz measured
Storage and interface	microSD up to 128 GB, or UVC; Type-C
Power	No battery
Weight	180 g with strap
From the seller's listing
Model number	XQ20
Cameras	2, dual 1080P, 2 MP, MP4
Lens	Fisheye
Frame rate	60 and 30 fps
Field of view	160°; axis not stated
Shutter	Global; not measured in these recordings
IMU	300 Hz; stamped at measurement end, 50 ppm, 32-bit µs
Measured in these recordings
Calibrated field of view	cam0 122° × 91° · cam1 122° × 91°
Not published
Open	Sensor part numbers

Dataset

Recordings	Unit A: 158 s, 4,736 frames. Unit B: 119 s, 3,573 frames
Format	MP4 plus per-segment IMU CSV, converted to EuRoC layout
Board	AprilGrid 6×6, tag 35.2 mm, spacing 0.3
Licence	CC BY-NC 4.0; commercial use by request
Hosted	Calibench community, xq20/ (access by automatic approval)

AprilGrid 6 × 6, 264 × 264 mm. Standard pattern, to scale

Detailed results

Lens Calibrated field of view

Calibrated field of view, horizontal × vertical. cam0: 122° × 91°, corners reach 64° from the axis · cam1: 122° × 91°, corners reach 64° from the axis

cam0 cam1

We read these from the calibrated lens fit at the image edges, through the principal point (where the optical axis meets the image). ~ marks an axis past where the corners reached, so the fit is extrapolated there. Past the fit means an edge lies beyond where the fit holds, so we give no value.

IMU Gyro–accel time offset t_ga −1.03 ms

This solve pinned the IMU intrinsics at identity, so it estimated no scales. t_ga, the time offset between the gyroscope and accelerometer channels: −1.03 ± 0.50 ms.

Timing cam0 td +5.77 ms

The time offsets, to scale. td is the camera–IMU time offset, with t_imu = t_cam + td.

Full results Every solved parameter

Rotation is log(R) in rad, camera in the IMU frame.

Parameter	MVIS
cam0
fx px	601.56
fy px	601.47
cx px	965.27
cy px	632.09
k1	0.22798
k2	0.50518
k3	−0.48094
k4	0.17631
Lever x mm	−29.78
Lever y mm	−9.79
Lever z mm	+18.01
log(R) rad	[−0.0073, −0.0032, +0.0029]
td ms	+5.770
Reprojection px	0.189
cam1
fx px	601.32
fy px	601.33
cx px	952.65
cy px	563.80
k1	0.23524
k2	0.49043
k3	−0.46904
k4	0.17017
Lever x mm	+30.53
Lever y mm	−9.72
Lever z mm	+17.63
log(R) rad	[+0.0017, +0.0032, +0.0034]
td ms	+5.745
Reprojection px	0.177
IMU
t_ga ms	−1.03 ± 0.50

Limits

• No independent reference. These values show what MVIS solved, not how close they are to the truth.
• The supplier's IMU noise file (gyro 0.002, accel 0.02) is tuned for Kalibr, about 10× and 3× the noise measured at rest. Used as is in MVIS, it put unit A's lever arm 20 to 45 mm too far forward. These results use the measured noise with a 2 to 3× margin.
• With IMU intrinsics free, the lever arm's y shifts about 9 mm on both units, more than its 3.5 mm uncertainty. Pinned, it agrees with the supplier's Kalibr, so these results pin the IMU intrinsics.
• The listing says 1080P and 160°, and the supplier's sheet says 1920×1200 and 123.6°. These recordings are 1920×1200. The listing's shutter value wasn't tested here.

How it was run

• cam0 at 30 Hz, cam1 at 10 Hz, t_ga estimated, camera intrinsics free, IMU intrinsics pinned. IMU noise padded from the measured noise: gyro 6e-4 rad/s/√Hz, accel 0.015 m/s²/√Hz. Batch of 2026-11-07.