Skip to content
RealSense ↗︎

RealSense D435i

Depth camera: global-shutter infrared stereo, rolling-shutter RGB and a 6-axis IMU

RealSense D435i depth camera, front view
Photo: RealSense, D435i product page ↗︎
Cameras
640 × 480 here: RGBMeasured
Shutter
Rolling (RGB)Maker ↗︎
Field of view
Calibrated 56° × 43°Measured
69° × 42° RGB, 87° × 58° depthMaker ↗︎
Baseline
50 mm, infrared pair (not used here)Maker ↗︎
IMU
BMI055, 200 HzMeasured
Output
USB-C, EuRoC mav0Maker ↗︎

Maker values checked 2026-11-14 · 3 recordings, 67.7 to 80 s

Depth camera from RealSense with a 50 mm global-shutter infrared stereo pair, a rolling-shutter RGB camera and a 6-axis IMU. Three hand-held recordings in front of an AprilGrid, each solved as RGB + IMU with the rolling-shutter readout estimated.

The 3D scene MVIS recovered from the first recording.
Reprojection median0.10 px
Readout time32.6 ms
cam0 td, first row+6.00 ms+22.28 ms at mid-frame
cam0 lever arm30.2 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.

The three recordings agree on the time offset (6.00 to 6.12 ms), the readout time (32.5 ms) and the lever arm across the image (within 0.4 and 1.4 mm), but not along the optical axis: z is 14.7 to 20.4 mm, against 11.3 mm in the RealSense SDK's design values. All three recordings turn the camera little about the two axes across the image, which is what makes that distance observable. Holding z at the SDK value fits nearly as well (within about 2 σ), so the SDK value is the safer choice for now.

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.

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.

To scale, from above and the front

From above
xzcam030.2 mm20 mm
From the front
xycam020 mm
CameraLever x, y, z (mm)Distance (mm)log(R) (rad)td (ms)Field of view
cam0−20.8, +8.0, +20.430.2[+0.003, −0.005, −0.001]+6.0056° × 43°

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

Published by the maker
MakerRealSense (formerly Intel RealSense)
ShutterGlobal on the infrared pair, rolling on the RGB camera
Infrared pair87° × 58° depth field of view; not used here
RGB camera2 MP, 69° × 42°, up to 1920×1080 at 30 fps; recorded here at 640×480, 30 Hz, exposure fixed at 3 ms
LinkUSB-C 3.1 Gen 1; 90 × 25 × 25 mm
Measured in these recordings
IMUBosch BMI055; gyro and accelerometer at 200 Hz in these recordings, accelerometer resampled onto the gyro stamps
Calibrated field of viewcam0 56° × 43°
Recordings3 recordings, 80, 67.7 and 77 s, hand-held. About 2,000 to 2,400 RGB frames each, IMU at 200 Hz
FormatEuRoC mav0: RGB PNG and IMU CSV
BoardAprilGrid, 6×6 tags of 88 mm, spacing 0.3
ExposureFixed, 3 ms; gain fixed after a short warm-up; white balance fixed
DataNot hosted
AprilGrid 6 × 6, 660 × 660 mm. Standard pattern, to scale
Lens Calibrated field of view

Calibrated field of view, horizontal × vertical. cam0: 56° × 43°, corners reach 33° from the axis

0°15°30°45°60°75°90°0.00.40.81.21.6Ray angle from the optical axisImage radius / f

cam0

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 −3.87 ms

Each bar shows a scale as its deviation from 1. IMU model: calyx. t_ga, the time offset between the gyroscope and accelerometer channels: −3.87 ± 0.53 ms.

Gyro scale x−0.03 %
Gyro scale y+0.23 %
Gyro scale z−0.51 %
Accel scale x+0.91 %
Accel scale y+1.17 %
Accel scale z+1.00 %
−8 %0+8 %
Timing cam0 td, first row +6.00 ms

The time offsets, to scale. td is the camera–IMU time offset, with t_imu = t_cam + td. For a rolling-shutter camera td is the time of the first image row; the middle row is td + readout / 2, which is what a global-shutter calibration reports.

-10 ms-5 ms+0 ms+5 ms+10 mscam0 td+6.00t_ga−3.87
Board shape peak to peak 5.8 mm

The calibration board isn't perfectly flat, so this solve estimated the height of every corner off the flat board along with the calibration. The board's mean plane and tilt belong to its pose, so a flat board shows only noise here.

+1.91 mm+2.03 mm+0.49 mm+0.43 mm+2.03 mm+2.20 mm+0.59 mm+0.53 mm+2.07 mm+2.25 mm+0.47 mm+0.59 mm+2.12 mm+2.02 mm+0.16 mm+0.62 mm+1.87 mm+2.04 mm+0.40 mm+0.43 mm+1.59 mm+1.41 mm-0.14 mm+0.39 mm+0.07 mm+0.11 mm-0.43 mm-0.53 mm+0.20 mm+0.21 mm-0.32 mm-0.46 mm+0.15 mm+0.17 mm-0.44 mm-0.33 mm+0.05 mm+0.06 mm-0.57 mm-0.34 mm-0.20 mm+0.29 mm-0.69 mm-0.39 mm-0.21 mm-0.29 mm-1.26 mm-0.40 mm-0.80 mm-0.57 mm-0.91 mm-1.32 mm-0.61 mm-0.41 mm-0.62 mm-0.90 mm-0.50 mm-0.47 mm-0.74 mm-0.58 mm-0.63 mm-0.45 mm-0.69 mm-0.66 mm-0.76 mm-0.63 mm-1.15 mm-0.69 mm-0.88 mm-1.33 mm-2.46 mm-1.26 mm-1.48 mm-0.93 mm-1.05 mm-1.72 mm-0.97 mm-0.57 mm-0.49 mm-0.91 mm-0.62 mm-0.68 mm-0.35 mm-0.48 mm-0.68 mm-0.74 mm-0.33 mm-0.41 mm-0.80 mm-1.28 mm-1.36 mm-0.66 mm-1.49 mm-2.53 mm-2.53 mm-1.63 mm-1.72 mm-0.93 mm-0.53 mm-1.50 mm-0.85 mm-0.42 mm+0.17 mm-0.46 mm-0.34 mm-0.26 mm+0.75 mm+0.13 mm-0.10 mm-0.39 mm+0.94 mm+0.64 mm-0.36 mm-1.33 mm+0.20 mm+0.59 mm-1.28 mm-2.28 mm-1.09 mm-0.09 mm-1.09 mm-0.24 mm+1.04 mm-0.17 mm-0.03 mm+0.48 mm+1.94 mm+1.03 mm+0.70 mm+0.99 mm+2.74 mm+1.85 mm+1.31 mm+1.26 mm+3.28 mm+2.63 mm+1.48 mm+0.77 mm+3.32 mm+3.19 mm+0.82 mm-0.23 mm+2.22 mm+2.97 mm-3.3 mmflat+3.3 mm
Height of each corner off the flat board: peak to peak 5.8 mm, rms 1.2 mm. Hover a corner for its value.
Full results Every solved parameter

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

ParameterMVIS
cam0
fx px599.37
fy px598.02
cx px333.53
cy px249.02
k10.10194
k2−0.20872
k30.00257
k4−0.00001
Lever x mm−20.76
Lever y mm+8.01
Lever z mm+20.40
log(R) rad[+0.0026, −0.0054, −0.0010]
td ms+6.003
Reprojection px0.124
Readout ms+32.55
IMU
t_ga ms−3.87 ± 0.53
  • The three recordings agree on the time offset (6.00 to 6.12 ms), the readout time (32.5 ms) and the lever arm across the image (within 0.4 and 1.4 mm), but not along the optical axis: z is 14.7 to 20.4 mm, against 11.3 mm in the RealSense SDK's design values. All three recordings turn the camera little about the two axes across the image, which is what makes that distance observable. Holding z at the SDK value fits nearly as well (within about 2 σ), so the SDK value is the safer choice for now.
  • The tag size, 88 mm, is the stated size of the printed board, and it scales every lever arm.
  • The board is not flat: its corners lie up to about 6 mm out of plane, the same shape in every recording.
  • The camera stores no IMU correction, so the IMU scale and bias values are those of the raw sensor.
  • RGB camera with the IMU in one solve per recording, as a rolling-shutter camera with the readout time estimated. Pinhole camera with radial-tangential distortion, seeded from a single-camera solve of the same recording and left free; the camera time offset, IMU scale and misalignment and the gyro-to-accel offset t_ga estimated. The board shape is estimated per corner. IMU noise is the MVIS default, since no measured values exist for this unit. Solved 2026-11-14.