2D Optical Calibration
This version of the calibration system allows you to calibrate your headset with a single stereo camera
Above: A video walking you through the 2D calibration process for northstar.
- 2.Setup python with the following dependencies:
1) Pyrealsense2 2) opencv-contrib-python 3) numpy
Run through the following steps to calibrate your headset.
- captureGraycodes.py - run
python captureGraycodes.pyto run this script
- Ensure that your headset is placed on the calibration stand, with the stand's camera looking through it where the users' eyes will be.
- On line 44 there is a line of code that changes the window offset for the graycode generator. This example here moves the window 1920 pixels to the right, and 0 up and down. You'll want the X value (1920) to be your main monitor's width in pixels.
cv2.moveWindow ("Graycode Viewport", 1920, 0)
- It helps to place a piece of cloth over the rig to shield the cameras + headset from ambient light.
- The sequence of binary codes will culminate in a 0-1 UV mapping, saved to "./WidthCalibration.png" "./HeightCalibration.png" in your main folder.
- Running this script will fit a 3rd-Degree 2D Polynomial to the left and right "eye"'s X and Y distortions.
- This polynomial will map from each display's 0-1 UV Coordinates to rectilinear coordinates (where a 3D ray direction is just (x, y, 1.0)).
When you are finished, you may paste the output of the calibrateGraycodes.py into this diagnostic shadertoy to check for alignment. Additionally, ensure that your headset is plugged in and displaying imagery from your desktop. Running this script will display a sequence of gray codes on your North Star, capturing them at the same time.
The shader toy should look like this when viewed through the headset
Additionally, there should be a
NorthStarCalibration.jsonin this directory which you may use in the Unity implementation.