Hi Nearpoint
The TOF calibration is available on the new_tof_calib branch of depthai.
As for the instructions:
If you have already installed depthai repository, you can just run:
git checkout new_tof_calib
git submodule update --init --recursive
to update the boards and:
python3 .\install_requirements.py
to update the depthai-python libraries.
To start calibration run calibrate.py with the right parameters for Charuco board you are using.
Example:
python3 calibrate.py -db -nx 12 -ny 9 -c 1 -cd 0 -s 6 -ms 4.7 -brd OAK-D-SR-POE
Where:
-db is default board so you are using charuco markers
-nx number of charuco markers in x direction
-ny number of charuco markers in y direction
-c number of pictures taken each time polygon is displayed (not needed, suggested to be left out in your case)
-cd countdown time before picture is taken in seconsd (not needed, suggested just in case, you need faster image calibration)
-s size os square around the charuco marker in centimeters
-ms size of markers in centimeters
-brd board of device (in our case OAK-D-SR-POE)
Because TOF is a little problematic, just because the board must be close to camera to detect charuco boards, you can also use modes with -m .
My suggestion is that you first try to run calibration and in case, that program gives you error
division by zero
or
Failed to detect markers in the image dataset/rgb\rgb_p3_10.png
you go to dataset folder and delete the picture, which has bad detection of charuco boards (in all camera folders) and run the same code again with added parameter -m process . This will just try the processing stage, so you will not have to take pictures of the board again.
I'll be adding this to the docs as well today.
Thanks,
Jaka