hi
I want to ask, when using calibrate.py to calibrate the FFC-4P, if a JSON file generated by a 9298 grayscale camera with a resolution of 1280X800 is used, I found that there is a significant error when using the camera parameters and distortion coefficients from this file to calibrate images from a 400P camera. How can this be resolved, or how can the resolution of the JSON file generated during calibration be set to 400P?
Thanks
Hi @Tjjl
The intrinsics change in the back when using a lower resolution. If you want to manually specify the new intrinsics at a lower (different resolution) you can use https://docs.luxonis.com/projects/api/en/latest/references/python/?highlight=getcameraintrinsics#depthai.CalibrationHandler.getCameraIntrinsics.
Thanks,
Jaka
Hi
Jaka
Thank you for your assistance. I have resolved the issue.
I have another question to consult. When I use the code depthBC->disparity.link(depthBCOut->input) to obtain depth information, I want to convert this type of data into actual depth values. I use the code BCMat.at<uchar>(stay, stax) to read the disparity value of depthBC, and then I use the formula from this link: Configuring Stereo Depth to calculate the depth. We are using a focal length $��$fx of 441.25 corresponding to 400p. However, the calculated depth is approximately twice the actual depth.
Thank you!
We created a mono node, and in the code "BCMat.at<uchar>(stay, stax)", 'stay' and 'stax' represent the y and x coordinates of the input image respectively. We did not use any extensions. Our experiment showed a distance of 115cm, whereas the actual measured distance is 70cm.
Thank you.
