Hi, I would like to ask if there's any intention for Aximmetry to update their methods for Lens Calibration as the current method involves moving the camera around a space, and this requires quite a bit of space as well.
I would suggest an alternate method be made available where the Camera is actually still, and instead, we move the checkerboard around the space, and gauging the distance using a similar method as current. Moving the board is much faster than moving the camera, and it will make calibration a lot more convenient and faster.
Secondly, I'm not sure if the current distortion values are adequate for lenses that are not perfect, eg. Zoom lenses. Will there be any consideration to include p1, p2 values for tangential distortion into the lens distortion calculations?
Personally I think my lens, which is a Sigma 18-35 F4.0 USM has issues at the edges that were not able to allow a perfect distortion calibration. Especially when panning left and right, reveals that the distortion calibrated is insufficient.
Please do consider these suggestions for implementation in the future versions of aximmetry
Hi Terence,
You can perform a lens calibration without moving your camera by selecting the PTZ camera tracking type when creating the Calibration Profile:
If you do so, then you can also create a 6DOF tracking calibration profile and then copy the lens calibration results from the PTZ calibration file into the 6DOF file.
Performing tracking calibration without moving the camera might be as challenging as with camera movement. However, this method is likely under consideration due to frequent requests from individuals with small studios.
Regarding the p1 and p2 values, I'm not sure about their specific impact. A better calibration tool is in development, and I am confident that the p1 and p2 values are being considered.
Warmest regards,