MartyG-RealSense
commented
1 month ago Hi @Robertleoj The On-Chip tool improves depth image quality, whilst Tare is the one that improves depth accuracy.
There is not necessarily a problem with the calibration of your camera though. Accuracy of the depth value of coordinates does tend to drift at the edges of an image compared to the center.
When you performed a Tare calibration, if you did the calibration in the RealSense Viewer tool then did you click the Get button to customize the calibration to the distance that you had positioned the camera from the target image, please? If not then further information can be found at https://github.com/IntelRealSense/librealsense/issues/11476#issuecomment-1441647959
If you wish to try to reset your camera's calibration to its factory-new default values then you can do so using instructions at https://github.com/IntelRealSense/librealsense/issues/10182#issuecomment-1019854487
Issue Description
My D405 has a warped depth map: If I position the camera normal to a plane that is 203.7mm away from the camera, it reads 203.7mm at the middle, but closer to the edges of the FOV, I'm seeing it go up to 207mm.
The textured target from the self-calibration docs is on the plane, stuck with an adhesive, so texture should not be an issue.
I've tried running the on-chip calibration in both intrinsic and extrinsic mode, in addition to running the tare calibration and the dynamic targeted calibration, none fix this issue.
The cameras in my company are mounted on an outdoor robot, and experiences heat fluctuations and vibrations, so I'd like a reliable way to restore calibration on these cameras.