-
Notifications
You must be signed in to change notification settings - Fork 177
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Camera factory calibration interpretation #554
Comments
Did you use the Luxonis ROS package for calibrating? You can also get the calibration data from the depthai SDK using the calibrate.py file. It will save a file of calibration data if used correctly. |
This is with the factory calibrations published on the |
I am not associated with Luxonis. I am just trying to help you. Are you using a FFC board? If you calibrate using the file I suggested, then that code will flash the device memory, I think, and the ROS messages will show the updated calibration data. |
I do appreciate your input; we can calibrate the camera, yet would prefer to use the factory calibration as recalibrating a camera adds a considerable cost to each camera. However, unlike with the Intel RealSense cameras, the factory calibration simply doesn't work out for us. |
Check if issue already exists
Here's the same issue on the Luxonis forum.
Describe the bug
Applying the factory calibration published in the CameraInfo message does not deliver sensible results. Recalibrating the camera using standard tools resulted in a very different calibration which does work, but preferably we would like to use what gets published on the camera_info topics.
Minimal Reproducible Example
Here's an MRE, including a bag file. It uses a marker detection library to generate a set of 3D points from the right camera image and then projects these back onto image space for all three cameras.
Key code used for projecting 3D points to image coordinates using the Luxonis calibration:
Expected behavior
The centers of the red circles align with the fiducial marker centers in all three images in the screenshot below.
Observed behavior
The centers align in the image of the right camera, indicating that the projections of the made node (see the linked MRE code) match those of the marker detection library. This is good. However, the centers don't align in the images of the left and the rgb camera, indicating a severe calibration (interpretation) issue.
Screenshots
![image](https://private-user-images.githubusercontent.com/107043154/344677206-d980bf95-01fe-4983-bb42-f4cddd66618c.png?jwt=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpc3MiOiJnaXRodWIuY29tIiwiYXVkIjoicmF3LmdpdGh1YnVzZXJjb250ZW50LmNvbSIsImtleSI6ImtleTUiLCJleHAiOjE3MjMzNTA3MDYsIm5iZiI6MTcyMzM1MDQwNiwicGF0aCI6Ii8xMDcwNDMxNTQvMzQ0Njc3MjA2LWQ5ODBiZjk1LTAxZmUtNDk4My1iYjQyLWY0Y2RkZDY2NjE4Yy5wbmc_WC1BbXotQWxnb3JpdGhtPUFXUzQtSE1BQy1TSEEyNTYmWC1BbXotQ3JlZGVudGlhbD1BS0lBVkNPRFlMU0E1M1BRSzRaQSUyRjIwMjQwODExJTJGdXMtZWFzdC0xJTJGczMlMkZhd3M0X3JlcXVlc3QmWC1BbXotRGF0ZT0yMDI0MDgxMVQwNDI2NDZaJlgtQW16LUV4cGlyZXM9MzAwJlgtQW16LVNpZ25hdHVyZT0wMDkwNjYyNTQ4N2ExNzViMDgwZDA0OGM5MjQwY2YwOWY0MTZlNWViMDQ0NmViNTU2NThiZDI3YjA0ZTEyZDFhJlgtQW16LVNpZ25lZEhlYWRlcnM9aG9zdCZhY3Rvcl9pZD0wJmtleV9pZD0wJnJlcG9faWQ9MCJ9.1ZX8RV77kHDMlIaBhrTqzGII5KGi3w-JOtclU6TvPzU)
Pipeline Graph
Standard pipeline
RGBStereo
.Attach system log
depthai-ros
built from source or installed from apt? Occurs with bothdepthai/depthai-core
library installed from rosdep or manually? For rosdep install, check ifros-<rosdistro>-depthai
package is installed, manual install can be checked withldconfig -p | grep depthai
. ros-noetic-depthai is installedapt show ros-$ROS_DISTRO-depthai ros-$ROS_DISTRO-depthai-ros ros-$ROS_DISTRO-depthai-bridge ros-$ROS_DISTRO-depthai-ros-msgs ros-$ROS_DISTRO-depthai-ros-driver
Versions:
DEPTHAI_DEBUG=1
and paste the logs, either from command line or from latest log in~/.ros/log
Additional context
Thanks for taking your time to look into this.
The text was updated successfully, but these errors were encountered: