Vehicle-Camera Calibration
- 1.Calibration list page, where the users can load existing dataset or create a new one.
2. New calibration selection modal.
3. Get started page of vehicle-camera setup.
4. Calibration settings modal.
- Dataset name can be added here.
- The user has to select the shape of the vehicle. Either rectangle or trapezoid.
For rectangle-shaped vehicles, users can input the measured values
For trapezoid-shaped vehicles, users can input the following measured values.
Description for vehicle details:
Checkerboard Config:
configure the checkerboard at the bottom of the modal
5. Enter the intrinsic parameters for the camera and the camera fov.
- Intrinsic parameters for the camera are to be added here. Users have three options.
- Users can use the intrinsic calibration tool and calibrate the results. Save them to profile and then load them here.
- Or users can also load the JSON file.
- Select the FOV for your camera, based on its actual position
6. Add images related to the mounted camera.
7. Enter the target configuration for each camera image
- For each image enter the target configuration, by mentioning the VRP to IRP, the target height and IRP to TRP
- Also mention the position in which is the board is placed can be either perpendicular or parallel to the ground
8. Detect the checkerboard corners for both the mounted camera images.
- Click on detect corners to get the checkerboard corners auto-detected.
9. Click on the run calibration button. This takes all the input configuration and the file data to get the calibrated results.
10. Top right bar shows the extrinsic parameters.
11. Visualize button shows the 3d representation of the car and its wheels. Along with the camera center and its frustum.
12. Export option helps the user to export the calibrated data for the mounted camera with the vehicle.
13. Users can check the error stats and add more images to see the change in error stats.
- Reprojection Error: It’s value is the mean delta of the marked wheel point and the reprojection of the calibrated wheel position. It's measured in pixels.
- Translation Error: The distance between the means of the 3d projections of the checkerboard in a 3d scene from the left images and right images
- Rotation Error: The angle between the planes of the 3d projections in a 3d scene of the checkerboard from the left images and right images
We have a save option on the top right corner. A user can click on the Save button to save the calibration dataset at any time during the calibration process.
If the checkerboard corners are not auto-detected. Users can select four boundary points in order (top-left, top-right , bottom-left, bottom-right). And then click on retry corner detection, to get the remaining inner corners of the checkerboard.
- roll, pitch, yaw are in degrees and px, py, pz are in meters.
- roll, pitch, yaw, px, py, pz are the extrinsic parameters downloaded from the calibration tool.
- vehiclePoint3D is the 3d coordinates of a point in the vehicle coordinate system.
- imagePoint3D is the 3d coordinates of a point in the camera coordinate system.
We currently show three different types of the camera sensor coordinate system. On selecting the camera coordinate system, the extrinsic parameters change accordingly. The export option exports the extrinsic parameters based on the selected camera coordinate system.
- Optical coordinate system: Its the default coordinate system which we follow.
- ROS REP 103: It is the coordinate system followed by ROS. On changing to this, you can see the change in the visualization and the extrinsic parameters.
- NED : This follows the north-east-down coordinate system.