This page lets users view, create, launch, and delete calibration datasets. Admins can manage users’ access to these datasets on this page.
Click on New Calibration to create a new calibration dataset.
Select LiDAR-Camera Calibration to create a new dataset.
Upon selecting LiDAR-Camera Calibration, the user is welcomed to the instructions page. Click on Get started to start the calibration setup.
Users can choose either the target-based or the targetless calibration. The target-based calibration uses the checkerboard as the calibration target, and the targetless calibration uses the features available in the scene.
Horizontal corners: Total number of inner corners from left to right. The blue dots shown in the above preview correspond to the horizontal corners.
Vertical corners: Total number of inner corners from top to bottom. The red dots shown in the above preview correspond to the vertical corners.
Square size: This is the length of the square's arm in meters. It corresponds to the length of the yellow square highlighted in the preview.
Intrinsic parameters for the camera are to be added here. Users have three options.
Users can use the Camera Intrinsic calibration tool to calibrate the results, save them to the profile, and then load them here. For more details, click here.
Users can also load the JSON file.
Users can manually enter the intrinsic parameters if they already have them.
Intrinsic parameters for the camera are to be added here. Users have three options.
Users can use the Camera Intrinsic calibration tool to calibrate the results, save them to the profile, and then load them here. For more details, click here.
Users can also load the JSON file.
Users can manually enter the intrinsic parameters if they already have them.
You need to add the images taken from the left and right cameras of the Overlapping Camera. This page allows you to pair them manually. It also has an “Auto Pair” button to pair the files automatically based on their filename order.
Choose the feature matching algorithm and click on Detect for feature detection
On this page, the user has to click Run calibration to get the calibration results. Once clicked, the images are processed, and the calibration results are shown on the right side of the page.
Each epiline is the projection of the corresponding point from the opposite image. Hence, the quality of the calibration can be judged by how close the epilines are to the corresponding points. Also, all the epilines in one image converge to the epipole (could be outside the image). For convenience, you can see the Epiline point distance (average distance of the point to its respective epiline) error stat.
You can toggle the error stats and select individual errors or average errors. Individual Errors show the errors related to an image pair chosen on the left pane.
Left Reprojection Error: The pixel error when the detected corners from the right image are projected onto the left image using the calibration results
Right Reprojection Error: The pixel error when the detected corners from the left image are projected onto the right image using the calibration results
Translation Error: The distance between the centroids (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
Epiline Point Distance: Average pixel distance of each point to its corresponding projected epiline.
Epipolar Error: Proportional to the distance of a point from its epiline. Does not have a physical meaning. It is the residual error from minimizing the epipolar constraints while calculating the fundamental/essential matrix.
Clicking on the visualize button takes the user to the visualization screen. The user can see the cameras and the selected image-pair checkerboard on the screen. To get a better sense of the view, the user can move around in the 3D scene using the mouse left (for pan) or right (for orbit) button click-drag. Also, at the top-right, in the options dropdown, you have the option to show/hide all the checkerboards from all the image pairs at once.
Only frustums are visible for the targetless calibration.
