Example Calibrate Planar Multi

From BoofCV
Revision as of 12:23, 2 September 2022 by Peter (talk | contribs) (Created page with "Example of calibrating a multi camera system from one planar calibration target. It's assumed that all the cameras are synchronized. They can be either stationary with a movin...")
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigationJump to search

Example of calibrating a multi camera system from one planar calibration target. It's assumed that all the cameras are synchronized. They can be either stationary with a moving target or moving with a stationary target.

Example File: ExampleCalibrateMulti.java

Calibration Tutorial: Wikipage


  • Camera calibration
  • Lens distortion
  • Intrinsic parameters

Related Examples:

Example Code

 * Demonstrates calibration of a N-camera system. This could be a three camera stereo system, 30 cameras in a ring,
 * or any similar variant.
 * @author Peter Abeles
public class ExampleCalibrateMulti {
	public static void main( String[] args ) {
		// Creates a detector and specifies its physical characteristics
		CalibrationDetectorMultiECoCheck detector = FactoryFiducialCalibration.ecocheck(null,
				ConfigECoCheckMarkers.parse("14x10n1", /* square size */1.0));

		String directory = UtilIO.pathExample("calibration/trinocular/");

		// Images for each camera are in their own directory
		List<String> left = UtilIO.listSmartImages(directory + "left", true);
		List<String> middle = UtilIO.listSmartImages(directory + "middle", true);
		List<String> right = UtilIO.listSmartImages(directory + "right", true);

		// Configure the calibration class for this scenario
		var calibrator = new CalibrateMultiPlanar();
		// Tell it what type of camera model to use
		calibrator.getCalibratorMono().configurePinhole(true, 3, false);
		// NOTE: For now only one calibration target is supported. Support for multiple targets is planned for
		//       the future.
		calibrator.initialize(/*num cameras*/3, /*num targets*/1);
		calibrator.setTargetLayout(0, detector.getLayout(0));
		calibrator.setCameraProperties(0, 1224, 1024);
		calibrator.setCameraProperties(1, 1224, 1024);
		calibrator.setCameraProperties(2, 1224, 1024);

		for (int imageIdx = 0; imageIdx < left.size(); imageIdx++) {
			System.out.print("image set " + imageIdx + ", landmark count:");
			// Detect calibration targets and save results into a synchronized frame. It's assumed that each
			// set of images was taken at the exact moment in time
			GrayF32 imageLeft = UtilImageIO.loadImage(left.get(imageIdx), GrayF32.class);
			GrayF32 imageMiddle = UtilImageIO.loadImage(middle.get(imageIdx), GrayF32.class);
			GrayF32 imageRight = UtilImageIO.loadImage(right.get(imageIdx), GrayF32.class);

			var syncObs = new SynchronizedCalObs();
			addCameraObservations(0, imageLeft, detector, syncObs);
			addCameraObservations(1, imageMiddle, detector, syncObs);
			addCameraObservations(2, imageRight, detector, syncObs);


		System.out.println("Performing calibration");

		// Print out optimization results. Can help you see if something has gone wrong
		calibrator.getBundleUtils().sba.setVerbose(System.out, null);
		BoofMiscOps.checkTrue(calibrator.process(), "Calibration Failed!");


		MultiCameraCalibParams params = calibrator.getResults();
		CalibrationIO.save(params, "multi_camera.yaml");

	private static void addCameraObservations( int cameraID, GrayF32 image,
											   CalibrationDetectorMultiECoCheck detector, SynchronizedCalObs dst ) {

		// Find calibration targets inside the image

		// Find the target which matches the expected target ID
		var set = dst.cameras.grow();
		set.cameraID = cameraID;
		for (int i = 0; i < detector.getDetectionCount(); i++) {
			if (detector.getMarkerID(i) != 0)

		// Print number of calibration points it found
		System.out.print(" " + set.targets.getTail().points.size());