boofcv.abst.geo.pose.PnPRefineRodrigues Maven / Gradle / Ivy

Show more of this group Show more artifacts with this name
Show all versions of boofcv-geo Show documentation
BoofCV is an open source Java library for real-time computer vision and robotics applications.
There is a newer version: 1.1.5
/*
 * Copyright (c) 2021, Peter Abeles. All Rights Reserved.
 *
 * This file is part of BoofCV (http://boofcv.org).
 *
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 *
 *   http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

package boofcv.abst.geo.pose;

import boofcv.abst.geo.RefinePnP;
import boofcv.abst.geo.optimization.ResidualsCodecToMatrix;
import boofcv.alg.geo.pose.PnPJacobianRodrigues;
import boofcv.alg.geo.pose.PnPResidualReprojection;
import boofcv.alg.geo.pose.PnPRodriguesCodec;
import boofcv.struct.geo.Point2D3D;
import georegression.struct.se.Se3_F64;
import org.ddogleg.fitting.modelset.ModelCodec;
import org.ddogleg.optimization.FactoryOptimization;
import org.ddogleg.optimization.UnconstrainedLeastSquares;

import java.util.List;

/**
 * Minimizes the projection residual error in a calibrated camera for a pose estimate.
 * Rotation is encoded using rodrigues coordinates.
 *
 * @author Peter Abeles
 */
public class PnPRefineRodrigues implements RefinePnP {

	ModelCodec paramModel = new PnPRodriguesCodec();
	ResidualsCodecToMatrix func;
	PnPJacobianRodrigues jacobian = new PnPJacobianRodrigues();

	double[] param;
	UnconstrainedLeastSquares minimizer;
	int maxIterations;
	double convergenceTol;

	public PnPRefineRodrigues( double convergenceTol, int maxIterations ) {
		this.maxIterations = maxIterations;
		this.convergenceTol = convergenceTol;
		this.minimizer = FactoryOptimization.levenbergMarquardt(null, false);

		func = new ResidualsCodecToMatrix<>(paramModel, new PnPResidualReprojection(), new Se3_F64());

		param = new double[paramModel.getParamLength()];
	}

	@Override
	public boolean fitModel( List obs, Se3_F64 worldToCamera, Se3_F64 refinedWorldToCamera ) {

		paramModel.encode(worldToCamera, param);

		func.setObservations(obs);
		jacobian.setObservations(obs);

		minimizer.setFunction(func, jacobian);

		minimizer.initialize(param, 0, convergenceTol*obs.size());

		boolean updated = false;
		for (int i = 0; i < maxIterations; i++) {
			boolean converged = minimizer.iterate();
			if (converged || minimizer.isUpdated()) {
				// save the results
				paramModel.decode(minimizer.getParameters(), refinedWorldToCamera);
				updated = true;
			}
			if (converged) {
				if (i == 0) {
					// if it converted on the first iteration then that means it already
					// meet convergence. use input to avoid introduction of small numerical errors
					refinedWorldToCamera.setTo(worldToCamera);
				}
				break;
			}
		}

		return updated;
	}

	@Override
	public double getFitScore() {
		return minimizer.getFunctionValue();
	}
}