/* * This file is part of the LIRE project: http://lire-project.net * LIRE is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * LIRE is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License * along with LIRE; if not, write to the Free Software * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA * * We kindly ask you to refer the any or one of the following publications in * any publication mentioning or employing Lire: * * Lux Mathias, Savvas A. Chatzichristofis. Lire: Lucene Image Retrieval – * An Extensible Java CBIR Library. In proceedings of the 16th ACM International * Conference on Multimedia, pp. 1085-1088, Vancouver, Canada, 2008 * URL: http://doi.acm.org/10.1145/1459359.1459577 * * Lux Mathias. Content Based Image Retrieval with LIRE. In proceedings of the * 19th ACM International Conference on Multimedia, pp. 735-738, Scottsdale, * Arizona, USA, 2011 * URL: http://dl.acm.org/citation.cfm?id=2072432 * * Mathias Lux, Oge Marques. Visual Information Retrieval using Java and LIRE * Morgan & Claypool, 2013 * URL: http://www.morganclaypool.com/doi/abs/10.2200/S00468ED1V01Y201301ICR025 * * Copyright statement: * -------------------- * (c) 2002-2013 by Mathias Lux (mathias@juggle.at) * http://www.semanticmetadata.net/lire, http://www.lire-project.net */ package net.semanticmetadata.lire.imageanalysis.features.local.sift; import java.awt.geom.AffineTransform; import java.util.Collection; import java.util.Random; /** * Abstract class for arbitrary geometric transformation models to be applied * to points in n-dimensional space. * <p/> * Provides methods for generic optimization and model extraction algorithms. * Currently, RANSAC and Monte-Carlo minimization implemented. Needs revision... * <p/> * TODO A model is planned to be a generic transformation pipeline to be * applied to images, volumes or arbitrary sets of n-dimensional points. E.g. * lens transformation of camera images, pose and location of mosaic tiles, * non-rigid bending of confocal stacks etc. * <p/> * License: GPL * <p/> * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License 2 * as published by the Free Software Foundation. * <p/> * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. * <p/> * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. * <p/> * NOTE: * The SIFT-method is protected by U.S. Patent 6,711,293: "Method and * apparatus for identifying scale invariant features in an image and use of * same for locating an object in an image" by the University of British * Columbia. That is, for commercial applications the permission of the author * is required. * * @author Stephan Saalfeld <saalfeld@mpi-cbg.de> * @version 0.1b */ abstract public class Model { // minimal number of point correspondences required to solve the model static final public int MIN_SET_SIZE = 0; // real random //final Random random = new Random( System.currentTimeMillis() ); // repeatable results final static Random rnd = new Random(69997); /** * error depends on what kind of algorithm is running * small error is better than large error */ public double error; /** * instantiates an empty model with maximally large error */ public Model() { error = Double.MAX_VALUE; } /** * fit the model to a minimal set of point correpondences * estimates a model to transform match.p2.local to match.p1.world * * @param min_matches minimal set of point correpondences * @return true if a model was estimated */ public abstract boolean fit(PointMatch[] min_matches); /** * apply the model to a point location * * @param point * @return transformed point */ public abstract float[] apply(float[] point); /** * apply the model to a point location * * @param point */ public abstract void applyInPlace(float[] point); /** * apply the inverse of the model to a point location * * @param point * @return transformed point */ public abstract float[] applyInverse(float[] point); /** * apply the inverse of the model to a point location * * @param point */ public abstract void applyInverseInPlace(float[] point); /** * test the model for a set of point correspondence candidates * <p/> * clears inliers and fills it with the fitting subset of candidates * * @param candidates set of point correspondence candidates * @param inliers set of point correspondences that fit the model * @param epsilon maximal allowed transfer error * @param min_inliers minimal ratio of inliers (0.0 => 0%, 1.0 => 100%) */ public boolean test( Collection<PointMatch> candidates, Collection<PointMatch> inliers, double epsilon, double min_inlier_ratio) { inliers.clear(); for (PointMatch m : candidates) { m.apply(this); if (m.getDistance() < epsilon) inliers.add(m); } float ir = (float) inliers.size() / (float) candidates.size(); error = 1.0 - ir; if (error > 1.0) error = 1.0; if (error < 0) error = 0.0; return (ir > min_inlier_ratio); } /** * less than operater to make the models comparable, returns false for error < 0 * * @param m * @return false for error < 0, otherwise true if this.error is smaller than m.error */ public boolean betterThan(Model m) { if (error < 0) return false; return error < m.error; } /** * randomly change the model a bit * <p/> * estimates the necessary amount of shaking for each single dimensional * distance in the set of matches * * @param matches point matches * @param scale gives a multiplicative factor to each dimensional distance (scales the amount of shaking) * @param center local pivot point for centered shakes (e.g. rotation) */ abstract public void shake( Collection<PointMatch> matches, float scale, float[] center); abstract public void minimize(Collection<PointMatch> matches); abstract public AffineTransform getAffine(); /** * string to output stream */ abstract public String toString(); /** * clone */ abstract public Model clone(); };