/* * 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 * * Updated: 07.08.13 12:18 */ package net.semanticmetadata.lire.imageanalysis.features.global; import net.semanticmetadata.lire.builders.DocumentBuilder; import net.semanticmetadata.lire.imageanalysis.features.GlobalFeature; import net.semanticmetadata.lire.imageanalysis.features.LireFeature; import net.semanticmetadata.lire.utils.ImageUtils; import net.semanticmetadata.lire.utils.MetricsUtils; import net.semanticmetadata.lire.utils.SerializationUtils; import java.awt.image.BufferedImage; import java.awt.image.WritableRaster; /** * Simple 64 bin Opponent Histogram, based on the Opponent color space as described in van de Sande, Gevers & Snoek (2010) * "Evaluating Color Descriptors for Object and Scene Recognition", IEEE PAMI (see BibTeX in the source code). * * * @author Mathias Lux, mathias@juggle.at * Date: 18.12.12 * Time: 11:53 */ /* @ARTICLE{Sande2010, author={van de Sande, K.E.A. and Gevers, T. and Snoek, C.G.M.}, journal={Pattern Analysis and Machine Intelligence, IEEE Transactions on}, title={Evaluating Color Descriptors for Object and Scene Recognition}, year={2010}, month={sept. }, volume={32}, number={9}, pages={1582 -1596}, doi={10.1109/TPAMI.2009.154}, ISSN={0162-8828}, } */ public class OpponentHistogram implements GlobalFeature { final double sq2 = Math.sqrt(2d); final double sq6 = Math.sqrt(3d); final double sq3 = Math.sqrt(6d); double[] descriptor; double o1, o2, o3; double tmpVal, tmpSum; byte[] histogram = new byte[64]; @Override public void extract(BufferedImage bimg) { // check if it's (i) RGB and (ii) 8 bits per pixel. bimg = ImageUtils.get8BitRGBImage(bimg); // extract: double[] histogram = new double[64]; for (int i = 0; i < histogram.length; i++) { histogram[i] = 0; } WritableRaster raster = bimg.getRaster(); int[] px = new int[3*(raster.getHeight()-2)]; int colorPos; for (int x = 1; x < raster.getWidth() - 1; x++) { raster.getPixels(x, 1, 1, raster.getHeight()-2, px); for (int y = 0; y < raster.getHeight() - 2; y++) { o1 = (double) (px[y*3] - px[y*3+1]) / sq2; o2 = (double) (px[y*3] + px[y*3+1] - 2 * px[y*3+2]) / sq6; o3 = (double) (px[y*3] + px[y*3+1] + px[y*3+2]) / sq3; // Normalize ... easier to handle. o1 = (o1 + 255d / sq2) / (510d / sq2); o2 = (o2 + 510d / sq6) / (1020d / sq6); o3 = o3 / (3d * 255d / sq3); // get the array position. colorPos = (int) Math.min(Math.floor(o1 * 4d), 3d) + (int) Math.min(Math.floor(o2 * 4d), 3d) * 4 + (int) Math.min(3d, Math.floor(o3 * 4d)) * 4 * 4; histogram[colorPos]++; } } // normalize with max norm & quantize to [0,127]: double max = 0; for (int i = 0; i < histogram.length; i++) { max = Math.max(histogram[i], max); } for (int i = 0; i < histogram.length; i++) { this.histogram[i] = (byte) Math.floor(127d * (histogram[i] / max)); } } @Override public byte[] getByteArrayRepresentation() { byte[] result = new byte[histogram.length]; for (int i = 0; i < result.length; i++) { result[i] = histogram[i]; } return result; } @Override public void setByteArrayRepresentation(byte[] in) { for (int i = 0; i < histogram.length; i++) { histogram[i] = in[i]; } } @Override public void setByteArrayRepresentation(byte[] in, int offset, int length) { for (int i = 0; i < length; i++) { histogram[i] = in[i+offset]; } } @Override public double[] getFeatureVector() { return SerializationUtils.castToDoubleArray(histogram); } @Override public double getDistance(LireFeature feature) { if (!(feature instanceof OpponentHistogram)) throw new UnsupportedOperationException("Wrong descriptor."); return MetricsUtils.jsd(((OpponentHistogram) feature).histogram, histogram); } public double getDistance(byte[] h1, byte[] h2) { return getDistance(h1, 0, h1.length, h2, 0, h2.length); } /** * Jeffrey Divergence or Jensen-Shannon divergence (JSD) from * Deselaers, T.; Keysers, D. & Ney, H. Features for image retrieval: * an experimental comparison Inf. Retr., Kluwer Academic Publishers, 2008, 11, 77-107 * @param h1 * @param offset1 * @param length1 * @param h2 * @param offset2 * @param length2 * @return */ public double getDistance(byte[] h1, int offset1, int length1, byte[] h2, int offset2, int length2) { // double sum = 0f; // for (int i = 0; i < h1.length; i++) { // sum += (h1[i] > 0 ? (h1[i] / 2f) * Math.log((2f * h1[i]) / (h1[i] + h2[i])) : 0) + // (h2[i] > 0 ? (h2[i] / 2f) * Math.log((2f * h2[i]) / (h1[i] + h2[i])) : 0); // } // return (float) sum; tmpSum = 0; for (int i = 0; i < length1; i++) { tmpVal = (double) (h1[i+offset1] + h2[i+offset2]); tmpSum += (h1[i+offset1] > 0 ? ((double) h1[i+offset1] / 2d) * Math.log((2d * h1[i+offset1]) / tmpVal) : 0) + (h2[i+offset2] > 0 ? ((double) h2[i+offset2] / 2d) * Math.log((2d * h2[i+offset2]) / tmpVal) : 0); } return tmpSum; } // public String getStringRepresentation() { // StringBuilder sb = new StringBuilder(histogram.length * 2 + 25); // sb.append("ophist"); // sb.append(' '); // sb.append(histogram.length); // sb.append(' '); // for (double aData : histogram) { // sb.append((int) aData); // sb.append(' '); // } // return sb.toString().trim(); // } // // public void setStringRepresentation(String s) { // StringTokenizer st = new StringTokenizer(s); // if (!st.nextToken().equals("ophist")) // throw new UnsupportedOperationException("This is not a OpponentHistogram descriptor."); // for (int i = 0; i < histogram.length; i++) { // if (!st.hasMoreTokens()) // throw new IndexOutOfBoundsException("Too few numbers in string representation."); // histogram[i] = (byte) Integer.parseInt(st.nextToken()); // } // // } @Override public String getFeatureName() { return "OpponentHistogram"; } @Override public String getFieldName() { return DocumentBuilder.FIELD_NAME_OPPONENT_HISTOGRAM; } }