[图像处理] 请熟悉图像处理的帮忙一下
 |
hailong4688199
2012-02-22
我在网上下载了使用java实现的SIFT算法，但不知道如何使用。我的使用目的是能够抽取某一张图片的特征向量值，和比较匹配两张图片特征的相似度。最好能贴上代码，不胜感激！ 文件列表如下所示： Java_SIFT-src
如有需要可以继续贴上来，这里面的SIFT_Test.java代码是
/**
* Extract and display Scale Invariant Features after the method of David Lowe
* \cite{Lowe04} in an image.
*
* BibTeX:
* <pre>
* @article{Lowe04,
* author = {David G. Lowe},
* title = {Distinctive Image Features from Scale-Invariant Keypoints},
* journal = {International Journal of Computer Vision},
* year = {2004},
* volume = {60},
* number = {2},
* pages = {91--110},
* }
* </pre>
*
*
* License: GPL
*
* 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.
*
* 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.
*
* 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.
*
* 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
*/
import mpi.cbg.fly.*;
import ij.plugin.*;
import ij.gui.*;
import ij.*;
import ij.process.*;
import java.util.Collections;
import java.util.Vector;
import java.awt.Color;
import java.awt.Polygon;
import java.awt.TextField;
import java.awt.event.KeyEvent;
import java.awt.event.KeyListener;
public class SIFT_Test implements PlugIn, KeyListener
{
// steps
private static int steps = 3;
// initial sigma
private static float initial_sigma = 1.6f;
// feature descriptor size
private static int fdsize = 4;
// feature descriptor orientation bins
private static int fdbins = 8;
// size restrictions for scale octaves, use octaves < max_size and > min_size only
private static int min_size = 64;
private static int max_size = 1024;
/**
* Set true to double the size of the image by linear interpolation to
* ( with * 2 + 1 ) * ( height * 2 + 1 ). Thus we can start identifying
* DoG extrema with $\sigma = INITIAL_SIGMA / 2$ like proposed by
* \citet{Lowe04}.
*
* This is useful for images scmaller than 1000px per side only.
*/
private static boolean upscale = false;
private static float scale = 1.0f;
/**
* draws a rotated square with center point center, having size and orientation
*/
static void drawSquare( ImageProcessor ip, double[] o, double scale, double orient )
{
scale /= 2;
double sin = Math.sin( orient );
double cos = Math.cos( orient );
int[] x = new int[ 6 ];
int[] y = new int[ 6 ];
x[ 0 ] = ( int )( o[ 0 ] + ( sin - cos ) * scale );
y[ 0 ] = ( int )( o[ 1 ] - ( sin + cos ) * scale );
x[ 1 ] = ( int )o[ 0 ];
y[ 1 ] = ( int )o[ 1 ];
x[ 2 ] = ( int )( o[ 0 ] + ( sin + cos ) * scale );
y[ 2 ] = ( int )( o[ 1 ] + ( sin - cos ) * scale );
x[ 3 ] = ( int )( o[ 0 ] - ( sin - cos ) * scale );
y[ 3 ] = ( int )( o[ 1 ] + ( sin + cos ) * scale );
x[ 4 ] = ( int )( o[ 0 ] - ( sin + cos ) * scale );
y[ 4 ] = ( int )( o[ 1 ] - ( sin - cos ) * scale );
x[ 5 ] = x[ 0 ];
y[ 5 ] = y[ 0 ];
ip.drawPolygon( new Polygon( x, y, x.length ) );
}
public void run( String args )
{
if ( IJ.versionLessThan( "1.37i" ) ) return;
final ImagePlus imp = WindowManager.getCurrentImage();
if ( imp == null ) { System.err.println( "There are no images open" ); return; }
final GenericDialog gd = new GenericDialog( "Test SIFT" );
gd.addNumericField( "steps_per_scale_octave :", steps, 0 );
gd.addNumericField( "initial_gaussian_blur :", initial_sigma, 2 );
gd.addNumericField( "feature_descriptor_size :", fdsize, 0 );
gd.addNumericField( "feature_descriptor_orientation_bins :", fdbins, 0 );
gd.addNumericField( "minimum_image_size :", min_size, 0 );
gd.addNumericField( "maximum_image_size :", max_size, 0 );
gd.addCheckbox( "upscale_image_first", upscale );
gd.showDialog();
if ( gd.wasCanceled() ) return;
steps = ( int )gd.getNextNumber();
initial_sigma = ( float )gd.getNextNumber();
fdsize = ( int )gd.getNextNumber();
fdbins = ( int )gd.getNextNumber();
min_size = ( int )gd.getNextNumber();
max_size = ( int )gd.getNextNumber();
upscale = gd.getNextBoolean();
if ( upscale ) scale = 2.0f;
ImageProcessor ip1 = imp.getProcessor().convertToFloat();
ImageProcessor ip2 = imp.getProcessor().duplicate().convertToRGB();
Vector< Feature > fs1;
FloatArray2DSIFT sift = new FloatArray2DSIFT( fdsize, fdbins );
FloatArray2D fa = ImageArrayConverter.ImageToFloatArray2D( ip1 );
Filter.enhance( fa, 1.0f );
if ( upscale )
{
FloatArray2D fat = new FloatArray2D( fa.width * 2 - 1, fa.height * 2 - 1 );
FloatArray2DScaleOctave.upsample( fa, fat );
fa = fat;
fa = Filter.computeGaussianFastMirror( fa, ( float )Math.sqrt( initial_sigma * initial_sigma - 1.0 ) );
}
else
fa = Filter.computeGaussianFastMirror( fa, ( float )Math.sqrt( initial_sigma * initial_sigma - 0.25 ) );
long start_time = System.currentTimeMillis();
System.out.print( "processing SIFT ..." );
sift.init( fa, steps, initial_sigma, min_size, max_size );
fs1 = sift.run( max_size );
Collections.sort( fs1 );
System.out.println( " took " + ( System.currentTimeMillis() - start_time ) + "ms" );
System.out.println( fs1.size() + " features identified and processed" );
ip2.setLineWidth( 1 );
ip2.setColor( Color.red );
for ( Feature f : fs1 )
{
//System.out.println( f.location[ 0 ] + " " + f.location[ 1 ] + " " + f.scale + " " + f.orientation );
drawSquare( ip2, new double[]{ f.location[ 0 ] / scale, f.location[ 1 ] / scale }, fdsize * 4.0 * ( double )f.scale / scale, ( double )f.orientation );
}
ImagePlus imp1 = new ImagePlus( imp.getTitle() + " Features ", ip2 );
imp1.show();
}
public void keyPressed(KeyEvent e)
{
if (
( e.getKeyCode() == KeyEvent.VK_F1 ) &&
( e.getSource() instanceof TextField) )
{
}
}
public void keyReleased(KeyEvent e) { }
public void keyTyped(KeyEvent e) { }
}
|
|
hailong4688199
2012-02-22
以上的代码中使用了imageJ，如何使用的也请教
|
