Facemorph.aam
Class CLMSVM

java.lang.Object
  Facemorph.aam.CLMSVM

public class CLMSVM
extends java.lang.Object

Constrained local model with SVM based patch detectors

Constructor Summary

CLMSVM()

Method Summary

void	build(java.lang.String file, HaarReader haar) Builds an SVM based CLM model
Template	delineate(java.awt.image.BufferedImage bimg, HaarReader haar, Template template, ImageZoomPanel izp) Delineate am image using this CLMSVM using the convex quadratic fitting (CQF) method
FloatImage[]	getStrip(Template tem, FloatImage fimg, java.awt.Rectangle rect, int minOffset, int maxOffset) Gets the patches from all points arranged in a strip
static void	main(java.lang.String[] args) Main method used for generating a random split of the data
boolean	processPatches(java.util.ArrayList<FloatImage[]> patches, java.util.ArrayList<FloatImage[]> negativePatches, boolean doCV) Build the SVM model from the patches
void	read(java.lang.String file) Reads PCI data from a directory
void	write(java.lang.String file) Writes CLMSVM data to a directory

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

CLMSVM

public CLMSVM()

Method Detail

build

public void build(java.lang.String file,
                  HaarReader haar)
           throws java.io.FileNotFoundException,
                  java.io.IOException

Builds an SVM based CLM model

Parameters:

file - the name of the file listing the training images

haar - the face detector

Throws:

java.io.FileNotFoundException

java.io.IOException

delineate

public Template delineate(java.awt.image.BufferedImage bimg,
                          HaarReader haar,
                          Template template,
                          ImageZoomPanel izp)
                   throws PowellException

Delineate am image using this CLMSVM using the convex quadratic fitting (CQF) method

Parameters:

bimg - the image to delineate

haar - the face detector

template - used to initialise if not null

izp - the image zoom panel to use if animating the searcg

Returns:

returns the estimated template

Throws:

PowellException

getStrip

public FloatImage[] getStrip(Template tem,
                             FloatImage fimg,
                             java.awt.Rectangle rect,
                             int minOffset,
                             int maxOffset)

Gets the patches from all points arranged in a strip

Parameters:

tem - the current template

fimg - the image to sample the patches from

rect - the face detection rectangle, if not null this indicates that the priors should be updated

minOffset - the minimum offset used when building an enriched training set

maxOffset - the maximum offset used when building an enriched training set

Returns:

returns the patch strip image

processPatches

public boolean processPatches(java.util.ArrayList<FloatImage[]> patches,
                              java.util.ArrayList<FloatImage[]> negativePatches,
                              boolean doCV)
                       throws java.io.IOException

Build the SVM model from the patches

Parameters:

patches - the positive example patches

negativePatches - the negative example patches

doCV - perform grid based cross validation

Returns:

returns true if the model was built OK

Throws:

java.io.IOException

read

public void read(java.lang.String file)
          throws java.io.IOException

Reads PCI data from a directory

Parameters:

file - The text file containing the names of the image component files and their variance

Throws:

java.io.IOException - if the file can't be read

write

public void write(java.lang.String file)
           throws java.io.IOException

Writes CLMSVM data to a directory

Parameters:

file - The text file containing the names of the image component files and their variance

Throws:

java.io.IOException - throws IOException if problem writing file

main

public static void main(java.lang.String[] args)
                 throws java.io.FileNotFoundException,
                        java.io.IOException

Main method used for generating a random split of the data

Parameters:

args - args[0] = name of file to read, args[1] = fraction to put in file1, args[2] = output file1, args[3] = output file2

Throws:

java.io.FileNotFoundException

java.io.IOException