Facemorph
Class Transformer

java.lang.Object
  Facemorph.Transformer

public class Transformer
extends java.lang.Object

Class to perform prototype (average) based transformations on images.

Constructor Summary

Transformer()

Method Summary

static java.awt.Image	averageImages(java.awt.Image[] images, Template[] templates, Template averageTemplate, int outW, int outH, java.awt.image.ImageObserver obs, boolean autoSize, int normalisation) Average a set of images
static java.awt.Image	averageImages(java.awt.Image[] images, Template[] templates, Template averageTemplate, int outW, int outH, java.awt.image.ImageObserver obs, boolean autoSize, int normalisation, TransformerListener warpListener, TransformerListener averageListener) Method to create an average from a set of delineated images.
static java.awt.Image	averageImagesTexture(java.awt.Image[] images, Template[] templates, Template averageTemplate, int outW, int outH, Filter[] filters, int bm, java.awt.image.ImageObserver obs, boolean autoSize, int normalisation) Method to create an average from a set of delineated images.
static java.awt.Image	averageImagesTexture(java.awt.Image[] images, Template[] templates, Template averageTemplate, int outW, int outH, Filter[] filters, int bm, java.awt.image.ImageObserver obs, boolean autoSize, int normalisation, TransformerListener warpListener, TransformerListener averageListener) Method to create an average from a set of delineated images.
static java.awt.image.BufferedImage	caricatureChimeric(java.awt.image.BufferedImage subImg, java.awt.image.BufferedImage cimg1, java.awt.image.BufferedImage cimg2, Template tmplt, float l1, float l2, float wid) Half face transform on the image (no warping)
static java.awt.image.BufferedImage	ImageToBufferedImage(java.awt.Image img, java.awt.image.ImageObserver iob) Converts an abstract (unknown type of) Image to a BufferedImage for extra functionality.
static void	mask(Template newTemplate, Mask mask, java.awt.image.BufferedImage warped, java.awt.image.BufferedImage transformed) Mask the colour changes to the array specified
static java.util.Vector<java.awt.Image>	morph(java.awt.Image img1, Template tmp1, java.awt.Image img2, Template tmp2, int steps, java.awt.image.ImageObserver ob, boolean fast) Creates a Morph between two delineated images
static java.util.Vector<java.awt.Image>	morph(java.awt.Image img1, Template tmp1, java.awt.Image img2, Template tmp2, int steps, java.awt.image.ImageObserver ob, boolean fast, java.util.Vector<Template> templates, TransformerListener listener) Creates a Morph between two delineated images
static void	normaliseColours(Template newTemplate, Mask mask, java.awt.image.BufferedImage warped, java.awt.image.BufferedImage transformed) Normalise the mean and standard deviation of colours in the masked area
static java.awt.image.BufferedImage	shiftColours(java.awt.image.BufferedImage subImg, java.awt.image.BufferedImage src, java.awt.image.BufferedImage dst, double amount) Perform the colour shifting part of the transform
static java.awt.image.BufferedImage	testTransform(int warpType, Template subTem, Template srcTem, Template dstTem, Template outTem, java.awt.image.BufferedImage subImg, java.awt.image.BufferedImage srcImg, java.awt.image.BufferedImage dstImg, double shapeAmount, double colourAmount, double textureAmount, Filter[] filters, int bm, Mask mask, boolean samples, boolean borders, boolean matchCols, int normalisation, int[] normPoints) Face transform method, more modular version
static java.awt.image.BufferedImage	testTransform(int warpType, Template subTem, Template srcTem, Template dstTem, Template outTem, java.awt.Image subImg, java.awt.Image srcImg, java.awt.Image dstImg, double shapeAmount, double colourAmount, double textureAmount, Filter[] filters, int bm, Mask mask, boolean samples, boolean borders, boolean matchCols, int normalisation, int[] normPoints) Face transform method, more modular version
static java.awt.image.BufferedImage	textureTransform(java.awt.image.BufferedImage orig, java.awt.image.BufferedImage sub, java.awt.image.BufferedImage src, java.awt.image.BufferedImage dst, Filter[] filters, int bm, double amount) Performs a texture (wavelet magnitude) transform on the image
static java.awt.Image	transform(Template subjectTemplate, Template sourceTemplate, Template destTemplate, Template newTemplate, java.awt.Image subjectImg, java.awt.Image sourceImg, java.awt.Image destImg, double shape, double colour, double texture, java.awt.image.ImageObserver ob, Mask mask, Filter[] filters, int bm, boolean fast, boolean conts) Transforms subjectImage using the prototypes given and returns the result.
static java.awt.Image	transform(Template subjectTemplate, Template sourceTemplate, Template destTemplate, Template newTemplate, java.awt.Image subjectImg, java.awt.Image sourceImg, java.awt.Image destImg, double s, java.awt.image.ImageObserver ob, boolean fast) Transforms subjectImage using the prototypes given and returns the result.
static java.awt.Image	transform(Template subjectTemplate, Template sourceTemplate, Template destTemplate, Template newTemplate, java.awt.Image subjectImg, java.awt.Image sourceImg, java.awt.Image destImg, double s, java.awt.image.ImageObserver ob, Mask mask, boolean fast) Transforms subjectImage using the prototypes given and returns the result.
static java.awt.Image	transformBald(Template subjectTemplate, Template avTemplate, Template rotAvTemplate, Template baldAvTemplate, Template rotBaldAvTemplate, Template hairAvTemplate, Template rotHairAvTemplate, Template newTemplate, java.awt.Image subjectImg, java.awt.Image sourceImg, java.awt.Image destImg, double s, java.awt.image.ImageObserver ob, boolean colour, boolean fast) Transforms subjectImage using the prototypes given and returns the result.
static java.awt.image.BufferedImage	transformChimeric(int warpType, Template subTem, Template srcTem, Template dstTem, Template outTem, java.awt.image.BufferedImage subImg, java.awt.image.BufferedImage srcImg, java.awt.image.BufferedImage dstImg, float leftAmount, float rightAmount, float width, boolean samples, boolean borders) Half face transform method
static java.awt.image.BufferedImage	warp(int warpType, java.awt.image.BufferedImage bimg, Template start, Template end, int outW, int outH, boolean samples, boolean borders) Image warp method

Methods inherited from class java.lang.Object

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

Constructor Detail

Transformer

public Transformer()

Method Detail

transform

public static java.awt.Image transform(Template subjectTemplate,
                                       Template sourceTemplate,
                                       Template destTemplate,
                                       Template newTemplate,
                                       java.awt.Image subjectImg,
                                       java.awt.Image sourceImg,
                                       java.awt.Image destImg,
                                       double s,
                                       java.awt.image.ImageObserver ob,
                                       boolean fast)

Transforms subjectImage using the prototypes given and returns the result. The transformed Template is returned in newTemplate. The transformation follows the formula: new = subject + s*(dest-source)

Parameters:

subjectTemplate - The Template outlining the features of subjectImg

sourceTemplate - The (average) Template outlining the features of sourceImg

destTemplate - The (average) Template outlining the features of destImg

newTemplate - Template to hold the Template corresponding to the transformed result. Useful for morphing afterwards.

subjectImg - The Image to transform

sourceImg - The (average) image of the source group

destImg - The (average) image of the destination group

s - The amount to scale this transform by. 0 = no transform, 1= complete transform.

ob - An ImageObserver used for obtaining Image sizes

fast - Flag indicating whether to use (fast?) linear warping with no sampling along the contours, or (slow?) multi-scale warping with full contour sampling.

Returns:

The transformed image.

transform

public static java.awt.Image transform(Template subjectTemplate,
                                       Template sourceTemplate,
                                       Template destTemplate,
                                       Template newTemplate,
                                       java.awt.Image subjectImg,
                                       java.awt.Image sourceImg,
                                       java.awt.Image destImg,
                                       double shape,
                                       double colour,
                                       double texture,
                                       java.awt.image.ImageObserver ob,
                                       Mask mask,
                                       Filter[] filters,
                                       int bm,
                                       boolean fast,
                                       boolean conts)

Transforms subjectImage using the prototypes given and returns the result. The transformed Template is returned in newTemplate. The transformation follows the formula: new = subject + s*(dest-source)

Parameters:

mask - The Mask object defining the boundary of the masked region.

subjectTemplate - The Template outlining the features of subjectImg

sourceTemplate - The (average) Template outlining the features of sourceImg

destTemplate - The (average) Template outlining the features of destImg

newTemplate - Template to hold the Template corresponding to the transformed result. Useful for morphing afterwards.

subjectImg - The Image to transform

sourceImg - The (average) image of the source group

destImg - The (average) image of the destination group

shape - The amount to scale the shape transform by. 0 = no transform, 1= complete transform.

colour - The amount to scale the colour transform by. 0 = no transform, 1= complete transform.

texture - The amount to scale the texture transform by. 0 = no transform, 1= complete transform.

ob - An ImageObserver used for obtaining Image sizes

filters - the filters to use if doint a texture transform

fast - Flag indicating whether to use (fast?) linear warping with no sampling along the contours, or (slow?) multi-scale warping with full contour sampling.

bm - the border model of the wavelet analysis filters (1 symmetric, -1 anti-sym), if doing a texture transform

conts - flag indicating that contours should be sampled along (rather than just using template points)

Returns:

The transformed image.

transform

public static java.awt.Image transform(Template subjectTemplate,
                                       Template sourceTemplate,
                                       Template destTemplate,
                                       Template newTemplate,
                                       java.awt.Image subjectImg,
                                       java.awt.Image sourceImg,
                                       java.awt.Image destImg,
                                       double s,
                                       java.awt.image.ImageObserver ob,
                                       Mask mask,
                                       boolean fast)

Transforms subjectImage using the prototypes given and returns the result. The transformed Template is returned in newTemplate. The transformation follows the formula: new = subject + s*(dest-source)

Parameters:

mask - The Mask object defining the boundary of the masked region.

subjectTemplate - The Template outlining the features of subjectImg

sourceTemplate - The (average) Template outlining the features of sourceImg

destTemplate - The (average) Template outlining the features of destImg

newTemplate - Template to hold the Template corresponding to the transformed result. Useful for morphing afterwards.

subjectImg - The Image to transform

sourceImg - The (average) image of the source group

destImg - The (average) image of the destination group

s - The amount to scale this transform by. 0 = no transform, 1= complete transform.

ob - An ImageObserver used for obtaining Image sizes

fast - Flag indicating whether to use (fast?) linear warping with no sampling along the contours, or (slow?) multi-scale warping with full contour sampling.

Returns:

The transformed image.

testTransform

public static java.awt.image.BufferedImage testTransform(int warpType,
                                                         Template subTem,
                                                         Template srcTem,
                                                         Template dstTem,
                                                         Template outTem,
                                                         java.awt.Image subImg,
                                                         java.awt.Image srcImg,
                                                         java.awt.Image dstImg,
                                                         double shapeAmount,
                                                         double colourAmount,
                                                         double textureAmount,
                                                         Filter[] filters,
                                                         int bm,
                                                         Mask mask,
                                                         boolean samples,
                                                         boolean borders,
                                                         boolean matchCols,
                                                         int normalisation,
                                                         int[] normPoints)

Face transform method, more modular version

Parameters:

warpType - the type of warp to use

subTem - the subject's Template

srcTem - the source group Template

dstTem - the target group Template

outTem - the output Template

subImg - the subject image

srcImg - the source group image

dstImg - the target group image

shapeAmount - the amount to transform the shape by

colourAmount - the amount to transform the colour by

textureAmount - the amount to transform the texture by

filters - the filters to use for texture transformation

bm - the border model (sym=1, anti-sym=-1) to use for the filters

mask - the mask to use (if any, or null)

samples - sample along template contours

borders - pin down the borders

matchCols - match the colours on the masked area

normalisation - indicates the kind of normalisation to use, one of ASM.RIGID_BODY_NORMALISATION, ASM.TWO_POINT_NORMALISATION, ASM.THREE_POINT_NORMALISATION or ASM.NO_NORMALISATION

normPoints - the normalisation points to use for ASM.TWO_POINT_NORMALISATION or ASM.THREE_POINT_NORMALISATION

Returns:

return the transformed image

See Also:

Warp.createWarp(int, int, int, int, int, boolean)

testTransform

public static java.awt.image.BufferedImage testTransform(int warpType,
                                                         Template subTem,
                                                         Template srcTem,
                                                         Template dstTem,
                                                         Template outTem,
                                                         java.awt.image.BufferedImage subImg,
                                                         java.awt.image.BufferedImage srcImg,
                                                         java.awt.image.BufferedImage dstImg,
                                                         double shapeAmount,
                                                         double colourAmount,
                                                         double textureAmount,
                                                         Filter[] filters,
                                                         int bm,
                                                         Mask mask,
                                                         boolean samples,
                                                         boolean borders,
                                                         boolean matchCols,
                                                         int normalisation,
                                                         int[] normPoints)

Face transform method, more modular version

Parameters:

warpType - the type of warp to use

subTem - the subject's Template

srcTem - the source group Template

dstTem - the target group Template

outTem - the output Template

subImg - the subject image

srcImg - the source group image

dstImg - the target group image

shapeAmount - the amount to transform the shape by

colourAmount - the amount to transform the colour by

textureAmount - the amount to transform the texture by

filters - the filters to use for texture transformation

bm - the border model (sym=1, anti-sym=-1) to use for the filters

mask - the mask to use (if any, or null)

samples - sample along template contours

borders - pin down the borders

matchCols - match the colours on the masked area

normalisation - indicates the kind of normalisation to use, one of ASM.RIGID_BODY_NORMALISATION, ASM.TWO_POINT_NORMALISATION, ASM.THREE_POINT_NORMALISATION or ASM.NO_NORMALISATION

normPoints - the normalisation points to use for ASM.TWO_POINT_NORMALISATION or ASM.THREE_POINT_NORMALISATION

Returns:

return the transformed image

See Also:

Warp.createWarp(int, int, int, int, int, boolean)

caricatureChimeric

public static java.awt.image.BufferedImage caricatureChimeric(java.awt.image.BufferedImage subImg,
                                                              java.awt.image.BufferedImage cimg1,
                                                              java.awt.image.BufferedImage cimg2,
                                                              Template tmplt,
                                                              float l1,
                                                              float l2,
                                                              float wid)

Half face transform on the image (no warping)

Parameters:

subImg - the subject image

cimg1 - the source image

cimg2 - the target image

tmplt - the shape template

l1 - the amount to transform the left

l2 - the amount to transform the right

wid - the width of the smoothing band

Returns:

return the transformed image

See Also:

transformChimeric(int, Facemorph.Template, Facemorph.Template, Facemorph.Template, Facemorph.Template, java.awt.image.BufferedImage, java.awt.image.BufferedImage, java.awt.image.BufferedImage, float, float, float, boolean, boolean)

transformChimeric

public static java.awt.image.BufferedImage transformChimeric(int warpType,
                                                             Template subTem,
                                                             Template srcTem,
                                                             Template dstTem,
                                                             Template outTem,
                                                             java.awt.image.BufferedImage subImg,
                                                             java.awt.image.BufferedImage srcImg,
                                                             java.awt.image.BufferedImage dstImg,
                                                             float leftAmount,
                                                             float rightAmount,
                                                             float width,
                                                             boolean samples,
                                                             boolean borders)

Half face transform method

Parameters:

warpType - the type of warp to use

subTem - the subject Template

srcTem - the source group Template

dstTem - the target group Template

outTem - the output Template

subImg - the subject Image

srcImg - the source group image

dstImg - the target group image

leftAmount - the amount to transform the left side

rightAmount - the amount to transform the right side

width - the width of the smoothing band

samples - if true sample along Template contours

borders - add points to pin down the borders

Returns:

returns the transformed image

See Also:

Warp.createWarp(int, int, int, int, int, boolean)

mask

public static void mask(Template newTemplate,
                        Mask mask,
                        java.awt.image.BufferedImage warped,
                        java.awt.image.BufferedImage transformed)

Mask the colour changes to the array specified

Parameters:

newTemplate - the template that defines the mask

mask - the mask labels

warped - the warped image

transformed - the transformed image

normaliseColours

public static void normaliseColours(Template newTemplate,
                                    Mask mask,
                                    java.awt.image.BufferedImage warped,
                                    java.awt.image.BufferedImage transformed)

Normalise the mean and standard deviation of colours in the masked area

Parameters:

newTemplate - the transformed template

mask - the mask labels

warped - the warped image

transformed - the transformed image

textureTransform

public static java.awt.image.BufferedImage textureTransform(java.awt.image.BufferedImage orig,
                                                            java.awt.image.BufferedImage sub,
                                                            java.awt.image.BufferedImage src,
                                                            java.awt.image.BufferedImage dst,
                                                            Filter[] filters,
                                                            int bm,
                                                            double amount)

Performs a texture (wavelet magnitude) transform on the image

Parameters:

orig - the warped original image

sub - the warped and colour transformed original image

src - the source group image

dst - the target group image

filters - the filters to use

bm - the border model to use

amount - the amount to transform the textures

Returns:

return the transformed image

warp

public static java.awt.image.BufferedImage warp(int warpType,
                                                java.awt.image.BufferedImage bimg,
                                                Template start,
                                                Template end,
                                                int outW,
                                                int outH,
                                                boolean samples,
                                                boolean borders)

Image warp method

Parameters:

warpType - the type of warp to use

bimg - the original image

start - the original Template

end - the target Template

outW - the output width

outH - the output height

samples - indicates if contours should be sampled

borders - indicates if the borders should be pinned down

Returns:

return the warped image

shiftColours

public static java.awt.image.BufferedImage shiftColours(java.awt.image.BufferedImage subImg,
                                                        java.awt.image.BufferedImage src,
                                                        java.awt.image.BufferedImage dst,
                                                        double amount)

Perform the colour shifting part of the transform

Parameters:

subImg - the warped original image

src - the warped source group image

dst - the warped target group image

amount - the amount to transform the colours by

Returns:

returns the colour transformed image

morph

public static java.util.Vector<java.awt.Image> morph(java.awt.Image img1,
                                                     Template tmp1,
                                                     java.awt.Image img2,
                                                     Template tmp2,
                                                     int steps,
                                                     java.awt.image.ImageObserver ob,
                                                     boolean fast)

Creates a Morph between two delineated images

Parameters:

img1 - The image to start from

tmp1 - The Template outlining the features of img1

img2 - The image to end at

tmp2 - The template outlining the features of img2

steps - The number of steps between the images i.e. the output list will contain steps+1 images

ob - AN ImageObserver, used for finding Image dimensions

fast - Flag to indicate if (fast?) linear warping with no contour sampling or (slow?) non-linear warping with full contour sampling.

Returns:

A Vector of Image objects representing the morph, includes the start and end points.

morph

public static java.util.Vector<java.awt.Image> morph(java.awt.Image img1,
                                                     Template tmp1,
                                                     java.awt.Image img2,
                                                     Template tmp2,
                                                     int steps,
                                                     java.awt.image.ImageObserver ob,
                                                     boolean fast,
                                                     java.util.Vector<Template> templates,
                                                     TransformerListener listener)

Creates a Morph between two delineated images

Parameters:

img1 - The image to start from

tmp1 - The Template outlining the features of img1

img2 - The image to end at

tmp2 - The template outlining the features of img2

steps - The number of steps between the images i.e. the output list will contain steps+1 images

ob - AN ImageObserver, used for finding Image dimensions

fast - Flag to indicate if (fast?) linear warping with no contour sampling or (slow?) non-linear warping with full contour sampling.

templates - Vector to hold output array of templates

listener - used for demonstrating output

Returns:

A Vector of Image objects representing the morph, includes the start and end points.

averageImages

public static java.awt.Image averageImages(java.awt.Image[] images,
                                           Template[] templates,
                                           Template averageTemplate,
                                           int outW,
                                           int outH,
                                           java.awt.image.ImageObserver obs,
                                           boolean autoSize,
                                           int normalisation)

Average a set of images

Parameters:

images - the images to average

templates - the templates specifying the location and shape of each face

averageTemplate - the average face shape

outW - the output width

outH - the output height

obs - an image observer

autoSize - flag indicating if the average should be auto resized to fit in the output image size

normalisation - the type of normalisation to use ASM.RIGID_BODY_NORMALISATION, ASM.TWO_POINT_NORMALISATION, ASM.NO_NORMALISATION

Returns:

returns the average image

averageImages

public static java.awt.Image averageImages(java.awt.Image[] images,
                                           Template[] templates,
                                           Template averageTemplate,
                                           int outW,
                                           int outH,
                                           java.awt.image.ImageObserver obs,
                                           boolean autoSize,
                                           int normalisation,
                                           TransformerListener warpListener,
                                           TransformerListener averageListener)

Method to create an average from a set of delineated images.

Parameters:

images - The array of images to be averaged.

templates - The array of templates outlining the features of the corresponding templates.

averageTemplate - The template in which the average shape description will be placed.

outW - The desired width of the output image.

outH - The desired height of the output image.

obs - An image observer used for getting the height and width of the image

normalisation - the type of normalisation to use ASM.RIGID_BODY_NORMALISATION, ASM.TWO_POINT_NORMALISATION, ASM.NO_NORMALISATION

autoSize - automatically resize the template to fill the output image

warpListener - used to display the process

averageListener - used to display the average emerging

Returns:

An int array of pixel values representing the average image.

averageImagesTexture

public static java.awt.Image averageImagesTexture(java.awt.Image[] images,
                                                  Template[] templates,
                                                  Template averageTemplate,
                                                  int outW,
                                                  int outH,
                                                  Filter[] filters,
                                                  int bm,
                                                  java.awt.image.ImageObserver obs,
                                                  boolean autoSize,
                                                  int normalisation)

Method to create an average from a set of delineated images.

Parameters:

images - The array of images to be averaged.

templates - The array of templates outlining the features of the corresponding templates.

averageTemplate - The template in which the average shape description will be placed.

outW - The desired width of the output image.

outH - The desired height of the output image.

filters - the filters to use

bm - the border model to use for the filters

obs - An image observer used for getting the height and width of the image

autoSize - indicates that the template should be expanded to fill the output image

normalisation - the type of normalisation to use ASM.RIGID_BODY_NORMALISATION, ASM.TWO_POINT_NORMALISATION, ASM.NO_NORMALISATION

Returns:

An int array of pixel values representing the average image.

averageImagesTexture

public static java.awt.Image averageImagesTexture(java.awt.Image[] images,
                                                  Template[] templates,
                                                  Template averageTemplate,
                                                  int outW,
                                                  int outH,
                                                  Filter[] filters,
                                                  int bm,
                                                  java.awt.image.ImageObserver obs,
                                                  boolean autoSize,
                                                  int normalisation,
                                                  TransformerListener warpListener,
                                                  TransformerListener averageListener)

Method to create an average from a set of delineated images.

Parameters:

images - The array of images to be averaged.

templates - The array of templates outlining the features of the corresponding templates.

averageTemplate - The template in which the average shape description will be placed.

outW - The desired width of the output image.

outH - The desired height of the output image.

filters - the filters to use

bm - the border model to use for the filters

obs - An image observer used for getting the height and width of the image

averageListener - used to display the averaging process

autoSize - resize the average template to fill the output image

normalisation - the type of normalisation to use ASM.RIGID_BODY_NORMALISATION, ASM.TWO_POINT_NORMALISATION, ASM.NO_NORMALISATION

warpListener - used to display the warping process

Returns:

An int array of pixel values representing the average image.

ImageToBufferedImage

public static java.awt.image.BufferedImage ImageToBufferedImage(java.awt.Image img,
                                                                java.awt.image.ImageObserver iob)

Converts an abstract (unknown type of) Image to a BufferedImage for extra functionality.

Parameters:

img - The image to convert

iob - An image observer (for getting the image size etc)

Returns:

The Image as a BufferedImage

transformBald

public static java.awt.Image transformBald(Template subjectTemplate,
                                           Template avTemplate,
                                           Template rotAvTemplate,
                                           Template baldAvTemplate,
                                           Template rotBaldAvTemplate,
                                           Template hairAvTemplate,
                                           Template rotHairAvTemplate,
                                           Template newTemplate,
                                           java.awt.Image subjectImg,
                                           java.awt.Image sourceImg,
                                           java.awt.Image destImg,
                                           double s,
                                           java.awt.image.ImageObserver ob,
                                           boolean colour,
                                           boolean fast)

Transforms subjectImage using the prototypes given and returns the result. The transformed Template is returned in newTemplate. The transformation follows the formula: new = subject + s*(dest-source)

Parameters:

avTemplate - The average template

rotAvTemplate - The rotated average template

baldAvTemplate - the face-on bald average template

rotBaldAvTemplate - the rotated bald average template

hairAvTemplate - the face-on with hair average template

rotHairAvTemplate - the rotated with-hair average template

newTemplate - The output template, delineatng the transformed image, useful for further processing.

colour - Flag indicating if to change the colours, false implies just the shape is changed

subjectTemplate - The Template outlining the features of subjectImg

subjectImg - The Image to transform

sourceImg - The (average) image of the source group

destImg - The (average) image of the destination group

s - The amount to scale this transform by. 0 = no transform, 1= complete transform.

ob - An ImageObserver used for obtaining Image sizes

fast - Flag indicating whether to use (fast?) linear warping with no sampling along the contours, or (slow?) multi-scale warping with full contour sampling.

Returns:

The transformed image.