The effect of texture representations on AAM performance

P. Kittipanya-ngam; T. F. Cootes

doi:10.1109/ICPR.2006.1117

The effect of texture representations on AAM performance

P. Kittipanya-ngam, T. F. Cootes

Research output: Chapter in Book/Conference proceeding › Conference contribution › peer-review

Abstract

The Active Appearance Model (AAM) algorithm matches statistical models of shape and texture to images rapidly by assuming a linear relationship between the texture residual and changes in the model parameters. When the texture is represented as raw intensity values, this has been shown to be a reasonable approximation in many cases. However, models built on them are sensitive to changes in illumination conditions. This paper examines the effect of using different representations of image texture to improve the accuracy and robustness of the AAM search. We show that normalising the gradient images by non-linear techniques can give much improved matching with higher accuracy and a wider effective range of convergence. © 2006 IEEE.

Original language	English
Title of host publication	Proceedings - International Conference on Pattern Recognition\|Proc. Int. Conf. Pattern Recognit.
Pages	328-331
Number of pages	3
Volume	2
DOIs	https://doi.org/10.1109/ICPR.2006.1117
Publication status	Published - 2006
Event	18th International Conference on Pattern Recognition, ICPR 2006 - Hong Kong Duration: 1 Jul 2006 → …

Conference

Conference	18th International Conference on Pattern Recognition, ICPR 2006
City	Hong Kong
Period	1/07/06 → …

Access to Document

10.1109/ICPR.2006.1117

Cite this

@inproceedings{164ba947bede425c85d7e85a81c987d3,

title = "The effect of texture representations on AAM performance",

abstract = "The Active Appearance Model (AAM) algorithm matches statistical models of shape and texture to images rapidly by assuming a linear relationship between the texture residual and changes in the model parameters. When the texture is represented as raw intensity values, this has been shown to be a reasonable approximation in many cases. However, models built on them are sensitive to changes in illumination conditions. This paper examines the effect of using different representations of image texture to improve the accuracy and robustness of the AAM search. We show that normalising the gradient images by non-linear techniques can give much improved matching with higher accuracy and a wider effective range of convergence. {\textcopyright} 2006 IEEE.",

author = "P. Kittipanya-ngam and Cootes, {T. F.}",

year = "2006",

doi = "10.1109/ICPR.2006.1117",

language = "English",

isbn = "0769525210",

volume = "2",

pages = "328--331",

booktitle = "Proceedings - International Conference on Pattern Recognition|Proc. Int. Conf. Pattern Recognit.",

note = "18th International Conference on Pattern Recognition, ICPR 2006 ; Conference date: 01-07-2006",

}

TY - GEN

T1 - The effect of texture representations on AAM performance

AU - Kittipanya-ngam, P.

AU - Cootes, T. F.

PY - 2006

Y1 - 2006

N2 - The Active Appearance Model (AAM) algorithm matches statistical models of shape and texture to images rapidly by assuming a linear relationship between the texture residual and changes in the model parameters. When the texture is represented as raw intensity values, this has been shown to be a reasonable approximation in many cases. However, models built on them are sensitive to changes in illumination conditions. This paper examines the effect of using different representations of image texture to improve the accuracy and robustness of the AAM search. We show that normalising the gradient images by non-linear techniques can give much improved matching with higher accuracy and a wider effective range of convergence. © 2006 IEEE.

AB - The Active Appearance Model (AAM) algorithm matches statistical models of shape and texture to images rapidly by assuming a linear relationship between the texture residual and changes in the model parameters. When the texture is represented as raw intensity values, this has been shown to be a reasonable approximation in many cases. However, models built on them are sensitive to changes in illumination conditions. This paper examines the effect of using different representations of image texture to improve the accuracy and robustness of the AAM search. We show that normalising the gradient images by non-linear techniques can give much improved matching with higher accuracy and a wider effective range of convergence. © 2006 IEEE.

U2 - 10.1109/ICPR.2006.1117

DO - 10.1109/ICPR.2006.1117

M3 - Conference contribution

SN - 0769525210

SN - 9780769525211

VL - 2

SP - 328

EP - 331

BT - Proceedings - International Conference on Pattern Recognition|Proc. Int. Conf. Pattern Recognit.

T2 - 18th International Conference on Pattern Recognition, ICPR 2006

Y2 - 1 July 2006

ER -

The effect of texture representations on AAM performance

Abstract

Conference

Access to Document

Fingerprint

Cite this