Homogeneous eigenspace for photometric analyses and recognition

Takeshi Shakunaga, Fumihiko Sakaue

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

Eigenspaces are often utilized in computer vision for various applications. However, careful manipulations are required for sensitive applications in photometric analyses and human interface applications. While several formulations have already been utilized in each application, a unified framework of eigenspace manipulations should be created for further progress of computer vision. This paper provides a concept of homogeneous eigenspaces for this purpose. After brief discussion of eigenspace construction, this paper focuses on how an optimal projection of an input image should be done. Three approaches are presented for the projection and normalization in order to minimize the effects of noise. Both the theoretical and experimental comparisons show that the normalization after projection can suppress the effects of noise more effectively than the other methods. The normalization-after-projection formulation can be reformulated by the homogeneous eigenspace. In the framework of homogeneous eigenspace, a projection onto the normalized eigenspace is accomplished without using an explicit normalization. Furthermore, an optimal partial projection to the normalized eigenspace is also reduced to linear mapping in this framework. Two applications are also discussed for illumination analysis and for photometric classification. These problems are efficiently solved by partial projections onto homogeneous eigenspaces.

Original languageEnglish
Title of host publicationProceedings of the 5th International Conference on Image and Graphics, ICIG 2009
Pages398-405
Number of pages8
DOIs
Publication statusPublished - 2010
Event5th International Conference on Image and Graphics, ICIG 2009 - Xi'an, Shanxi, China
Duration: Sep 20 2009Sep 23 2009

Other

Other5th International Conference on Image and Graphics, ICIG 2009
CountryChina
CityXi'an, Shanxi
Period9/20/099/23/09

Fingerprint

Computer vision
Lighting

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Computer Vision and Pattern Recognition

Cite this

Shakunaga, T., & Sakaue, F. (2010). Homogeneous eigenspace for photometric analyses and recognition. In Proceedings of the 5th International Conference on Image and Graphics, ICIG 2009 (pp. 398-405). [5437887] https://doi.org/10.1109/ICIG.2009.139

Homogeneous eigenspace for photometric analyses and recognition. / Shakunaga, Takeshi; Sakaue, Fumihiko.

Proceedings of the 5th International Conference on Image and Graphics, ICIG 2009. 2010. p. 398-405 5437887.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Shakunaga, T & Sakaue, F 2010, Homogeneous eigenspace for photometric analyses and recognition. in Proceedings of the 5th International Conference on Image and Graphics, ICIG 2009., 5437887, pp. 398-405, 5th International Conference on Image and Graphics, ICIG 2009, Xi'an, Shanxi, China, 9/20/09. https://doi.org/10.1109/ICIG.2009.139
Shakunaga T, Sakaue F. Homogeneous eigenspace for photometric analyses and recognition. In Proceedings of the 5th International Conference on Image and Graphics, ICIG 2009. 2010. p. 398-405. 5437887 https://doi.org/10.1109/ICIG.2009.139
Shakunaga, Takeshi ; Sakaue, Fumihiko. / Homogeneous eigenspace for photometric analyses and recognition. Proceedings of the 5th International Conference on Image and Graphics, ICIG 2009. 2010. pp. 398-405
@inproceedings{9c1e0fd300e94c988ac73fac5d13885b,
title = "Homogeneous eigenspace for photometric analyses and recognition",
abstract = "Eigenspaces are often utilized in computer vision for various applications. However, careful manipulations are required for sensitive applications in photometric analyses and human interface applications. While several formulations have already been utilized in each application, a unified framework of eigenspace manipulations should be created for further progress of computer vision. This paper provides a concept of homogeneous eigenspaces for this purpose. After brief discussion of eigenspace construction, this paper focuses on how an optimal projection of an input image should be done. Three approaches are presented for the projection and normalization in order to minimize the effects of noise. Both the theoretical and experimental comparisons show that the normalization after projection can suppress the effects of noise more effectively than the other methods. The normalization-after-projection formulation can be reformulated by the homogeneous eigenspace. In the framework of homogeneous eigenspace, a projection onto the normalized eigenspace is accomplished without using an explicit normalization. Furthermore, an optimal partial projection to the normalized eigenspace is also reduced to linear mapping in this framework. Two applications are also discussed for illumination analysis and for photometric classification. These problems are efficiently solved by partial projections onto homogeneous eigenspaces.",
author = "Takeshi Shakunaga and Fumihiko Sakaue",
year = "2010",
doi = "10.1109/ICIG.2009.139",
language = "English",
isbn = "9780769538839",
pages = "398--405",
booktitle = "Proceedings of the 5th International Conference on Image and Graphics, ICIG 2009",

}

TY - GEN

T1 - Homogeneous eigenspace for photometric analyses and recognition

AU - Shakunaga, Takeshi

AU - Sakaue, Fumihiko

PY - 2010

Y1 - 2010

N2 - Eigenspaces are often utilized in computer vision for various applications. However, careful manipulations are required for sensitive applications in photometric analyses and human interface applications. While several formulations have already been utilized in each application, a unified framework of eigenspace manipulations should be created for further progress of computer vision. This paper provides a concept of homogeneous eigenspaces for this purpose. After brief discussion of eigenspace construction, this paper focuses on how an optimal projection of an input image should be done. Three approaches are presented for the projection and normalization in order to minimize the effects of noise. Both the theoretical and experimental comparisons show that the normalization after projection can suppress the effects of noise more effectively than the other methods. The normalization-after-projection formulation can be reformulated by the homogeneous eigenspace. In the framework of homogeneous eigenspace, a projection onto the normalized eigenspace is accomplished without using an explicit normalization. Furthermore, an optimal partial projection to the normalized eigenspace is also reduced to linear mapping in this framework. Two applications are also discussed for illumination analysis and for photometric classification. These problems are efficiently solved by partial projections onto homogeneous eigenspaces.

AB - Eigenspaces are often utilized in computer vision for various applications. However, careful manipulations are required for sensitive applications in photometric analyses and human interface applications. While several formulations have already been utilized in each application, a unified framework of eigenspace manipulations should be created for further progress of computer vision. This paper provides a concept of homogeneous eigenspaces for this purpose. After brief discussion of eigenspace construction, this paper focuses on how an optimal projection of an input image should be done. Three approaches are presented for the projection and normalization in order to minimize the effects of noise. Both the theoretical and experimental comparisons show that the normalization after projection can suppress the effects of noise more effectively than the other methods. The normalization-after-projection formulation can be reformulated by the homogeneous eigenspace. In the framework of homogeneous eigenspace, a projection onto the normalized eigenspace is accomplished without using an explicit normalization. Furthermore, an optimal partial projection to the normalized eigenspace is also reduced to linear mapping in this framework. Two applications are also discussed for illumination analysis and for photometric classification. These problems are efficiently solved by partial projections onto homogeneous eigenspaces.

UR - http://www.scopus.com/inward/record.url?scp=77952281902&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77952281902&partnerID=8YFLogxK

U2 - 10.1109/ICIG.2009.139

DO - 10.1109/ICIG.2009.139

M3 - Conference contribution

AN - SCOPUS:77952281902

SN - 9780769538839

SP - 398

EP - 405

BT - Proceedings of the 5th International Conference on Image and Graphics, ICIG 2009

ER -