Reconstructing a whole face image from a partially damaged or occluded image by multiple matching

Bon W. Hwang, Seong Whan Lee

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

3 Citations (Scopus)

Abstract

The problem we address in this paper is, given a facial image that is partially occluded or damaged by noise, to reconstruct a whole face, A key process for the reconstruction is to obtain the correspondences between the input image and the reference face. We present a method that matches an input image with multiple example images that are generated from a morphable face model. From the matched feature points, shape and texture of the full face are inferred by the non-iterative data completion algorithm. Compared with single matching with the particular "reference face", this multiple matching method increases the robustness of the matching. The experimental results of applying the algorithm to face images that are contaminated by Gaussian noise and those which are partially occluded show that the reconstructed faces are plausible and similar to the original ones.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Pages692-701
Number of pages10
Volume4642 LNCS
Publication statusPublished - 2007 Dec 1
Event2007 International Conference on Advances in Biometrics, ICB 2007 - Seoul, Korea, Republic of
Duration: 2007 Aug 272007 Aug 29

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume4642 LNCS
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other2007 International Conference on Advances in Biometrics, ICB 2007
CountryKorea, Republic of
CitySeoul
Period07/8/2707/8/29

Fingerprint

Face
Noise
Textures
Feature Point
Gaussian Noise
Completion
Texture
Correspondence
Robustness
Experimental Results
Model

Keywords

  • Data completion
  • Face reconstruction
  • Morphable face model
  • SIFT feature

ASJC Scopus subject areas

  • Biochemistry, Genetics and Molecular Biology(all)
  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Hwang, B. W., & Lee, S. W. (2007). Reconstructing a whole face image from a partially damaged or occluded image by multiple matching. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 4642 LNCS, pp. 692-701). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4642 LNCS).

Reconstructing a whole face image from a partially damaged or occluded image by multiple matching. / Hwang, Bon W.; Lee, Seong Whan.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 4642 LNCS 2007. p. 692-701 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4642 LNCS).

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

Hwang, BW & Lee, SW 2007, Reconstructing a whole face image from a partially damaged or occluded image by multiple matching. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 4642 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 4642 LNCS, pp. 692-701, 2007 International Conference on Advances in Biometrics, ICB 2007, Seoul, Korea, Republic of, 07/8/27.
Hwang BW, Lee SW. Reconstructing a whole face image from a partially damaged or occluded image by multiple matching. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 4642 LNCS. 2007. p. 692-701. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
Hwang, Bon W. ; Lee, Seong Whan. / Reconstructing a whole face image from a partially damaged or occluded image by multiple matching. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 4642 LNCS 2007. pp. 692-701 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{316a9699ac794872953180bd9afa85d0,
title = "Reconstructing a whole face image from a partially damaged or occluded image by multiple matching",
abstract = "The problem we address in this paper is, given a facial image that is partially occluded or damaged by noise, to reconstruct a whole face, A key process for the reconstruction is to obtain the correspondences between the input image and the reference face. We present a method that matches an input image with multiple example images that are generated from a morphable face model. From the matched feature points, shape and texture of the full face are inferred by the non-iterative data completion algorithm. Compared with single matching with the particular {"}reference face{"}, this multiple matching method increases the robustness of the matching. The experimental results of applying the algorithm to face images that are contaminated by Gaussian noise and those which are partially occluded show that the reconstructed faces are plausible and similar to the original ones.",
keywords = "Data completion, Face reconstruction, Morphable face model, SIFT feature",
author = "Hwang, {Bon W.} and Lee, {Seong Whan}",
year = "2007",
month = "12",
day = "1",
language = "English",
isbn = "9783540745488",
volume = "4642 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "692--701",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - GEN

T1 - Reconstructing a whole face image from a partially damaged or occluded image by multiple matching

AU - Hwang, Bon W.

AU - Lee, Seong Whan

PY - 2007/12/1

Y1 - 2007/12/1

N2 - The problem we address in this paper is, given a facial image that is partially occluded or damaged by noise, to reconstruct a whole face, A key process for the reconstruction is to obtain the correspondences between the input image and the reference face. We present a method that matches an input image with multiple example images that are generated from a morphable face model. From the matched feature points, shape and texture of the full face are inferred by the non-iterative data completion algorithm. Compared with single matching with the particular "reference face", this multiple matching method increases the robustness of the matching. The experimental results of applying the algorithm to face images that are contaminated by Gaussian noise and those which are partially occluded show that the reconstructed faces are plausible and similar to the original ones.

AB - The problem we address in this paper is, given a facial image that is partially occluded or damaged by noise, to reconstruct a whole face, A key process for the reconstruction is to obtain the correspondences between the input image and the reference face. We present a method that matches an input image with multiple example images that are generated from a morphable face model. From the matched feature points, shape and texture of the full face are inferred by the non-iterative data completion algorithm. Compared with single matching with the particular "reference face", this multiple matching method increases the robustness of the matching. The experimental results of applying the algorithm to face images that are contaminated by Gaussian noise and those which are partially occluded show that the reconstructed faces are plausible and similar to the original ones.

KW - Data completion

KW - Face reconstruction

KW - Morphable face model

KW - SIFT feature

UR - http://www.scopus.com/inward/record.url?scp=37849043951&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=37849043951&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:37849043951

SN - 9783540745488

VL - 4642 LNCS

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 692

EP - 701

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

ER -