Face recognition using facial shape indices with two different three-dimensional sensors

Hwanjong Song, Hyoungchul Shin, Sang Youn Lee, Kwanghoon Sohn

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

This paper describes a three-dimensional (3-D) face recognition system based on two different 3-D sensors. These sensors were used to overcome pose variation problems that cannot be effectively solved when working with 2-D images. We acquired input data based on a structured light system and compared them with 3-D faces acquired by a 3-D laser scanner. Due to differing data structures, we generated a selection of probe images and stored images (not only for head pose estimation but also for face recognition). Given an unknown range image, we extracted invariant facial features based on facial geometry and utilized the previously developed error-compensated singular-value decomposition method to estimate a head pose. Distinctive facial shape indices were defined and extracted based on facial curvature characteristics. The extracted indices have a different number and different distribution on each face image. When multiple matching possibilities are involved, dynamic programming (DP) is useful matching algorithm. DP merges data points in order to achieve better point-to-point matching by finding a matching path at minimum cost. Experimental results show that the proposed method obtained a 96.8% face recognition rate when working with 300 individuals under different pose variations.

Original languageEnglish
Article number067201
JournalOptical Engineering
Volume46
Issue number6
DOIs
Publication statusPublished - 2007 Jun 1

Fingerprint

Face recognition
Dynamic programming
sensors
Sensors
dynamic programming
Singular value decomposition
Data structures
data structures
Geometry
Lasers
scanners
curvature
Costs
costs
decomposition
probes
estimates
geometry
lasers

All Science Journal Classification (ASJC) codes

  • Atomic and Molecular Physics, and Optics
  • Engineering(all)

Cite this

@article{215b2b507b6847deb05420f222d7678b,
title = "Face recognition using facial shape indices with two different three-dimensional sensors",
abstract = "This paper describes a three-dimensional (3-D) face recognition system based on two different 3-D sensors. These sensors were used to overcome pose variation problems that cannot be effectively solved when working with 2-D images. We acquired input data based on a structured light system and compared them with 3-D faces acquired by a 3-D laser scanner. Due to differing data structures, we generated a selection of probe images and stored images (not only for head pose estimation but also for face recognition). Given an unknown range image, we extracted invariant facial features based on facial geometry and utilized the previously developed error-compensated singular-value decomposition method to estimate a head pose. Distinctive facial shape indices were defined and extracted based on facial curvature characteristics. The extracted indices have a different number and different distribution on each face image. When multiple matching possibilities are involved, dynamic programming (DP) is useful matching algorithm. DP merges data points in order to achieve better point-to-point matching by finding a matching path at minimum cost. Experimental results show that the proposed method obtained a 96.8{\%} face recognition rate when working with 300 individuals under different pose variations.",
author = "Hwanjong Song and Hyoungchul Shin and Lee, {Sang Youn} and Kwanghoon Sohn",
year = "2007",
month = "6",
day = "1",
doi = "10.1117/1.2744386",
language = "English",
volume = "46",
journal = "Optical Engineering",
issn = "0091-3286",
publisher = "SPIE",
number = "6",

}

Face recognition using facial shape indices with two different three-dimensional sensors. / Song, Hwanjong; Shin, Hyoungchul; Lee, Sang Youn; Sohn, Kwanghoon.

In: Optical Engineering, Vol. 46, No. 6, 067201, 01.06.2007.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Face recognition using facial shape indices with two different three-dimensional sensors

AU - Song, Hwanjong

AU - Shin, Hyoungchul

AU - Lee, Sang Youn

AU - Sohn, Kwanghoon

PY - 2007/6/1

Y1 - 2007/6/1

N2 - This paper describes a three-dimensional (3-D) face recognition system based on two different 3-D sensors. These sensors were used to overcome pose variation problems that cannot be effectively solved when working with 2-D images. We acquired input data based on a structured light system and compared them with 3-D faces acquired by a 3-D laser scanner. Due to differing data structures, we generated a selection of probe images and stored images (not only for head pose estimation but also for face recognition). Given an unknown range image, we extracted invariant facial features based on facial geometry and utilized the previously developed error-compensated singular-value decomposition method to estimate a head pose. Distinctive facial shape indices were defined and extracted based on facial curvature characteristics. The extracted indices have a different number and different distribution on each face image. When multiple matching possibilities are involved, dynamic programming (DP) is useful matching algorithm. DP merges data points in order to achieve better point-to-point matching by finding a matching path at minimum cost. Experimental results show that the proposed method obtained a 96.8% face recognition rate when working with 300 individuals under different pose variations.

AB - This paper describes a three-dimensional (3-D) face recognition system based on two different 3-D sensors. These sensors were used to overcome pose variation problems that cannot be effectively solved when working with 2-D images. We acquired input data based on a structured light system and compared them with 3-D faces acquired by a 3-D laser scanner. Due to differing data structures, we generated a selection of probe images and stored images (not only for head pose estimation but also for face recognition). Given an unknown range image, we extracted invariant facial features based on facial geometry and utilized the previously developed error-compensated singular-value decomposition method to estimate a head pose. Distinctive facial shape indices were defined and extracted based on facial curvature characteristics. The extracted indices have a different number and different distribution on each face image. When multiple matching possibilities are involved, dynamic programming (DP) is useful matching algorithm. DP merges data points in order to achieve better point-to-point matching by finding a matching path at minimum cost. Experimental results show that the proposed method obtained a 96.8% face recognition rate when working with 300 individuals under different pose variations.

UR - http://www.scopus.com/inward/record.url?scp=34548033285&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=34548033285&partnerID=8YFLogxK

U2 - 10.1117/1.2744386

DO - 10.1117/1.2744386

M3 - Article

VL - 46

JO - Optical Engineering

JF - Optical Engineering

SN - 0091-3286

IS - 6

M1 - 067201

ER -