Occlusions in Face Recognition: a 3D Approach

Colombo, Alessandro; Cusano, Claudio; Schettini, Raimondo

doi:10.5772/6644

Search citation statements

Order By: Relevance

Paper Sections

Select...

Introduction1

Citation Types

Supporting

Mentioning

Contrasting

Year Published

2015

Publication Types

Select...

Article1

Relationship

Self Cite0

Independent1

Authors

Journals

Cited by 1 publication

(1 citation statement)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Facial surfaces are approached by indexed collections of radial curves. Occlusions detection and restoration of partially occluded faces was also approached by Colombo et al [9,10,11]. The detection method considers occlusions as local deformations of the face and the restoration task is based on Principal Component Analysis (PCA) reconstruction.…”

Section: Introductionmentioning

confidence: 99%

Aided 3D Facial Restoration for Plastic Surgery Planning using Elastic Radial Curves

Hermessi¹,

Mourali²

2015

IJCA

View full text Add to dashboard Cite

Improving appearance and guarantying healthcare are the important objectives of facial plastic surgery. Face can be the seat of tumors that can be restored using 3D images in a way to surpass 2D limits such as change in illumination conditions and pose variation. This work proposes a Riemannian geometry approach to illustrate the degradation of facial tumefaction which is based on geodesic paths between healthy and swollen half of the face. The progressivity of tumor volume assists plastic surgeons in approaching the harmonious shapes for facial restoration. We introduce a preprocessing step to prepare 3D scans chosen from FRAV 3D database then sculpted to simulate unilateral tumors. We present facial surfaces by indexed collections of radial curves emanating from the nose tips. Symmetric indexed radial curves are compared using elastic shape analysis. General TermsElastic 3D face analysis for plastic surgery planning.

show abstract

Section: Introductionmentioning

confidence: 99%