Registration of 3D face images of JUDeiTy3DK database using POFaceReg8 algorithm

The application of depth cameras for 3D visual data based measurements (VBM) is an active research area. 3D face recognition is one such niche area that is drawing the attention of researchers. This paper introduces a new 3D facial database named JUDeiTy3DK developed by a research group at Jadavpur...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Sadhana (Bangalore) 2023-08, Vol.48 (3), Article 158
Hauptverfasser: Bagchi, Parama, Bhattacharjee, Debotosh, Nasipuri, Mita, Krejcar, Ondrej
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:The application of depth cameras for 3D visual data based measurements (VBM) is an active research area. 3D face recognition is one such niche area that is drawing the attention of researchers. This paper introduces a new 3D facial database named JUDeiTy3DK developed by a research group at Jadavpur University, Kolkata, India, using Kinect as a vision-based depth measuring instrument to capture 3D facial depth data. This database aims to provide the research community with a new 3D face database containing various challenging situations in the context of face recognition. This database consists of 3D face models of 200 male and female individuals with various poses, facial expressions, real-world occlusions, and combinations of various poses, expressions, and occlusions we encounter daily. In addition, ground truth face images are also provided, where 14 visible facial landmark points on each face are manually annotated. Also, we have formulated a new 3D face registration algorithm named POFaceReg8, which combines some selected features of landmark and holistic-based algorithms, followed by a new face recognition algorithm Diag_DCT . We have presented the registration and recognition performance results by comparing our proposed methods with other standard state-of-the-art techniques using pre-existing metrics.
ISSN:0973-7677
0973-7677
DOI:10.1007/s12046-023-02225-w