文档介绍:Journal of King Saud University – Computer and Information Sciences (2018) 30, 141–151
King Saud University
Journal of King Saud University –
Computer and Information Sciences
Face recognition using Angular Radial Transform
Bensenane Hamdan *, Keche Mokhtar
Laboratoire Signals and Images, Dept. of Electronique, Universite´ des Sciences et de la Technologie d’Oran Mohamed
Boudiaf, USTO-MB, BP 1505, 3100 Oran, Algeria
Received 8 May 2016; revised 24 October 2016; accepted 27 October 2016
Available online 5 December 2016
KEYWORDS Abstract Moment-based Angular Radial Transform, Legendre moment invariants and Zernike
Angular Radial Transform moments are a family of orthogonal functions which allow the generation of non-redundant
(ART); descriptors by the projection of an image onto an orthogonal basis. These descriptors can be used
Legendre moment invariants for classification, such as in face recognition. Zernike moments and Legendre moments have
(LMI); already been used for this purpose.
Euclidean distance (ED); This paper proposes to use moment-based Angular Radial Transform for extracting the face
Pseudo-Zernike moments characteristics that feed a Support Vector Machine or a Nearest Neighbor Classifier for face recog-
(PZM); nition. Facial images from the ORL database, Essex Faces94 database, Essex Faces96 database,
Nearest Neighbor Classifier and Yale database were used for testing the proposed approach. The experimental results obtained
(NNC);
show that the proposed method is more efficient, i