Title: Multi-scale neighbourhood based-tree binary pattern: a new feature descriptor for face recognition

Authors: Shekhar Karanwal

Addresses: Computer Science and Engineering Department, Graphic Era Deemed to be University, Dehradun, Uttarakhand, India

Abstract: This research paper proposes the novel face recognition (FR) descriptor for pose and expression variations so-called multi-scale neighbourhood based-tree binary pattern (MNB-TBP). To produce the entire feature size of the MNB-TBP descriptor, two local descriptors are introduced. These two local descriptors are called as NB-TBP and mean (M_n)NB-TBP. The methodologies adopted for NB-TBP and M_nNB-TBP descriptors are totally novel and are based on tree structure. The multi-scale feature extraction and combination of both the proposed descriptors gives the emergence of MNB-TBP descriptor. To produce the compact and robust feature for classification principal component analysis (PCA) is applied further. Finally classification is performed by support vector machines (SVM) and nearest neighbour (NN). The three challenging databases used for the performance evaluation are Olivetti Research Laboratory (ORL), Georgia Technology (GT) and Faces94. The proposed FR approach achieves very effective results which comprehensively outperforms the several state-of-art approaches from the literature.

Keywords: neighbourhood based-tree binary pattern; NB-TBP; mean neighbourhood based-tree binary pattern; M_nNB-TBP; multi-scale neighbourhood based-tree binary pattern; MNB-TBP; principal component analysis; PCA; support vector machines; SVM; nearest neighbour.

DOI: 10.1504/IJBM.2021.114643

International Journal of Biometrics, 2021 Vol.13 No.2/3, pp.322 - 342

Received: 29 Feb 2020
Accepted: 26 May 2020
Published online: 29 Apr 2021 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article