Mixed texture descriptors for facial expression recognition
International Journal of Informatics and Applied Mathematics · 02.08.2026 · 9 (1) · 50-65
DOI: 10.53508/ijiam.1602587 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Informatics and Applied Mathematics, online erschienen am 02.08.2026. Verfasst von Chebah Ouafa und 1 weiteren Autor:innen. Die Arbeit nennt 25 Literaturangaben.
Abstract (Originalsprache)
This article proposes a novel approach to recognize facial expression using a combination of three local texture descriptors such as Local Binary Pattern (LBP), Gradient Direction Pattern (GDP) and Local Directional Pattern (LDP). Firstly, we need to perform the pre-process of face image, which includes face detection and face image cropping. Next, we calculate texture descriptors based on LBP, GDP and LDP operators. Histogram sequence concatenation is then applied to these descriptors after dividing the facial image into a number of non-overlapping blocks. Finally, the concatenated histograms represent an input to a Support Vector Machine (SVM), which will be used to classify facial images in one of the six universal expressions. Experimental results using CK+ and JAFFE database show that the proposed approach achieves superior recognition performance compared to the existing studies with classification accuracy of 97.62% and 93.84% respectively.
Zitieren
APA
Ouafa, C., & Anissa, C. (2026). Mixed texture descriptors for facial expression recognition. International Journal of Informatics and Applied Mathematics, 9(1), 50-65. https://doi.org/10.53508/ijiam.1602587
Vancouver
Ouafa C, Anissa C. Mixed texture descriptors for facial expression recognition. International Journal of Informatics and Applied Mathematics. 2026;9(1):50-65. doi:10.53508/ijiam.1602587
BibTeX
@article{Ouafa2026,
author = {Ouafa, Chebah and Anissa, Chebah},
title = {Mixed texture descriptors for facial expression recognition},
journal = {International Journal of Informatics and Applied Mathematics},
year = {2026},
volume = {9},
number = {1},
pages = {50--65},
doi = {10.53508/ijiam.1602587}
}
EUVIMED