14 March 2013 Face recognition with directional ternary pattern (DTP)
Author Affiliations +
Proceedings Volume 8768, International Conference on Graphic and Image Processing (ICGIP 2012); 87687E (2013) https://doi.org/10.1117/12.2009559
Event: 2012 International Conference on Graphic and Image Processing, 2012, Singapore, Singapore
Abstract
This paper presents an effective and robust facial feature descriptor based on the directional ternary pattern (DTP) for face recognition. The DTP operator encodes the texture information of a local region by labeling the edge response values in all eight directions around a pixel with three different levels. The coding scheme exploits a threshold in order to differentiate between smooth and high-textured face regions while forming the ternary code. The location and occurrence information of the DTP micro-patterns within the facial image is then used as the feature descriptor. The effectiveness of the proposed method is evaluated with the FERET face image database using template matching (TM). Extensive experiments show the superiority of the DTP feature descriptor against some well-known local pattern-based feature representation methods.
© (2013) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Md. Hasanul Kabir, Md. Hasanul Kabir, Faisal Ahmed, Faisal Ahmed, } "Face recognition with directional ternary pattern (DTP)", Proc. SPIE 8768, International Conference on Graphic and Image Processing (ICGIP 2012), 87687E (14 March 2013); doi: 10.1117/12.2009559; https://doi.org/10.1117/12.2009559
PROCEEDINGS
5 PAGES


SHARE
Back to Top