Augmented Small-Scale Database to Improve the Performance of Eigenface Recognition Technique

Allam's picture
Journal Title, Volume, Page: 
IJCVIP, Vol.2 Issue 2, pp 59-70, 2012
Year of Publication: 
Allam Mousa
An-Najah National University, Palestine
Current Affiliation: 
Department Of Electrical Engineering, Faculty Of Engineering, An-Najah National University, Nablus, Palestine
Rana Salameh
ALLESCO, Palestine
Rawan Abu Shmais
JAWWAL, Palestine
Preferred Abstract (Original): 
Eigenface recognition technique reserves limitations in achieving good performance. This includes the large-scale database required, its sensitivity to improper illumination, as well as to different expressions of a human face, and image background. This paper presents an efficient and accessible solution for some of these limitations by improving the database’s design. Face recognition accuracy has been enhanced via the inclusion of a modified version of the images in the database. Illumination and various face positions have been integrated into the already available small-scale database. Recognition is sensitive to the illuminated side of a face under consideration. Applying the proposed approach and choosing proper pre-processing values, has improved the system’s performance.