Face Recognition using Singular Value Decomposition (SVD)

Raziq Irfan, Bakri (2019) Face Recognition using Singular Value Decomposition (SVD). [Final Year Project Report] (Unpublished)

PDF (Please get the password by email to repository@unimas.my, or call ext: 3914/ 3942/ 3933) Raziq Irfan ft.pdf Restricted to Registered users only Download (1MB)

Abstract

Security is one of the important aspects when it comes to personal privacy. Access to this personal privacy information needs a high-security system. The biometrics scanner is one of the best security systems these days. Other than the thumbprint and iris scanner, the face recognition system is also one of the most focus systems to be improved. This system has been implemented everywhere across the globe, from smartphones to accessing a secured building. Although it is widely used everywhere, the system has a few flaws that needed to be upgraded. The number of correctly identify a person is low due to some factors such as the variation of light, the various facial expressions and the difference in pose. The time taken for a complete face recognition also quite high. This is because of the method implemented in the face recognition system and the size of the database. This report is to show the implementation of the Singular Value Decomposition method into the face recognition system. Singular Value Decomposition unlike Principal Component Analysis, does not cause high computation complexity when dealing with a large database. This will reduce the time taken for face recognition to completely done. Singular Value Decomposition is not a complex approach for face recognition system. It can identify a person with a single image of a face. This will increase the rate of success of correctly identified a person. The proposed method, Singular Value Decomposition will improve the face recognition system in term of the rate of success and time taken for a complete face recognition.

Actions (For repository members only: login required)

View Item