Implementation of an Improved Facial Recognition Algorithm in a Web based Learning System

The study focuses on proffering solution to some identified data insecurity problems in software development using Webbased learning system as a test bed, by development of an hybrid crypto-biometric security system, and the use of an enhanced eigen-based facial recognition algorithm. The methodology is by implementation of an optimized principal component analysis eigen facial recognition algorithm for black faces using matlab. A comparative analysis of performance of the optimized principal component analysis (OPCA) and (PCA) principal component analysis is done and it was found out that OPCA performed better than PCA. Also a web based learning system using Hypertext Pre-processor (PHP), Scripting Language for the Web-based pages, Asynchronous JavaScript and XML (AJAX) is developed as a test bed for the crypto biometric system. With this work a prototype for the secured Web-based learning infrastructure is developed and its contextual framework, also an optimized principal component analysis algorithm for black face recognition evolve as contributions to knowledge. hence it will foster indigenization of electronic learning technology which will adequately address the related challenges in the phenomenon of system security in terms of confidentiality and integrity of the system.