ENHANCING USER AUTHENTICATION OF ONLINE CREDIT CARD PAYMENT USING FACE IMAGE COMPARISON WITH MPEG7-EDGE HISTOGRAM DESCRIPTOR

Abstract

In the past decade, progressions of internet and information technology support most of public and individual activities. Online payment transaction is one of those activities which also play an important role in people life. It is an alternative option for persons to choose for convenience. However, user information can leak easily without strictly protection. Biometric verification was considered as a key to overcome this security problem. In this thesis, face was selected as an input for biometric verification in online payment process. MPEG7-Edge Histogram Descriptor was proposed to be used in the feature extraction of face image. For the experimental results, this descriptor outperformed the other descriptors. Therefore, it can be applied to the new verification process of online payment system. Then, uncomplicated process with high performance was designed and created. The experimental results based on statistical analysis showed that the process with face matching verification can increase system security as well as improve the usability, capability, and user satisfaction.