Development of a Lips Recognition System using MATLAB

Tan, Zhi Shyuan (2019) Development of a Lips Recognition System using MATLAB. Final Year Project (Bachelor), Tunku Abdul Rahman University College.

Text Tan Zhi Shyuan.pdf Restricted to Registered users only Download (2MB)

Abstract

Comparing to the traditional way of authentication, biometric authentication is way more consistent and efficient as it is more difficult to be stolen or falsified. Lip is emerging as a new trend in biometric. As lips biometric is relative new in this field, algorithm and results to support the usefulness of this biometric method is lacking. The main idea of this research is to develop a lips recognition system using MATLAB as the platform. The proposed algorithm is able to detect the lips from an image and extract lip envelope parameters as the feature of interest. Recognition of users’ identity was performed based on the lip recognition using logic equations and calculations. 17 out of 20 tests are correctly identified and given a True Positive rate of 85% if the system is presented with lip images from same subjects. The False Negative rate is found as 15%. While randomized test yield TN rate which is correctly rejected of 73.3%, and FN which is incorrectly rejected rate is 26.7%. The overall accuracy of the lips recognition system is 80%. The results endorse the feasibility of lips as biometric.