Security System by Using Fingerprints And Voice Recognition

Kong, Sin Yi (2017) Security System by Using Fingerprints And Voice Recognition. Final Year Project (Bachelor), Tunku Abdul Rahman University College.

Text Kong SinYi.pdf Restricted to Registered users only Download (1MB)

Abstract

Fingerprint and voice recognition are the example of visual and auditory biometric which is also an intrinsic property of an individual that is unique and difficult to be shared or duplicated by anyone else. Therefore, the main focus on this project was on the design of a security system using fingerprints and voice recognition to provide a much more accurate and reliable user authentication method and hence enhance security and privacy of a user. User do not require to memorize on password or even carry keys whenever they leave home or workplace. This security system uses MATLAB and Arduino for voice recognition and fingerprints recognition. Voice recognition will be done by signal processing in MATLAB to identify and verify the user identity. MFCC approach is used in order to process the voice signal and extract its features. Then vector quantization will be calculated in MATLAB to measure it with the Euclidean distance. Lastly, the result is compared to the threshold value to determine the authorized user. Arduino is connected to fingerprint scanner to enrol user fingerprint and to determine the authorized user using Arduino IDE. Communication between MATLAB and Arduino via wireless connection uses ESP8266 Wi-Fi module to utilize the internet technology for system flexibility.