Hardware Based Voice Authenticated Security System

This paper explains a model that reflects technical development where computers and humans interact through voice UI. This model contains the code that uses comparisons between the common pitch of a recorded.wav file with that of test signal file. The vector differences between formant peaks within the Power Spectral Density of every file and most accurate one is detected. Feature extraction is obtained using MFCC. These features are trained using Vector Quantization for pattern matching [5]. Once signal matched then MATLAB code will send signal to the microcontroller unit via serial communication, so that microcontroller will control door open and close with time delay and send feedback message to user using Internet communication [1]. These entire operation users can monitor from remote location using BLYNK IOT application. So that this system will provide high security based locking systems for houses, banks etc.