Identifikasi Biometrik Menggunakan Suara Ketukan Gigi Dengan Fitur Mel Frequency Cepstral Coefficients (MFCC)

Radityo, Muhammad Rafif (2021) Identifikasi Biometrik Menggunakan Suara Ketukan Gigi Dengan Fitur Mel Frequency Cepstral Coefficients (MFCC). Undergraduate thesis, Institut Teknologi Sepuluh Nopember.

Text 02311740000107-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2023. Download (5MB) | Request a copy

Abstract

Biometrik merupakan salah satu metode yang dapat digunakan sebagai keamanan perangkat. Salah satu jenis biometrik yang banyak digunakan yaitu biometrik dengan input suara. Masing-masing manusia memiliki identitas suara yang berbeda-beda dan suara yang dihasilkan oleh manusia dapat bersumber dari ketukan gigi. Sinyal dari ketukan gigi akan memiliki koefisien tertentu setelah diolah dengan fitur MFCC (Mel Frequency Cepstral Coefficients). Pada penelitian kali ini digunakan alat yang kami sebut in-ear microphone sebagai penerima sinyal suara dan sinyal hasil pengolahan MFCC tersebut kemudian dijadikan sebagai data training dan sisanya sebagai data testing yang diolah menggunakan beberapa jenis classifier untuk dianalisa akurasinya. Hasil menunjukan bahwa akurasi data training dan data testing sebesar 99,3% dan 98% menggunakan machine learning jenis SVM (support vector machine). Sedangkan akurasi data training dan data testing menggunakan machine learning jenis neural network adalah 100% dan 98%.
==========================================================
=======================================
Biometric is one method that can be used as a security device. One type of biometric that is widely used is biometrics with voice input. Each human has a different voice identity and the sound produced by humans can be sourced from the dental occlusion or tooth click. The signal from the dental occlusion will have some certain coefficients after being processed with the MFCC (Mel Frequency Cepstrals Coefficients) feature. In this study, we used a tool called an in-ear microphone as a voice signal receiver and the MFCC processing signal was then used as data training and the rest as data testing, which was processed using machine learning to analyze its accuracy. The results show that the accuracy of training data and testing data is 99.3% and 98% using machine learning type SVM (support vector machine). Meanwhile, the accuracy of training data and testing data using machine learning type of neural network is 100% and 98%.

Item Type:	Thesis (Undergraduate)
Uncontrolled Keywords:	biometrik, ketukan gigi, MFCC, machine learning, biometric, tooth click, MFCC, machine learning.
Subjects:	Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines.
Divisions:	Faculty of Industrial Technology and Systems Engineering (INDSYS) > Physics Engineering > 30201-(S1) Undergraduate Thesis
Depositing User:	Muhammad Rafif Radityo
Date Deposited:	24 Aug 2021 06:04
Last Modified:	24 Aug 2021 06:04
URI:	http://repository.its.ac.id/id/eprint/90009

Actions (login required)

View Item