KLASIFIKASI TINGKAT KEMATANGAN BUAH KELAPA MENGGUNAKAN DEEP LEARNING BERBASIS FITUR AKUSTIK

MUHLIS AHMAD ABDILLAH, . (2023) KLASIFIKASI TINGKAT KEMATANGAN BUAH KELAPA MENGGUNAKAN DEEP LEARNING BERBASIS FITUR AKUSTIK. Sarjana thesis, UNIVERSITAS NEGERI JAKARTA.

[img] Text
1_COVER.pdf

Download (724kB)
[img] Text
2_BAB 1.pdf

Download (225kB)
[img] Text
3_BAB 2.pdf
Restricted to Registered users only

Download (461kB) | Request a copy
[img] Text
4_BAB 3.pdf
Restricted to Registered users only

Download (336kB) | Request a copy
[img] Text
5_BAB 4.pdf
Restricted to Registered users only

Download (612kB) | Request a copy
[img] Text
6_BAB 5.pdf
Restricted to Registered users only

Download (120kB) | Request a copy
[img] Text
7_DAFTAR PUSTAKA.pdf

Download (229kB)
[img] Text
8_LAMPIRAN.pdf
Restricted to Registered users only

Download (612kB) | Request a copy

Abstract

*ABSTRAK* ------- Penelitian ini membahas tentang pemanfaatan bunyi ketukan buah kelapa untuk mengklasifikasikan tingkat kematangannya berbasis fitur akustik. Terdapat kekurangan dalam mengklasifikasikan kematangan kelapa dengan mendengarkan suara ketukannya secara manual sangatlah bergantung dari kemampuan pendengar dalam menentukan kematangan kelapa. Sehingga diperlukan sistem yang dapat melakukan klasifikasi secara otomatis. Fitur akustik dieksplorasi dengan memvarisasikan fitur-fitur yang diekstraksi dari domain frekuensi dan waktu sebagai masukan untuk model deep learning. Fitur yang diekstraksi meliputi MelFrequency Cepstral Coefficients (MFCC) dan Power-Normalized Cepstral Coefficients (PNCC) dari domain frekuensi serta Amplitude Envelope (AE), Zero Crossing Rate (ZCR), dan RMS Energy (RMS Energy) dari domain waktu. Dalam penelitian ini, digunakan adalah Long Short-Term Memory (LSTM) dan Deep Neural Network (DNN). Hasil penelitian menunjukkan bahwa model LSTM dan DNN memperoleh akurasi 92,86% dan 89,29% dengan fitur domain frekuensi. This research discusses the utilization of coconut beats to classify the level of maturity based on acoustic features. There are shortcomings in classifying coconut maturity by listening to the sound of the knock manually is very dependent on the ability of the listener in determining the maturity of the coconut. So a system is needed that can do the classification automatically. Acoustic features are explored by varying the features extracted from the frequency and time domains as input to the deep learning model. The extracted features include Mel-Frequency Cepstral Coefficients (MFCC) and Power-Normalized Cepstral Coefficients (PNCC) from the frequency domain and Amplitude Envelope (AE), Zero Crossing Rate (ZCR), and RMS Energy from the time domain. In this study, Long Short-Term Memory (LSTM) and Deep Neural Network (DNN) were used. The results showed that the LSTM and DNN models obtained 92.86% and 89.29% accuracy with frequency domain features.

Item Type: Thesis (Sarjana)
Additional Information: 1) Dr. rer. nat. Bambang Heru Iswanto, M.Si. ; 2) Haris Suhendar, M.Si.
Subjects: Sains > Fisika
Divisions: FMIPA > S1 Fisika
Depositing User: Users 18984 not found.
Date Deposited: 07 Sep 2023 01:56
Last Modified: 07 Sep 2023 01:56
URI: http://repository.unj.ac.id/id/eprint/41253

Actions (login required)

View Item View Item