Rahmadani, Baihaq (2024) Deteksi Keretakan Beton Menggunakan Pengolahan Citra B-Scan pada Data Ground Penetrating Radar dengan Metode Convolutional Neural Network. Other thesis, Institut Teknologi Sepuluh Nopember.
![]() |
Text
5024201006-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2026. Download (13MB) | Request a copy |
Abstract
Dalam penelitian ini,mengkaji pendekatan deteksi retakan pada data Ground Penetrating
Radar (GPR) menggunakan Convolutional Neural Networks (CNN). Metodologi yang kami
terapkan meliputi empat tahap pokok, yaitu pengumpulan data, perancangan arsitektur CNN,
pelatihan model, dan evaluasi kinerja model. Pada tahap pengumpulan data, kami menghim
pun dua jenis data utama: data input B-Scan dan data output B-scan GPR. Kemudian, pada
tahap perancangan arsitektur CNN, kami mengembangkan model CNN yang dirancang khusus
untuk deteksi retakan dalam data GPR.Selanjutnya, tahap pelatihan model melibatkan proses
pelatihan arsitektur CNN yang telah dibuat dengan menggunakan data yang telah dikumpulkan
sebelumnya. Terakhir, tahap evaluasi kinerja model dilakukan untuk mengukur sejauh mana
model yangtelah dilatih dapat berhasil mendeteksi retakan dalam data GPR.Penelitian ini bertu
juan memberikan panduan yang komprehensif untuk mengevaluasi model CNN dalam konteks
deteksi retakan pada data GPR B-Scan. Kami berharap hasil penelitian ini dapat memberikan
wawasan yang berharga dalam pengembangan teknik deteksi retakan yang lebih efektif dalam
aplikasi GPR.
========================================================================================================================
In this research, examine crack detection approaches on Ground Penetrating Radar (GPR)
data using Convolutional Neural Networks (CNN). The methodology we apply includes four
main stages, namely data collection, CNN architecture design, model training, and model per
formance evaluation. In the data collection stage, we collected two main types of data: B Scan
input data and GPR B-scan output data. Then, at the CNN architecture design stage, we de
veloped a CNN model specifically designed for crack detection in GPR data. Next, the model
training stage involved the training process of the CNN architecture that had been created us
ing previously collected data. Finally, the model performance evaluation stage was carried out
to measure the extent to which the trained model can successfully detect cracks in GPR data.
This research aims to provide a comprehensive guide for evaluating CNN models in the context
of crack detection in B-Scan GPR data. We hope that the results of this research will pro
vide valuable insights in the development of more effective crack detection techniques in GPR
applications
Item Type: | Thesis (Other) |
---|---|
Uncontrolled Keywords: | B-Scan, Crack, CNN, GPR, Retak |
Subjects: | T Technology > T Technology (General) T Technology > TK Electrical engineering. Electronics Nuclear engineering T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5102.9 Signal processing. |
Divisions: | Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Computer Engineering > 90243-(S1) Undergraduate Thesis |
Depositing User: | Baihaq Rahmadani |
Date Deposited: | 06 Aug 2024 05:39 |
Last Modified: | 06 Aug 2024 05:39 |
URI: | http://repository.its.ac.id/id/eprint/111417 |
Actions (login required)
![]() |
View Item |