High Precision Deep Learning Model for Road Damage Classification using Transfer Learning

Muhammad Abdul Ghofur, Murdifin Murdifin, Awan Gumilang Hardandrito, Shofwatul 'Uyun

Abstract


Roads are critical infrastructure that frequently experience damage, directly impacting transportation safety and efficiency. Manual road damage inspection is time-consuming and resource-intensive, highlighting the need for automated, image-based approaches. This study compares two Convolutional Neural Network (CNN) architectures—MobileNetV2 with transfer learning and a custom-built CNN—for classifying road surface damage severity. The dataset consists of 1,800 road surface images evenly distributed across three categories: good, minor damage, and severe damage. All images were normalized, augmented, and resized, followed by evaluation using 5-Fold Cross-Validation to ensure robust performance. Experimental results show that MobileNetV2 achieved an accuracy of 98%, outperforming the custom CNN, which achieved 89%. These findings demonstrate the effectiveness of transfer learning in improving classification accuracy with limited data and highlight the potential of MobileNetV2 for efficient, real-time road damage detection systems that can be integrated into intelligent infrastructure monitoring solutions.

Keywords


CNN Custom; K-Fold Cross Validation; MobileNetV2

Full Text:

PDF

References


R. Rustam, M., Susanto, A., & Hidayat, “Deteksi Kerusakan Jalan Menggunakan Deep Learning dan Citra UAV,” J. Teknol. dan Sist. Komput., Vol. 11(1), pp. 33–41, 2023.

K. P. U. dan P. Rakyat, “Statistik Infrastruktur Jalan 2021,” Jakarta: Direktorat Jenderal Bina Marga.

N. Fitri, “Analisis Kerusakan Jalan Akibat Curah Hujan dan Beban Lalu Lintas di Indonesia,” J. Transp. dan Infrastruktur, Vol. 8(2), pp. 77–85, 2022.

M. Tan and Q. V. Le, “EfficientNet: Rethinking model scaling for convolutional neural networks,” 36th Int. Conf. Mach. Learn. ICML 2019, Vol. 2019-June, pp. 10691–10700, 2019.

U. Brawijaya, M. A. Maulana, B. Darma Setiawan, and R. S. Perdana, “Fakultas Ilmu Komputer Klasifikasi Kerusakan Permukaan Jalan Menggunakan Model MobileNetV3-Small,” Vol. X, No. X, No. X, pp. 2548–964, 2024, [Online]. Available: http://j-ptiik.ub.ac.id

A. Abubakar, M. Ajuji, and I. U. Yahya, “Comparison of deep transfer learning techniques in human skin burns discrimination,” Appl. Syst. Innov., Vol. 3, No. 2, pp. 1–15, 2020, DOI: 10.3390/asi3020020.

Y. Yulianto and A. Wibowo, “Neural Network,” Vol. 4, No. 2, pp. 581–594, 2023.

A. W. Mulia, I. Ruslianto, D. M. Midyanti, J. Rekayasa, and S. Komputer, “Pontianak Dan Sekitarnya Dengan Menggunakan,” Vol. 11, No. 01, pp. 11–20, 2023.

R. L. Gaho, I. T. Ali, and E. Prakasa, “Klasifikasi Kualitas Permukaan Jalan Raya Menggunakan Metode CNN Berbasis Arsitektur Xception,” pp. 354–365, 2024.

I. A. Pradana, A. D. Rahajoe, and A. N. Sihananto, “Jalan Berbasis Android Dengan Implementasi Algoritma Hybrid CNN-LSTM,” Vol. 5, No. 2, pp. 1–10, 2024.

N. Khairunisa, . C., and A. Jamaludin, “Analisis Perbandingan Algoritma Cnn Dan Yolo Dalam Mengidentifikasi Kerusakan Jalan,” J. Inform. dan Tek. Elektro Terap., Vol. 12, No. 3, 2024, DOI: 10.23960/jitet.v12i3.4434.

San Gabriel Vanness Kenrick Erwi and Hafiz Irsyad, “Implementasi Deteksi Objek Pada Jalan Rusak Menggunakan Metode YOLOv8,” Bul. Ilm. Inform. Teknol., Vol. 3, No. 1, pp. 1–9, 2024, DOI: 10.58369/biit.v3i1.76.

Y. N. Fuadah, I. D. Ubaidullah, N. Ibrahim, F. F. Taliningsing, N. K. Sy, And M. A. Pramuditho, “Optimasi Convolutional Neural Network dan K-Fold Cross Validation pada Sistem Klasifikasi Glaukoma,” ELKOMIKA J. Tek. Energi Elektr. Tek. Telekomun. Tek. Elektron., Vol. 10, No. 3, p. 728, 2022, DOI: 10.26760/elkomika.v10i3.728.

G. Litjens et al., “A survey on deep learning in medical image analysis,” Med. Image Anal., Vol. 42, No. December 2012, pp. 60–88, 2017, DOI: 10.1016/j.media.2017.07.005.

U. Multi, D. Palembang, J. Rajawali, and N. Palembang, “Klasifikasi Kanker Kulit Pada Citra Dermatoskopi Menggunakan CNN 1,2,” Vol. 5, No. 1, 2024, DOI: 10.35957/algoritme.xxxx.




DOI: https://doi.org/10.32520/stmsi.v14i6.5707

Article Metrics

Abstract view : 6 times
PDF - 1 times

Refbacks

  • There are currently no refbacks.


Creative Commons License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.