COMPARATIVE EVALUATION OF CNN ARCHITECTURES FOR ROAD DAMAGE CLASSIFICATION USING DIGITAL IMAGES IN SLEMAN REGENCY
Abstract
Reliable road condition monitoring is fundamental to maintenance decision-making and transportation safety, particularly in regional contexts where data resources are often scarce. This study presents a comparative evaluation of convolutional neural network (CNN) architectures for classifying road damage types using digital images collected in Sleman Regency. Three widely used CNN architectures, VGGNet-16, InceptionV3, and Xception, were evaluated under a unified experimental framework employing transfer learning, consistent preprocessing, explicit hyperparameter tuning, and four-fold cross-validation. The dataset comprises three road damage categories, alligator crack, corrugation, and pothole, captured under heterogeneous pavement and lighting conditions. Image preprocessing includes resizing, augmentation, and contrast enhancement using Contrast Limited Adaptive Histogram Equalization (CLAHE). To assess the contribution of preprocessing choices, an ablation study was conducted by comparing model performance with and without CLAHE. Experimental results indicate that all evaluated architectures achieve high classification performance. Among them, Xception consistently demonstrates the best overall performance across validation and test sets, achieving the highest accuracy, precision, recall, and F1-score. The ablation analysis further shows that including CLAHE consistently improves performance, particularly in recall and F1-score, indicating enhanced robustness under uneven illumination conditions. Although the contribution of this study is incremental rather than algorithmically novel, the findings provide empirical insights into the comparative behavior of CNN architectures under region-specific road conditions. The results highlight the importance of systematic preprocessing and controlled evaluation in CNN-based road-damage classification and provide practical guidance for selecting suitable architectures for regional road maintenance decision-support systems.
Downloads
References
K. Ceni and E. Ina, “IDENTIFIKASI JENIS DAN PENANGANAN KERUSAKAN JALAN (STUDI KASUS JL. G. OBOS XII, JL. SAMUDIN AMAN, JL. JATI KOTA PALANGKA RAYA),” Narotama Jurnal Teknik Sipil, vol. 5, no. 2, pp. 28–36, Nov. 2021, doi: https://doi.org/10.31090/njts.v5i2.1567
M. S. Karim, A. T. Handayani, and H. P. Astutik, “KINERJA RUAS JALAN SAAT KONDISI NEW NORMAL (STUDI KASUS JALAN LAKSDA ADISUTJIPTO, YOGYAKARTA KM 6,3-6,8),” EQUILIB, vol. 2, no. 1, pp. 13–20, 2021.
S. Elmaningtyas and S. Andayani, “FUZZY TOPSIS APPLICATION TO DETERMINE PRIORITIES OF ROAD MAINTENANCE IN SLEMAN REGENCY,” Jurnal Kajian dan Terapan Matematika, vol. 8, no. 2, pp. 138–148, Jul. 2022, [Online]. Available: http://journal.student.uny.ac.id/ojs/index.php/jktm
A. Kusnadi and Ranny, “KERUSAKAN JALAN FLEXIBLE PAVEMENT DENGAN MENGGUNAKAN ALGORITMA PCA,” ULTIMATICS, vol. 8, no. 2, pp. 125–130, Dec. 2016, doi: https://doi.org/10.31937/ti.v8i2.521
A. M. Wira, I. Ruslianto, and D. M. Midyanti, “KLASIFIKASI KERUSAKAN JALAN PADA CITRA JALAN RAYA PONTIANAK DAN SEKITARNYA MENGGUNAKAN CONVOLUTIONAL NEURAL NETWORK,” Coding: Jurnal Komputer dan Aplikasi, vol. 11, no. 1, 2023, doi: https://doi.org/10.26418/coding.v11i1.57905
R. Octavia, “IMPLEMENTASI CONVOLUTIONAL NEURAL NETWORK UNTUK KLASIFIKASI KERUSAKAN JALAN BERBASIS CITRA DIGITAL (STUDI KASUS: KERUSAKAN JALAN ASPAL DI BANYUMAS),” 2024, Yogyakarta, Indonesia.
D. Arya et al., “DEEP LEARNING-BASED ROAD DAMAGE DETECTION AND CLASSIFICATION FOR MULTIPLE COUNTRIES,” Autom. Constr., vol. 132, p. 103935, Dec. 2021, doi: https://doi.org/10.1016/j.autcon.2021.103935
L. Manoni, S. Orcioni, and M. Conti, “RECENT ADVANCEMENTS IN DEEP LEARNING TECHNIQUES FOR ROAD CONDITION MONITORING: A COMPREHENSIVE REVIEW,” IEEE Access, vol. 12, pp. 154271–154293, 2024, doi: https://doi.org/10.1109/ACCESS.2024.3481649
B. Omarov and B. Kulambayev, “DEEP LEARNING-BASED IMAGE PROCESSING FOR REAL-TIME DETECTION OF ROAD SURFACE DAMAGE,” Procedia Comput. Sci., vol. 251, pp. 609–614, 2024, doi: https://doi.org/10.1016/j.procs.2024.11.157
L. Alzubaidi et al., “REVIEW OF DEEP LEARNING: CONCEPTS, CNN ARCHITECTURES, CHALLENGES, APPLICATIONS, FUTURE DIRECTIONS,” J. Big Data, vol. 8, no. 1, 2021. https://doi.org/10.1186/s40537-021-00444-8
Y. Jiang, “ROAD DAMAGE DETECTION AND CLASSIFICATION USING DEEP NEURAL NETWORKS,” Discover Applied Sciences, vol. 6, no. 8, p. 421, 2024, doi: https://doi.org/10.1007/s42452-024-06129-0
S. Umar, Maimunah, and H. A. Meidar, “KLASIFIKASI JALAN RUSAK MENGGUNAKAN TRANSFER LEARNING ARSITEKTUR VGG16,” Journal of Information Systems and Informatics Engineering, vol. 8, no. 1, pp. 75–85, Jun. 2024.
T. J. W. Adi, P. Suprobo, and Y. E. P. R. Waliulu, “IRODD (INTELLIGENT-ROAD DAMAGE DETECTION) FOR REAL-TIME INFRASTRUCTURE PRESERVATION IN DETECTION, CLASSIFICATION, CALCULATION, AND VISUALIZATION,” Journal of Infrastructure Policy and Development, vol. 8, no. 11, Oct. 2024, doi: https://doi.org/10.24294/jipd.v8i11.6162
S. Swain and A. K. Tripathy, “AUTOMATIC DETECTION OF POTHOLES USING VGG-16 PRE-TRAINED NETWORK AND CONVOLUTIONAL NEURAL NETWORK,” Heliyon, vol. 10, no. 10, p. e30957, 2024, doi: https://doi.org/10.1016/j.heliyon.2024.e30957
X. Ren, S. Huang, Y. Hu, K. Ye, Z. Chen, and Z. Wang, “A LIGHTWEIGHT CONVOLUTIONAL NEURAL NETWORK FOR DETECTING ROAD CRACKS,” Signal Image Video Process., vol. 18, no. 10, pp. 6729–6743, 2024, doi: https://doi.org/10.1007/s11760-024-03347-2
S. Zhang, Z. Liu, K. Wang, W. Huang, and P. Li, “OBC-YOLOV8: AN IMPROVED ROAD DAMAGE DETECTION MODEL BASED ON YOLOV8,” PeerJ Comput. Sci., vol. 11, p. e2593, 2025, doi: https://doi.org/10.7717/peerj-cs.2593
Y. Li, C. Yin, Y. Lei, J. Zhang, and Y. Yan, “RDD-YOLO: ROAD DAMAGE DETECTION ALGORITHM BASED ON IMPROVED YOU ONLY LOOK ONCE VERSION 8,” Applied Sciences, vol. 14, no. 8, p. 3360, 2024, doi: https://doi.org/10.3390/app14083360
Z. Zhang et al., “ARDS-YOLO: INTELLIGENT DETECTION OF ASPHALT ROAD DAMAGES AND EVALUATION OF PAVEMENT CONDITION IN COMPLEX SCENARIOS,” Measurement, vol. 242, p. 115946, 2025, doi: https://doi.org/10.1016/j.measurement.2024.115946
S. Youwai, A. Chaiyaphat, and P. Chaipetch, “YOLO9TR: A LIGHTWEIGHT MODEL FOR PAVEMENT DAMAGE DETECTION UTILIZING A GENERALIZED EFFICIENT LAYER AGGREGATION NETWORK AND ATTENTION MECHANISM,” J. Real. Time. Image Process., vol. 21, no. 5, p. 163, 2024. https://doi.org/10.1007/s11554-024-01545-2
M. W. Khan, M. S. Obaidat, K. Mahmood, B. Sadoun, H. M. S. Badar, and W. Gao, “REAL-TIME ROAD DAMAGE DETECTION USING AN OPTIMIZED YOLOV9S-FUSION IN IOT INFRASTRUCTURE,” IEEE Internet Things J., vol. 12, no. 11, pp. 17649–17660, 2025, doi: https://doi.org/10.1109/JIOT.2025.3537640
Z. Demirel, S. T. Nasraldeen, Ö. Pehlivan, S. Shoman, M. Albdairi, and A. Almusawi, “COMPARATIVE EVALUATION OF YOLO AND GEMINI AI MODELS FOR ROAD DAMAGE DETECTION AND MAPPING,” Future Transportation, vol. 5, no. 3, p. 91, 2025, doi: https://doi.org/10.3390/futuretransp5030091
P. I. Wayan and others, “KLASIFIKASI CITRA MENGGUNAKAN CONVOLUTIONAL NEURAL NETWORK (CNN) PADA CALTECH 101,” Teknik ITS, vol. 5, 2016, doi: https://doi.org/10.12962/j23373539.v5i1.15696
J. P. Tanjung, “CLASSIFICATION OF WHEAT SEEDS USING NEURAL NETWORK BACKPROPAGATION ALGORITHM,” Journal of Informatics and Telecommunication Engineering, vol. 4, no. 2, pp. 335–342, Jan. 2021, doi: https://doi.org/10.31289/jite.v4i2.4449
N. Shakhovska, V. Yakovyna, M. Mysak, S. A. Mitoulis, S. Argyroudis, and Y. Syerov, “REAL-TIME MONITORING OF ROAD NETWORKS FOR PAVEMENT DAMAGE DETECTION BASED ON PREPROCESSING AND NEURAL NETWORKS,” Big Data and Cognitive Computing, vol. 8, no. 10, p. 136, 2024, doi: https://doi.org/10.3390/bdcc8100136
C. S. Wijaya, I. Wijaya, M. T. Shidqi, and D. Novita, “ANALISIS IMPLEMENTASI ARTIFICIAL INTELLIGENCE UNTUK BISNIS: SYSTEMATIC LITERATURE REVIEW,” Computer Science and Information Technology, vol. 4, no. 2, p. 133, 2023, doi: https://doi.org/10.46576/device.v4i2.4037
Copyright (c) 2026 Anggun Puspita Sari, Kariyam Kariyam, Feri Wijayanto, Edy Widodo
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Authors who publish with this Journal agree to the following terms:
- Author retain copyright and grant the journal right of first publication with the work simultaneously licensed under a creative commons attribution license that allow others to share the work within an acknowledgement of the work’s authorship and initial publication of this journal.
- Authors are able to enter into separate, additional contractual arrangement for the non-exclusive distribution of the journal’s published version of the work (e.g. acknowledgement of its initial publication in this journal).
- Authors are permitted and encouraged to post their work online (e.g. in institutional repositories or on their websites) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published works.
1.gif)
