KLASIFIKASI KERUSAKAN JALAN RAYA BERBASIS CITRA UDARA MENGGUNAKAN OBJECT-BASED IMAGE-ANALYSIS (OBIA)
DOI:
https://doi.org/10.21009/03.1201.FA05Abstract
Abstrak
Kerusakan jalan merupakan masalah serius yang dapat menyebabkan kemacetan, kecelakaan, dan risiko keselamatan. Pada makalah ini, diusulkan penggunaan citra drone untuk mengidentifikasi kerusakan jalan raya. Pendekatan pixel-to-pixel pada citra drone yang beresolusi spasial tinggi sulit dilakukan karena kerusakan jalan menyebar ke beberapa piksel. Oleh karena itu, pendekatan OBIA digunakan untuk mengklasifikasikan kerusakan jalan raya dengan memfokuskan objek sebagai kesatuan. Klasifikasi citra dilakukan menggunakan pendekatan Object-Based Image-Analysis (OBIA) dengan mengimplementasikan algoritma Simple Linear Iterative Clustering (SLIC). SLIC akan dieksplorasi dengan memvariasikan jumlah cluster untuk mendapatkan metode ekstraksi ciri yang tepat sebelum dilakukan klasifikasi menggunakan CNN. Terdapat dua objek yang dipilih untuk memvalidasi hasil klasifikasi, yaitu jalan rusak dan jalan tidak rusak. Hasil eksperimen menunjukkan bahwa pendekatan OBIA mampu mengidentifikasi objek kerusakan jalan dengan optimal. Bahkan, pemilihan jumlah cluster juga mempengaruhi nilai akurasi klasifikasi. Penggunaan jumlah maksimal 300 cluster memberikan hasil akurasi klasifikasi terbaik dengan peningkatan 13,87% dibandingkan dengan 500 cluster.
Kata-kata kunci: Kerusakan Jalan, OBIA, SLIC, Klasifikasi, Unmanned Aerial Vehicles (UAV)
Abstract
Road damage is a serious problem that can lead to congestion, accidents, and safety risks. This paper proposes the use of drone imagery for identifying road damage. The pixel-to-pixel approach on high-resolution drone imagery is challenging as road damage can span across multiple pixels. Therefore, an Object-Based Image Analysis (OBIA) approach is employed to classify road damage by focusing on objects as a whole. The classification of images is conducted using the Object-Based Image Analysis (OBIA) approach, implementing the Simple Linear Iterative Clustering (SLIC) algorithm. SLIC will be explored by varying the number of clusters to obtain appropriate feature extraction methods before performing classification using a Convolutional Neural Network (CNN). Two objects, namely damaged road and undamaged road, are selected to validate the classification results. The experimental results demonstrate that the OBIA approach can effectively identify road damage objects. Moreover, the selection of the number of clusters also influences the classification accuracy. The use of a maximum of 300 clusters yields the highest classification accuracy with a 13.87% improvement compared to 500 clusters.
Keywords: Road Damage, OBIA, SLIC, Classification, Unmanned Aerial Vehicles (UAV)
References
[2] D. Šulyová, J. Vodák, “Benefits and Limitations of Using UAVs in Different Areas with a Focus on the Environment,” Journal of Information Control and Management System, vol. 18, no. 2, 2021.
[3] R. Han et al., “Advantage of Combining OBIA and Classifier Ensemble Method for Very High-Resolution Satellite Imagery Classification,” Journal of Sensors, pp. 1-15, 2020, doi: 10.1155/2020/8855509.
[4] D. Ventura et al., “Mapping and Classification of Ecologically Sensitive Marine Habitats Using Unmanned Aerial Vehicle (UAV) Imagery and Object-Based Image Analysis (OBIA),” Remote Sensing, vol. 10, no. 9, p. 1331, 2018, doi: 10.3390/rs10091331.
[5] T. Blaschke, “Object based image analysis for remote sensing,” ISPRS journal of photogrammetry and remote sensing, vol. 65, no. 1, pp. 2-16, 2010, doi: 10.1016/j.isprsjprs.2009.06.004.
[6] T. Blaschke et al., “Geographic Object-Based Image Analysis - Towards a new paradigm,” ISPRS journal of photogrammetry and remote sensing, vol. 87, pp. 180-191, 2014, doi: 10.1016/j.isprsjprs.2013.09.014.
[7] J. Naranjo-Torres et al., “A review of convolutional neural network applied to fruit image processing,” Applied Sciences, vol. 10, no. 10, p. 3443, 2020, doi: 10.3390/app10103443.
[8] V. Aggarwal, G. Kaur, “A review:deep learning technique for image classification,” ACCENTS transactions on image processing and computer vision, vol. 4, no. 11, pp. 21-25, 2018, doi: 10.19101/tipcv.2018.411003.
[9] L. Liu et al., “Deep Learning for Generic Object Detection: A Survey,” International journal of computer vision, vol. 128, no. 2, pp. 261-318, 2020, doi: 10.1007/s11263-019-01247-4.
[10] D. Shen, G. Wu, H. Suk, “Deep Learning in Medical Image Analysis,” Annual review of biomedical engineering, vol. 19. no. 3, pp. 221-248, 2017.
[11] L. A. Silva et al., “An Architectural Multi-Agent System for a Pavement Monitoring System with Pothole Recognition in UAV Images,” Sensors, vol. 20, no. 21, p. 6205, 2020, doi: 10.3390/s20216205.
[12] V. Mandal, A. R. Mussah, Y. Adu-Gyamfi, “Deep Learning Frameworks for Pavement Distress Classification: A Comparative Analysis,” IEEE International Conference on Big Data (Big Data), pp. 5577-5583, 2020, doi: 10.1109/BigData50022.2020.9378047.
[13] T. Ishtiak et al., “A convolutional neural network approach for road anomalies detection in Bangladesh with image thresholding,” Third World Conference on Smart Trends in Systems Security and Sustainablity WorldS4, pp. 376-382, 2019, doi: 10.1109/WorldS4.2019.8903936.
[14] A. Fauzan, “Klasifikasi Tutupan Lahan Berbasis SLIC-Convolutional Neural Networks (CNN) dan Citra Pleiades,” Universitas Negeri Jakarta, 2022.
[15] H. Wang et al., “BSLIC: SLIC Superpixels Based on Boundary Term,” Symmetry, vol. 9, no. 3. 2017, doi: 10.3390/sym9030031.
[16] M. Zheng et al., “A Method for Analyzing the Performance Impact of Imbalanced Binary Data on Machine Learning Models,” Axioms, vol. 11, no. 11, 2022, doi: 10.3390/axioms11110607.
