Pothole Detection utilising Machine Learning: A Review

Year : 2025 | Volume : 16 | Issue : 01 | Page : 35 43
    By

    Aayush Chandorkar,

  • Swaraj Salvi,

  • Purva Mhatre,

  • Bhagyalaksmi V,

  1. Student, Department of Electronics and Telecommunications , K. C. (Kishinchand Chellaram) College of Engineering, Thane(W), Maharashtra, India
  2. Student, Department of Electronics and Telecommunications , K. C. (Kishinchand Chellaram) College of Engineering, Thane(W), Maharashtra, India
  3. Student, Department of Electronics and Telecommunications , K. C. (Kishinchand Chellaram) College of Engineering, Thane(W), Maharashtra, India
  4. Assistant Professor, Department of Electronics and Telecommunications , K. C. (Kishinchand Chellaram) College of Engineering, Thane(W), Maharashtra, India

Abstract

Potholes must be found and fixed quickly in order to maintain infrastructure, maximize transportation systems, and guarantee road safety. Using the Sequential API and the Keras library, this study presents a neural network model for pothole detection. Convolutional layers with ReLU activation, global average pooling, dense layers with dropout, and softmax activation for binary classification make up the model architecture. Image loading, resizing, array conversion, labeling, shuffling, normalization, and one-hot encoding are all examples of preprocessing steps. By precisely identifying potholes, this technology seeks to improve overall road maintenance and traffic flow management while facilitating prompt repairs. Every road, regardless of its engineering features or traffic volume, needs routine maintenance on a regular basis. Most routine maintenance tasks are small-scale, geographically distributed, and frequently involve manual labor. To a considerable extent, the requirement for routine maintenance can be predicted and is planned for at specific times throughout the year. To avoid early road deterioration, routine maintenance should be performed on all roads on a regular basis. It is the engineers’ duty to make sure that funds are allocated in the annual maintenance plan. The activities vary in frequency. Moreover, routine maintenance tasks can be classified as either reactive or cyclic, albeit it’s not always easy to tell the difference between the two.

Keywords: Road Maintenance, Pothole Detection, Machine Learning, Arduino Microcontroller, Central Monitoring System.

[This article belongs to Journal of Control & Instrumentation ]

How to cite this article:
Aayush Chandorkar, Swaraj Salvi, Purva Mhatre, Bhagyalaksmi V. Pothole Detection utilising Machine Learning: A Review. Journal of Control & Instrumentation. 2025; 16(01):35-43.
How to cite this URL:
Aayush Chandorkar, Swaraj Salvi, Purva Mhatre, Bhagyalaksmi V. Pothole Detection utilising Machine Learning: A Review. Journal of Control & Instrumentation. 2025; 16(01):35-43. Available from: https://journals.stmjournals.com/joci/article=2025/view=194162


References

[1] Buza, Emir, Samir Omanovic, and Alvin Huseinovic. “Pothole detection with image processing and spectral clustering.” In Proceedings of the 2nd International Conference on Information Technology and Computer Networks, vol. 810, p. 4853. 2013.

[2]       Boopathi, D., B. P. Gagana, S. Tevari Girish, M. Kiran, and D. V. Mr. Channappa Gowda. “AUTOMATIC DETECTION AND FILLING OF POTHOLES.”

[3] Dhiman, Amita, and Reinhard Klette. “Pothole detection using computer vision and learning.” IEEE Transactions on Intelligent Transportation Systems 21, no. 8 (2019): 3536-3550.

[4] Ren, Jinting, and Duo Liu. “PADS: A reliable pothole detection system using machine learning.” In International Conference on Smart Computing and Communication, pp. 327-338. Cham: Springer International Publishing, 2016.

[5] Egaji, Oche Alexander, Gareth Evans, Mark Graham Griffiths, and Gregory Islas. “Real-time machine learning-based approach for pothole detection.” Expert Systems with Applications 184 (2021): 115562.

[6]. Varona, Braian, Ariel Monteserin, and Alfredo Teyseyre. “A deep learning approach to automatic road surface monitoring and pothole detection.” Personal and Ubiquitous Computing 24, no. 4 (2020): 519-534.

[7]. Ping, Ping, Xiaohui Yang, and Zeyu Gao. “A deep learning approach for street pothole detection.” In 2020 IEEE Sixth International Conference on Big Data Computing Service and Applications (BigDataService), pp. 198-204. IEEE, 2020

[8]. Bharat, Ricardo, Abiodun Ikotun, Absalom Ezugwu, Laith Abualigah, Mohammad Shehab, and Raed Abu Zitar10. “A real-time automatic pothole detection system using convolution neural networks.” Applied and Computational Engineering 6, no. 1 (2023): 879-886.

[9] Kotha, Maheshwari, Mounika Chadalavada, Sri Harshitha Karuturi, and Hrishikesh Venkataraman. “Potsense: Pothole detection on indian roads using smartphone sensors.” In Proceedings of the 1st ACM Workshop on Autonomous and Intelligent Mobile Systems, pp. 1-6. 2020.

[10] Song, Hyunwoo, Kihoon Baek, and Yungcheol Byun. “Pothole detection using machine learning.” Advanced Science and Technology (2018): 151-155.

[11] Kumar, Abhishek, Dhruba Jyoti Kalita, and Vibhav Prakash Singh. “A modern pothole detection technique using deep learning.” In 2nd International Conference on Data, Engineering and Applications (IDEA), pp. 1-5. IEEE, 2020.

[12] Kulkarni, Aniket, Nitish Mhalgi, Sagar Gurnani, and Nupur Giri. “Pothole detection system using machine learning on Android.” International Journal of Emerging Technology and Advanced Engineering 4, no. 7 (2014): 360-364.

[13] Acharjee, Chandrika, Somya Singhal, and Suman Deb. “Machine learning approaches rapid pothole detection from 2D images.” In International Conference on Computational Intelligence, Security and Internet of Things, pp. 108-119. Cham: Springer International Publishing, 2020.

[14] Tahir, Hassam, and Eun-Sung Jung. “Comparative study on distributed lightweight deep learning models for road pothole detection.” Sensors 23, no. 9 (2023): 4347.

[15] Kharel, Subash, and Khaled R. Ahmed. “Potholes detection using deep learning and area estimation using image processing.” In Proceedings of SAI Intelligent Systems Conference, pp. 373-388. Cham: Springer International Publishing, 2021.

[16] Kavati, Ilaiah. “Deep Learning-Based Pothole Detection for Intelligent Transportation Systems.” In Edge Analytics: Select Proceedings of 26th International Conference—ADCOM 2020, pp. 257-267. Singapore: Springer Singapore, 2022.

[17] Owusu, Kalysa Abena Owusua. “Building a pothole detection and tracking system.” (2019).

[18] Yebes, J. Javier, David Montero, and Ignacio Arriola. “Learning to automatically catch potholes in worldwide road scene images.” IEEE Intelligent Transportation Systems Magazine 13, no. 3 (2020): 192-205.

[19] Anik, Fahim Shahriare, Asif Ahmed, Avita Maheen, and Puily Alam. “Identification and comparative analysis of potholes using image processing algorithms.” PhD diss., BRAC Univeristy, 2018.

[20] Ryu, Seung-Ki, Taehyeong Kim, and Young-Ro Kim. “Image-based pothole detection system for ITS service and road management system.” Mathematical Problems in Engineering 2015 (2015): 1-10.

Regular Issue Subscription Review Article
Volume 16
Issue 01
Received 09/10/2024
Accepted 06/01/2025
Published 15/01/2025
Publication Time 98 Days
346

Login


My IP

PlumX Metrics