Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{187969,
        author = {Dr.MONIKA D.ROKADE and Sarfaraz Sajid Inamdar and Shakil Akil Halwai and Kartik Raju Rathod and Nikhil Dhammasang Kasbe and Dr.Sunil S.Khatal},
        title = {A Comprehensive Review on Software-Based Deep Learning Techniques for Pothole Detection in Smart Transportation Systems},
        journal = {International Journal of Innovative Research in Technology},
        year = {2025},
        volume = {12},
        number = {7},
        pages = {281-288},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=187969},
        abstract = {Road surface degradation, particularly in the form of potholes, poses a serious challenge to transportation safety and infrastructure management. Over the past decade, several researchers have explored software-based detection approaches leveraging advancements in deep learning and computer vision. This review paper presents a comprehensive analysis of existing pothole detection methods that utilize software-driven frameworks without the need for additional hardware sensors. The study categorizes existing techniques based on their underlying algorithms such as Convolutional Neural Networks (CNN), You Only Look Once (YOLO) architectures, and image segmentation models. It discusses key aspects including dataset diversity, computational complexity, model accuracy, and deployment feasibility in real-world environments. Furthermore, this review highlights the evolution from traditional edge-based and vibration sensor methods to fully automated software-based detection systems powered by artificial intelligence. The paper also identifies research gaps related to model generalization across varying environmental conditions and real-time integration challenges. The review concludes that software-based deep learning frameworks, particularly YOLOv8 and similar architectures, demonstrate strong potential for scalable,real-time pothole detection within smart city and intelligent transportation applications.},
        keywords = {Pothole Detection, Deep Learning, YOLOv8, Computer Vision, Road Surface Analysis, Software-Based Detection, Smart Transportation, Image Processing, Convolutional Neural Networks (CNN), Object Detection.},
        month = {November},
        }