Implementation of Computer Vision in VIDAS for Road Damage Detection Using the SSD Algorithm

DOI

10.2991/978-94-6463-678-9_55How to use a DOI?

Keywords

Road damage detection; SSD model; MobileNet V2; road infrastructure monitoring

Abstract

Detecting road damage is essential for maintaining safe and efficient transportation infrastructure. Automated systems for identifying road defects, such as cracks and potholes, play a crucial role in enhancing the speed and accuracy of road inspections, enabling timely and cost-effective maintenance. This study focuses on implementing a Single Shot Detector (SSD) model to detect various types of road damage. Built on the Mobile-Net V2 architecture with FPN-Lite, the SSD model was trained on a dataset containing images of road defects and optimized over 50,000 iterations with a batch size of 16 for real-time detection. The model's effectiveness was evaluated using the mean Average Precision (mAP) metric at a threshold of 0.5:0.95, yielding an overall mAP score of 13.22%. The results showed that the model performed best in detecting faded markings (mAP 24.77%) and alligator cracks (mAP 20.49%), while struggling with transverse cracks (mAP 1.06%) and potholes (mAP 9.18%). These findings highlight the potential of the SSD model to support real-time road maintenance applications by reliably detecting certain defect types, but also indicate areas for further enhancement. Improvements in model training and data augmentation could elevate detection accuracy across all defect categories, contributing to more effective and proactive road infrastructure monitoring. This advancement in automated road damage detection ultimately supports faster, data-driven maintenance decisions, leading to safer and longer-lasting road networks.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Rian Rahmanda Putra
AU  - Anggi Nidya Sari
AU  - Agum Try Wardhana
AU  - Yulian Mirza
AU  - Muhammad Fikri
AU  - Norca Praditya
PY  - 2025
DA  - 2025/05/01
TI  - Implementation of Computer Vision in VIDAS for Road Damage Detection Using the SSD Algorithm
BT  - Proceedings of the 8th FIRST 2024 International Conference on Global Innovations (FIRST-ESCSI 2024 )
PB  - Atlantis Press
SP  - 576
EP  - 586
SN  - 2352-5401
UR  - https://doi.org/10.2991/978-94-6463-678-9_55
DO  - 10.2991/978-94-6463-678-9_55
ID  - Putra2025
ER  -