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{206678,
        author = {Prof. Nivin K S and Nithya BP and Krishna Sharan S Pandith and Nived C and Roshedh S U},
        title = {Structural Damage Detection In Enclosed Spaces: Using Machine Learning},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {13},
        number = {no},
        pages = {159-168},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=206678},
        abstract = {The structural integrity of enclosed spaces in marine vessels—such as ballast tanks, cargo holds, voids, and engine compartments—is vital for ship safety and performance. These areas are often difficult to access, poorly lit, and vulnerable to corrosion, fatigue, and stress, making traditional inspection methods—manual visual checks and non-destructive testing—time-consuming, hazardous, and inconsistent. This project introduces an AI-powered, automated damage detection system designed for such confined ship environments. Using drones or robotic crawlers equipped with RGB and infrared cameras, the system captures high-resolution imagery, which is enhanced through preprocessing techniques like CLAHE, gamma correction, and bilateral filtering. To manage large images effectively, it applies Slicing Aided Hyper Inference (SAHI), breaking them into smaller overlapping segments. These segments are analyzed using an ensemble of YOLOv8 and RetinaNet—object detection models chosen for their balance of speed and accuracy. Their outputs are fused to improve detection reliability and reduce false results. The system generates labeled damage areas, severity levels, and timestamped logs, all displayed on a real-time dashboard. It supports CSV, JSON, and PDF exports, with alert features for critical damage. A feedback and retraining loop allow personnel to annotate errors, helping improve model performance over time. By combining autonomous navigation, deep learning, and smart visualization, this solution offers a safer, more efficient alternative to manual inspections, promising reduced costs and enhanced safety across maritime operations.},
        keywords = {Structural Damage Detection, Enclosed Ship Spaces, Computer Vision, Deep Learning, Drone Inspection, Maritime Safety.},
        month = {July},
        }