Enhancing underwater archaeology with computer vision: A deep learning approach for shipwreck detection

View at publisher Summarise with VIRAT
Authors: Ansel Cyril Paul, R. Gunavathi, Milan Chakraborty
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Underwater archaeology, a vital field for under-standing human history and prehistory, faces numerous challenges, including poor visibility, limited accessibility, and the fragility of submerged sites. This research leverages the power of computer vision and deep learning to enhance the detection and classification of shipwrecks, a crucial aspect of underwater archaeological exploration. We trained a Convolutional Neural Network (CNN) on a comprehensive dataset of underwater images, employing histogram equalization as a preprocessing technique to improve image quality. Our model achieved a remarkable 99.84% test accuracy, showcasing its potential to automate the analysis of underwater imagery and significantly accelerate the discovery and documentation of submerged cultural heritage. This paper provides a detailed account of our methodology, the results obtained, and the broader implications for the future of under water archaeology.

Fields of Study & Subjects

Maritime and Coastal Archaeology Advanced Neural Network Applications Image Processing and 3D Reconstruction

Publication details

Journal / source
CRC Press eBooks
Pages
138–145
Publication date
25 March 2026
Language
EN
Document type
Book chapter