Year: 2024
Type: article
Abstract: This study explores an advanced methodology for detecting marine vessels via satellite imagery, utilizing a deep learning framework tailored for maritime surveillance. A novel convolutional neural network (CNN) architecture, optimized for varying vessel sizes and adverse conditions, achieved a detection accuracy of 94%, significantly surpassing traditional computer vision techniques. The approach integrates innovative preprocessing techniques and a robust post-processing framework to ensure reliability under challenging scenarios like partial cloud cover and rough sea states. This work demonstrates the potential of AI-driven solutions for enhancing real-world maritime operations, including national security and traffic management.
Authors: Abizer Jesawada -, Harsh Singh -, Kavita Patel -, Dipti Kumari -, Rohini Bhosale -
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4405979823