Research on ship object detection based on deep learning

Junkuan Jin; Yingjie Xiao

doi:10.1117/12.3030591

16 May 2024 Research on ship object detection based on deep learning

Junkuan Jin, Yingjie Xiao

Proceedings Volume 13160, Fourth International Conference on Smart City Engineering and Public Transportation (SCEPT 2024); 131600P (2024) https://doi.org/10.1117/12.3030591
Event: 4th International Conference on Smart City Engineering and Public Transportation (SCEPT 2024), 2024, Beijin, China

Abstract

In order to enhance ship detection accuracy and improve processing speed, an enhanced ship detection algorithm based on the YOLOv5 algorithm is introduced in this paper. Firstly, by adopting the Bidirectional Weighted Feature Pyramid Network, the model achieves higher accuracy in ship detection. Secondly, all the regular convolutions in the model are replaced with Ghost convolutions to achieve model lightweighting. The experimental results show that the average precision has been improved to 83.5%, exhibiting a 2.6 percentage point increase compared to the original model. The improved algorithm reduces the model's parameter size and computational complexity while maintaining high precision in ship detection.

(2024) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

Junkuan Jin and Yingjie Xiao "Research on ship object detection based on deep learning", Proc. SPIE 13160, Fourth International Conference on Smart City Engineering and Public Transportation (SCEPT 2024), 131600P (16 May 2024); https://doi.org/10.1117/12.3030591

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available