Research on natural image semantic segmentation method based on improved U-Net model

Xiangtian Ji; Shuo Jiang

doi:10.1117/12.3009019

27 October 2023 Research on natural image semantic segmentation method based on improved U-Net model

Xiangtian Ji, Shuo Jiang

Proceedings Volume 12922, Third International Conference on Electronics, Electrical and Information Engineering (ICEEIE 2023); 129222I (2023) https://doi.org/10.1117/12.3009019
Event: The Third International Conference on Electronics, Electrical and Information Engineering (ICEEIE 2023), 2023, Xiamen, China

Abstract

This project intends to combine the U-Net algorithm and the super pixel algorithm to study the high-resolution remote sensing image semantic segmentation algorithm based on the U-Net algorithm. A pixel-based end-to-end semantic segmentation method is proposed. By extending the original dataset, a dual classification model is built for each type of ground object. The final semantically segmented image is obtained by combining individual predicted sub-maps. The rapid identification of remote sensing images such as agricultural land, houses and rivers based on multi-band SAR images is completed using small-sample SAR images. The overall classification accuracy is 95.7%. Experimental results show that this algorithm can adapt to various environments and handle complex semantic information well.

(2023) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

Xiangtian Ji and Shuo Jiang "Research on natural image semantic segmentation method based on improved U-Net model", Proc. SPIE 12922, Third International Conference on Electronics, Electrical and Information Engineering (ICEEIE 2023), 129222I (27 October 2023); https://doi.org/10.1117/12.3009019

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