Linking feature combination and Swin Transformer model for extracting dragon fruit from Sentinel-2A remote sensing image

Zhiwei Qi; Bo Wei

doi:10.1117/12.3024192

21 February 2024 Linking feature combination and Swin Transformer model for extracting dragon fruit from Sentinel-2A remote sensing image

Zhiwei Qi, Bo Wei

Author Affiliations +

Proceedings Volume 12988, Second International Conference on Environmental Remote Sensing and Geographic Information Technology (ERSGIT 2023); 129880G (2024) https://doi.org/10.1117/12.3024192
Event: Second International Conference on Environmental Remote Sensing and Geographic Information Technology (ERSGIT 2023), 2023, Xi’an, China

Abstract

As an important tropical fruit, the yield and quality of dragon fruit play a key role in stabilizing the fruit market. Remote sensing technology can quickly and efficiently extract dragon fruit information over a wide area, providing important data support for orchard management, disease prevention and control, and ecological environment monitoring, and thus contributing to the sustainable development of agriculture. However, the small spectral difference between dragon fruit and other fruit trees on remote sensing images makes the extraction of dragon fruit a challenge. In this study, feature combination and the Swin Transformer deep learning model have been selected as techniques for extracting dragon fruit of the study area. The meticulous band selection is carried out for the original Sentinel-2A multispectral image, and a feature combination scheme is generated after adding the vegetation indices NDVI, EVI, RVI and DVI, which is a dataset including 14 features for the Swin Transformer model to extract dragon fruit. Based on the semantic segmentation of the Swin Transformer model, the dragon fruit is extracted and its effect is compared with other classical deep learning models FCN, Unet and DeepLabV3. The results show that the Swin Transformer model obtains the best extraction accuracy for dragon fruit compared with the compared models.

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

Citation Download Citation

Zhiwei Qi and Bo Wei "Linking feature combination and Swin Transformer model for extracting dragon fruit from Sentinel-2A remote sensing image", Proc. SPIE 12988, Second International Conference on Environmental Remote Sensing and Geographic Information Technology (ERSGIT 2023), 129880G (21 February 2024); https://doi.org/10.1117/12.3024192

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

Members: $17.00

Non-members: $21.00 ADD TO CART

PROCEEDINGS
6 PAGES

DOWNLOAD PAPER SAVE TO MY LIBRARY

GET CITATION

RIGHTS & PERMISSIONS

Get copyright permission Get copyright permission on Copyright Marketplace

KEYWORDS

Transformers

Vegetation

Image segmentation

Remote sensing

Performance modeling

Deep learning

Feature extraction

Show All Keywords

Keywords/Phrases

Search In:

Publication Years