Multi-scale fully convolutional network-based land cover classification method for remote sensing images

Jun Yao; Wei Feng; Le Yang; Lin Ju; Xin Li; Xiangwei Zhao

doi:10.1117/12.2624366

27 March 2022 Multi-scale fully convolutional network-based land cover classification method for remote sensing images

Jun Yao, Wei Feng, Le Yang, Lin Ju, Xin Li, Xiangwei Zhao

Proceedings Volume 12169, Eighth Symposium on Novel Photoelectronic Detection Technology and Applications; 121695F (2022) https://doi.org/10.1117/12.2624366
Event: Eighth Symposium on Novel Photoelectronic Detection Technology and Applications, 2021, Kunming, China

Abstract

With the rapid development of aviation technology, computer technology and sensor technology, remote sensing images have been widely used in transportation planning, environmental protection, fine agriculture, military monitoring and other fields. Land cover classification of remote sensing images has been a hot issue in recent years, and the early traditional methods can only extract shallow features, which is difficult to guarantee the classification accuracy. In recent years, deep learning theory has been developed rapidly with its advantage of automatically extracting deep abstract features, which has achieved great success in various fields and has also been introduced into the field of remote sensing image land cover classification. However, most of the deep learning-based land cover classification methods use local connectivity, which leads to the difficulty of simultaneous extraction of multi-scale features and a large loss of global information. In order to solve the above problems, this paper proposes a multi-scale fully convolutional network, which can accurately extract features at different scales, take into account the global feature distribution while retaining local details, and exploit the rich information contained in remote sensing images, while greatly reducing the number of parameters. Comparative experiments show that the proposed method can effectively extract the spatial distribution of farmland and houses at large scales while preserving the boundary contours of water systems and roads at small scales, which significantly improves the classification effect.

Citation Download Citation

Jun Yao, Wei Feng, Le Yang, Lin Ju, Xin Li, and Xiangwei Zhao "Multi-scale fully convolutional network-based land cover classification method for remote sensing images", Proc. SPIE 12169, Eighth Symposium on Novel Photoelectronic Detection Technology and Applications, 121695F (27 March 2022); https://doi.org/10.1117/12.2624366

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

Image classification

Remote sensing

Image processing

Convolution

Principal component analysis

Roads

Image segmentation

Show All Keywords

Keywords/Phrases

Search In:

Publication Years