Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Water Body Extraction From Very High-Resolution Remote Sensing Imagery Using Deep U-Net and a Superpixel-Based Conditional Random Field Model

Feng, Wenqing ; Sui, Haigang ; Huang, Weiming LU ; Xu, Chuan and An, Kaiqiang (2019) In IEEE Geoscience and Remote Sensing Letters 16(4). p.618-622
Abstract

Water body extraction (WBE) has attracted considerable attention in the field of remote sensing image analysis. Herein, we present an enhanced deep convolutional encoder-decoder (DCED) network (or Deep U-Net) specifically tailored to WBE from remote sensing images by applying superpixel segmentation and conditional random fields (CRFs). First, we preclassify the entire remote sensing image into the water and nonwater areas via Deep U-Net, using the results of class membership probabilities as the unary potential in the CRF model. The pairwise potential of CRF is defined by a linear combination of Gaussian kernels, which forms a fully connected neighbor structure. Next, regional restriction is incorporated into the approach to enhance... (More)

Water body extraction (WBE) has attracted considerable attention in the field of remote sensing image analysis. Herein, we present an enhanced deep convolutional encoder-decoder (DCED) network (or Deep U-Net) specifically tailored to WBE from remote sensing images by applying superpixel segmentation and conditional random fields (CRFs). First, we preclassify the entire remote sensing image into the water and nonwater areas via Deep U-Net, using the results of class membership probabilities as the unary potential in the CRF model. The pairwise potential of CRF is defined by a linear combination of Gaussian kernels, which forms a fully connected neighbor structure. Next, regional restriction is incorporated into the approach to enhance the consistency of the connected area. We use the simple linear iterative clustering algorithm to generate superpixels and correct the binary classification results by calculating their average posterior probabilities. Finally, a highly efficient approximate inference algorithm, mean-field inference, is generated for the final model. The results from the experimental application to GaoFen-2 images and WorldView-2 images demonstrate that the proposed approach exhibits competitive quantitative and qualitative performance, which effectively reduces salt-and-pepper noise and retains the edge structures of water bodies. Compared to existing state-of-the-art methods, our proposed method achieves superior final results.

(Less)
Please use this url to cite or link to this publication:
author
; ; ; and
organization
publishing date
type
Contribution to journal
publication status
published
subject
keywords
Conditional random fields (CRFs), Deep U-Net, Feature extraction, Image segmentation, Kernel, regional restriction (RR), Remote sensing, Semantics, superpixel, water body extraction (WBE)., Water conservation, Water resources
in
IEEE Geoscience and Remote Sensing Letters
volume
16
issue
4
pages
618 - 622
publisher
IEEE - Institute of Electrical and Electronics Engineers Inc.
external identifiers
  • scopus:85058664332
ISSN
1545-598X
DOI
10.1109/LGRS.2018.2879492
language
English
LU publication?
yes
id
373714a4-5e3b-471b-b484-6b840bb2a976
date added to LUP
2019-01-10 09:04:01
date last changed
2022-04-25 20:01:46
@article{373714a4-5e3b-471b-b484-6b840bb2a976,
  abstract     = {{<p>Water body extraction (WBE) has attracted considerable attention in the field of remote sensing image analysis. Herein, we present an enhanced deep convolutional encoder-decoder (DCED) network (or Deep U-Net) specifically tailored to WBE from remote sensing images by applying superpixel segmentation and conditional random fields (CRFs). First, we preclassify the entire remote sensing image into the water and nonwater areas via Deep U-Net, using the results of class membership probabilities as the unary potential in the CRF model. The pairwise potential of CRF is defined by a linear combination of Gaussian kernels, which forms a fully connected neighbor structure. Next, regional restriction is incorporated into the approach to enhance the consistency of the connected area. We use the simple linear iterative clustering algorithm to generate superpixels and correct the binary classification results by calculating their average posterior probabilities. Finally, a highly efficient approximate inference algorithm, mean-field inference, is generated for the final model. The results from the experimental application to GaoFen-2 images and WorldView-2 images demonstrate that the proposed approach exhibits competitive quantitative and qualitative performance, which effectively reduces salt-and-pepper noise and retains the edge structures of water bodies. Compared to existing state-of-the-art methods, our proposed method achieves superior final results.</p>}},
  author       = {{Feng, Wenqing and Sui, Haigang and Huang, Weiming and Xu, Chuan and An, Kaiqiang}},
  issn         = {{1545-598X}},
  keywords     = {{Conditional random fields (CRFs); Deep U-Net; Feature extraction; Image segmentation; Kernel; regional restriction (RR); Remote sensing; Semantics; superpixel; water body extraction (WBE).; Water conservation; Water resources}},
  language     = {{eng}},
  number       = {{4}},
  pages        = {{618--622}},
  publisher    = {{IEEE - Institute of Electrical and Electronics Engineers Inc.}},
  series       = {{IEEE Geoscience and Remote Sensing Letters}},
  title        = {{Water Body Extraction From Very High-Resolution Remote Sensing Imagery Using Deep U-Net and a Superpixel-Based Conditional Random Field Model}},
  url          = {{http://dx.doi.org/10.1109/LGRS.2018.2879492}},
  doi          = {{10.1109/LGRS.2018.2879492}},
  volume       = {{16}},
  year         = {{2019}},
}