Paper Info
Title
An Automatic Extraction Method For Lakes And Reservoirs Using Satellite Images
Abstract
Lakes and reservoirs are important parts of ecosystems and social developments. Separating scientific research on lakes and reservoirs is not conducive to measure water regulation services or defend against natural disasters. Accurate surface water information is of great significance to lake and reservoir management. To address this problem, we proposed a new automated lake and reservoir extraction process (LREP), which applied the modified two-mode (MTM) method to extract water from partial images by calculating thresholds in the normalized difference water index (NDWI) and the near-infrared (NIR) band; then, we estimated the segmentation thresholds from a set of NDWI thresholds obtained for the remaining images. Subsequently, a series of feature factors was used to remove rivers, shadows, and ponds. Finally, a new shape parameter called complexity was introduced to describe the edge characteristics of water bodies, distinguishing lakes and reservoirs from paddyfields and remaining ponds. We verified the algorithm using 154 Gaofen-2 images covering three cities: Liaoyuan, Xianning, and Zigong. The results indicated that the new LREP performed well with the overall accuracies of 99.38%, 99.27%, and 98.72% and the kappa coefficients of 0.92, 0.96, and 0.92 in Liaoyuan, Xianning, and Zigong, respectively. This method has the potential to monitor a variety of lakes and reservoirs, including lost lakes, new reservoirs, and surface water quality.
Year
DOI
Venue
2019
10.1109/ACCESS.2019.2916148
IEEE ACCESS
Keywords
Field
DocType
GF-2, lakes, reservoirs, two-mode, NDWI, threshold
Satellite,Computer science,Remote sensing,Distributed computing
Journal
Volume
ISSN
Citations 
7
2169-3536
0
PageRank 
References 
Authors
0.34
0
7
Name
Order
Citations
PageRank
Lingkui Meng13811.56
Zhiyuan Zhang200.34
Wen Zhang331636.67
Jinan Ye400.34
Chenhan Wu500.68
DeQing Chen6896.95
Chao Song710015.52