TY - JOUR
T1 - Texture-sensitive superpixeling and adaptive thresholding for effective segmentation of sea ice floes in high-resolution optical images
AU - Chai, Yanmei
AU - Ren, Jinchang
AU - Hwang, Byongjun
AU - Wang, Jian
AU - Fan, Dan
AU - Yan, Yijun
AU - Zhu, Shiwen
N1 - Funding Information:
Manuscript received June 30, 2020; revised October 19, 2020 and November 17, 2020; accepted November 23, 2020. Date of publication November 25, 2020; date of current version January 6, 2021. This work was supported in part by the Natural and Environmental Research Council under Grant NE/S002545/1, U.K., in part by the Visiting Scholarship funded by the China Scholarship Council under Grant 201906495007, in part by the Dazhi Scholarship of the Guangdong Polytechnic Normal University, National Natural Science Foundation of China under Grant 62072122, and in part by the Education Department of Guangdong Province under Grant 2019KSYS009. (Corresponding author: Jinchang Ren.) Yanmei Chai, Jian Wang, Dan Fan, and Shiwei Zhu are with the School of Information, Central University of Finance and Economics, Beijing 100081, China (e-mail: [email protected]; [email protected]; fandan@ cufe.edu.cn; [email protected]).
Publisher Copyright:
© 2020 IEEE.
Copyright:
Copyright 2021 Elsevier B.V., All rights reserved.
PY - 2021/1/1
Y1 - 2021/1/1
N2 - Efficient and accurate segmentation of sea ice floes from high-resolution optical (HRO) remote sensing images is crucial for understanding of sea ice evolutions and climate changes, especially in coping with the large data volume. Existing methods suffer from noise interference and the mixture of water and ice caused high segmentation error and less robustness. In this article, we propose a novel sea ice floe segmentation algorithm from HRO images based on texture-sensitive superpixeling and two-stage thresholding. First, sparse components are extracted from the HRO images using the robust principal component analysis (RPCA), and noise is removed by the bilateral filter. The enhanced image is obtained by combining the low-rank matrix and the sparse components. Second, a texture-sensitive simple linear iterative clustering (SLIC) superpixel algorithm is introduced for presegmentation of the enhanced HRO image. Third, a learning-based adaptive thresholding in the two stages is employed to generate the refined segmentation from the derived superpixels blocks. The efficacy of the proposed method is validated on two HRO images using visual assessment, quantitative evaluation (with seven metrics), and histogram comparison. The superior performance of the proposed method has demonstrated its efficacy for sea ice floe segmentation.
AB - Efficient and accurate segmentation of sea ice floes from high-resolution optical (HRO) remote sensing images is crucial for understanding of sea ice evolutions and climate changes, especially in coping with the large data volume. Existing methods suffer from noise interference and the mixture of water and ice caused high segmentation error and less robustness. In this article, we propose a novel sea ice floe segmentation algorithm from HRO images based on texture-sensitive superpixeling and two-stage thresholding. First, sparse components are extracted from the HRO images using the robust principal component analysis (RPCA), and noise is removed by the bilateral filter. The enhanced image is obtained by combining the low-rank matrix and the sparse components. Second, a texture-sensitive simple linear iterative clustering (SLIC) superpixel algorithm is introduced for presegmentation of the enhanced HRO image. Third, a learning-based adaptive thresholding in the two stages is employed to generate the refined segmentation from the derived superpixels blocks. The efficacy of the proposed method is validated on two HRO images using visual assessment, quantitative evaluation (with seven metrics), and histogram comparison. The superior performance of the proposed method has demonstrated its efficacy for sea ice floe segmentation.
KW - adaptive two-stage thresholding
KW - high-resolution optical image
KW - low-rank sparse representation
KW - Sea ice floe segmentation
KW - texture-sensitive superpixeling
KW - sea ice floe segmentation
KW - Adaptive two-stage thresholding
KW - high-resolution optical (HRO) image
UR - http://www.scopus.com/inward/record.url?scp=85097178134&partnerID=8YFLogxK
U2 - 10.1109/JSTARS.2020.3040614
DO - 10.1109/JSTARS.2020.3040614
M3 - Article
AN - SCOPUS:85097178134
VL - 14
SP - 577
EP - 586
JO - IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
JF - IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
SN - 1939-1404
M1 - 9271815
ER -