文献类型：期刊文献

英文题名：Oil spill segmentation in marine radar images based on the improved Sauvola algorithm

作者：Liu, Peng[1];Wang, Xuchong[1];Shao, Pengzhe[1];Zhao, Xingquan[1];Liu, Bingxin[1];Chen, Peng[1];Zhu, Xueyuan[1];Xu, Jin[2];Li, Ying[1]

机构：[1]Dalian Maritime Univ, Nav Coll, Dalian 116026, Peoples R China;[2]Guangdong Ocean Univ, Maritime Coll, Zhanjiang, Peoples R China

年份：2026

卷号：17

期号：2

起止页码：182

外文期刊名：REMOTE SENSING LETTERS

收录：SCI-EXPANDED(收录号：WOS:001658850400001)、、EI(收录号：20260319913747)、Scopus(收录号：2-s2.0-105027307621)、WOS

基金：This research was funded by the National Natural Science Foundation of China [52271359] and Fundamental Research Funds for the Central Universities [3132025141].

语种：英文

外文关键词：Oil spill segmentation; marine radar image; improved sauvola algorithm

外文摘要：Marine oil spill incidents pose persistent threats to coastal ecosystems and maritime operations. This study presents an enhanced segmentation methodology for oil spill detection in X-band marine radar images through an improved Sauvola algorithm. The proposed method integrates negative exponential modelling of the background clutter distribution with adaptive threshold optimization. Validated using operational data from the 2010 Dalian Xingang Harbour oil spill incident, the improved Sauvola algorithm demonstrates the segmentation results consistent with visual interpretation. Moreover, compared with six other methods, both the user accuracy and the producer accuracy of the improved Sauvola algorithm are larger than 96%.