文献类型：期刊文献

英文题名：Oil spill discrimination of multi-time-domain shipborne radar images using active contour model

作者：Xu, Jin[1];Pan, Xinxiang[1];Wu, Xuerui[2,3,4];Jia, Baozhu[1];Fei, Juan[5];Wang, Haixia[6];Li, Bo[7];Cui, Can[8]

机构：[1]Guangdong Ocean Univ, Maritime Coll, Zhanjiang, Guangdong, Peoples R China;[2]Chifeng Univ, Sch Resources Environm & Architectural Engn, Chifeng, Inner Mongolia, Peoples R China;[3]Chinese Acad Sci, Shanghai Astron Observ, Shanghai, Peoples R China;[4]Shanghai Key Lab Space Nav & Positioning Tech, Shanghai, Peoples R China;[5]Lingnan Normal Univ, Informat Engn Coll, Zhanjiang, Guangdong, Peoples R China;[6]Dalian Maritime Univ, Nav Coll, Dalian, Liaoning, Peoples R China;[7]Guangdong Ocean Univ, Continuing Educ Coll, Zhanjiang, Guangdong, Peoples R China;[8]Shenyang Aerosp Univ, Civil Aviat Coll, Shenyang, Liaoning, Peoples R China

年份：2021

卷号：8

期号：1

外文期刊名：GEOSCIENCE LETTERS

收录：SCI-EXPANDED(收录号：WOS:000626085300001)、、Scopus(收录号：2-s2.0-85102187912)、WOS

基金：This work was supported by the National Natural Science Foundation of China, Grant Numbers 51709031 and 51979045; the Fundamental Research Funds for the Central Universities, Grant Number 3132019138; the Innovation Support Project of Dalian, Grant Number 2018RQ22; the Enterprise-university-research Cooperation Project of the Ministry of Education of China, Grant Number 201702043016, the Special projects in key fields (Artificial Intelligence) of Universities in Guangdong Province, Grant Number 2019KZDZX1035; the program for Scientific Research Start-up Funds of Guangdong Ocean University.

语种：英文

外文关键词：Oil spill; Shipborne radar; Active contour model; Multi-time-domain

外文摘要：Accidental oil spills cause serious pollution to the ocean and are difficult to control in short time. It is an important guarantee for emergency disposal to effectively monitor oil spills. Remote sensing is the main means to monitor oil spills. High false alarm rate has been an important bottleneck of this technology. In this paper, a multi-time-domain shipborne radar images discrimination mechanism was proposed. Based on the improved Sobel operator, Otsu and linear interpolation, the co-frequency interference noises were detected and suppressed. Gray intensity correction model (GICM) and dual-threshold method were used to eliminate highlighted continuous pixels. Oil films were extracted by using an active contour model (ACM). Finally, a multi-time-domain discrimination mechanism based on variation range tolerance of identified oil films centroids was designed to reduce the false alarm rate. It can provide technical support for decision-making and emergency response.