基于Chan-Vese模型的海面油膜分割应用  

Oil Film Segmentation Based on Chan-Vese Model

作　　者：靳卫卫 于新生[2] 裴建新[2] 栗宝鹃 邱照宇 JIN Wei-wei;YU Xin-sheng;PEI Jian-xin;LI Bao-juan;QIU Zhao-yu(CNOOC Offshore Environmental Services Co.Ltd.,Tianjin 300457,China;College of Marine Geosciences,Ocean University of China,Qingdao Shandong 266061,China)

机构地区：[1]中海石油环保服务(天津)有限公司,天津300457 [2]中国海洋大学海洋地球科学学院,山东青岛266061

出　　处：《船海工程》2018年第2期77-80,共4页Ship & Ocean Engineering

基　　金：中国海洋石油集团有限公司科研项目(RLKJ-2014-01)

摘　　要：考虑到采用高空间辨率的光学图像开展海面油膜的监测由于海面风、浪、流等环境因素的影响,采集到的视频图像普遍存在成像光照不均匀、低对比度、低信噪比等问题,采用自适应高斯滤波进行图像噪声去除,基于一种局部图像对比度增强算法进行去雾和对比对增强处理,通过颜色空间转换配合Chan-Vese模型进行图像分割,使得海面油膜图像处理取得了良好效果。The monitoring of surface oil film by optical image with high spatial resolution is an effective method,but,due to the influence of sea surface wind,wave,flow and other environmental factors,the collected images are generally uneven illumination,low contrast and low signal-to-noise ratio,this makes the image segmentation difficult.The image noise can be removed through a Gaussian filter,and a local image contrast enhancement algorithm was used to enhance image contrast.The image segmentation was performed through the color space conversion and the Chan-Vese model,and the good results were obtained through the oil film image processing.

关 键 词：油膜分割 CHAN-VESE模型 高斯滤波 局部对比度增强 

分 类 号：U698.7[交通运输工程—港口、海岸及近海工程]