快速分割SAR影像海岸线的多尺度水平集方法

摘要

针对C-V模型对SAR影像海岸线分割速度慢的问题,以高分辨率珠海南岸ENVISAT ASAR影像为研究样本,提出了一种快速、高精度的多尺度C-V模型(multi-scale-C-V,M-C-V)水平集分割海岸线方法.该方法在传统C-V模型中加入多尺度技术缩小影像尺度获得不同空间分辨率下的影像序列,利用低通滤波磨光小尺度影像序列以形成边界相对光滑的影像序列,在具有不同尺度和磨光程度的影像序列中逐一进行基于C-V模型的水平集海岸线分割.在海岸线分割中,高空间分辨率影像继承上一级低空间分辨率影像中提取的边界并通过C-V模型进一步细化海岸线.实验表明,该方法加速形成初始水平集,缩短了提取海岸线的时间,同时较好地剔除零散的非主体海岸线部分,提高了主体海岸线的识别精度,使提取海岸线过程具有鲁棒性.%To improve the speed of coastline segmentation,a fast multi-scale C-V model(multi-scale-C-V,M-C-V)level set method with high accuracy is proposed.In this paper,the high resolution ENVISAT ASAR images of Zhuhai South Bank are used as the research sample.The method includes three distinct aspects:it adds multi-scale technology to the traditional C-V model,which reduces the image size and obtains image sequence under different spatial resolution;low-pass filter polishes small scale image sequence,which is easy to form the image sequence with the relatively smooth boundary;image sequences in the different scales polished scales,which can make the coastline based on the C-V model of level set easily extracted one by one.In the division of coastline,high spatial resolution images inherit boundary from the small scale images to the bigger scale images,which will further refine coastline through the C-V model.Experiments show that the method accelerates the formation of the initial level set,and shortens the time of the extraction of coastline.At the same time,this method removes the non-coastline body part,and improves the identification precision of the main body coastline,which makes the process of coastline extraction robust.