摘要
合成孔径雷达SAR(Synthetic Apenure Radar)图像海岸线检测在自动导航、地图绘制等海洋应用方面具有重要意义。国内外针对海岸线检测提出了不少方法,但是这些方法都有一个很大的缺点:速度慢。为此提出了一种快速SAR图像海岸线提取算法。该算法基于种子点增长的思想,从一个选定的初始种子点出发,计算出所有连通的海洋区域点,从而可以确定出海岸线位置。将该算法应用于真实的SAR图像进行实验,证明了该算法的有效性。
Coastline detection in Synthetic Aperture Radar （SAR） images plays a significant role in image applications, such as autonomous navigation, geographic mapping, etc. There are many methods have been proposed to solve this task, but they have a common defect, i.e. low speed. So, this paper proposes a quick coastline detection algorithm of SAR images. Based on the idea of speeds of growing, all connected sea pixels from one start seed pixel predetermined can be gotten, and the position of the coastline can be gotten. The feasibility of the algorithm has been proved by applying it to real SAR images.
出处
《电脑开发与应用》
2006年第6期2-3,6,共3页
Computer Development & Applications
基金
国家863计划资助项目(编号:2003AA131152)。
