三种遥感反演海岸线方法的实验对比分析  

Experimental Contrast Analysis of Three Methods for Retrieving Coastline by Remote Sensing

作　　者：马立婷[1] 陶秋香[1] 张庆云[1] 

机构地区：[1]山东科技大学测绘科学与工程学院,山东青岛266590

出　　处：《海洋测绘》2013年第6期66-69,共4页Hydrographic Surveying and Charting

基　　金：山东省自然科学基金项目(ZR2012DM001);山东科技大学科研创新团队支持计划(2011KYTD103);海岛(礁)测绘技术国家测绘地理信息局重点实验室项目(2011B04)

摘　　要：在分析海岸线提取算法基础上,以环渤海的Landsat7 ETM+影像为数据源,以基岩海岸线和淤泥质海岸线两种典型的海岸线为研究对象,采用Sobel边缘检测算法、阈值分割法和K最邻近分类算法进行海岸线提取实验。通过实验对比分析,归纳和总结了针对不同类型海岸线提取算法的适用性。Based on the analysis of the coastline extraction algorithm, the paper takes the Landsat7 ETM＋ image of Bohai Sea Riving Area as data source and studies such two typical coastlines, bedrock coastline and muddy coastline. Then carries on the coastline extraction experiment by using Sobel edge detection algorithm, threshold segmentation method and K-Nearest Neighbor. According to the experimental analysis, the applicability of the extraction algorithm for different types of coastline has been summarized.

关 键 词：海岸线提取 Sobel边缘检测算法 阈值分割法 K最邻近分类算法 

分 类 号：P237[天文地球—摄影测量与遥感]