| 基于改进标记分水岭的高分辨率遥感影像海岸水边线提取方法 |
| |
| 引用本文： | 栾奎峰,刘帅,潘与佳,朱卫东,李丕学,裘诚,邱振戈,沈蔚,王洁,王振华.基于改进标记分水岭的高分辨率遥感影像海岸水边线提取方法[J].海洋学研究,2021,39(1):20-28. |
| |
| 作者姓名： | 栾奎峰 刘帅 潘与佳 朱卫东 李丕学 裘诚 邱振戈 沈蔚 王洁 王振华 |
| |
| 作者单位： | 1.上海海洋大学 海洋科学学院,上海 201306; 2.上海市海洋监测预报中心,上海 200062; 3.上海河口海洋测绘工程技术研究中心,上海 201306 |
| |
| 基金项目： | 国家重点研发计划(2016YFC1400904);上海市海洋局科研项目(沪海科2019-5,沪海科2020-5);上海市地方能力建设项目(19050502100) |
| |
| 摘 要： | 利用遥感技术快速提取海岸线是一种重要的技术手段,针对传统分水岭算法在高分辨率多光谱卫星数据处理中存在的过分割和抗干扰能力差的问题,本文提出了一种基于扩展极值变换标记分水岭的算法。首先通过形态学重建、扩展极值变换等方法建立前景和背景标记,初步抑制灰度极小值和极大值区域,然后依据这两类标记对梯度图像进行修正,进而进行分水岭变换,提取岛屿水边线。以南海典型海岛为研究区域,利用2017年GF-2卫星数据对本文提出的方法进行验证和精度评价,研究结果表明:改进的分水岭算法对GF-2数据的人工岸线提取质量在1个像素(4 m)之内高于90%,沙质岸线提取质量在1.5个像素(6 m)之内高于90%,可以用于高分辨率多光谱影像的分割和海岛水边线的提取。
|
| 关 键 词： | 岛屿水边线 GF-2 扩展极值变换 标记分水岭 |
| 收稿时间： | 2020-03-04 |
Research on shoreline extraction for high-resolution remote sensing image based on improved marked watershed algorithm |
| |
| LUAN Kuifeng,LIU Shuai,PAN Yujia,ZHU Weidong,LI Pixue,QIU Cheng,QIU Zhen'ge,SHEN Wei,WANG Jie,WANG Zhenhua.Research on shoreline extraction for high-resolution remote sensing image based on improved marked watershed algorithm[J].Journal of Marine Sciences,2021,39(1):20-28. |
| |
| Authors: | LUAN Kuifeng LIU Shuai PAN Yujia ZHU Weidong LI Pixue QIU Cheng QIU Zhen'ge SHEN Wei WANG Jie WANG Zhenhua |
| |
| Institution: | 1. College of Marine Sciences, Shanghai Ocean University, Shanghai 201306, China; 2. Marine Monitoring and Forecasting Center of Shanghai, Shanghai 200062, China; 3. Shanghai Engineering Research Center of Estuarine and Oceanographic Mapping,Shanghai 201306, China |
| |
| Abstract: | The use of remote sensing technology to quickly extract the shorelines of islands is an important technology. An approach of marked watershed based on extended extremum transform was proposed in this study, solving the problems of over-segmentation and poor anti-interference ability of traditional watershed algorithm in high-resolution multispectral satellite data processing. Firstly, the foreground and background markers were established by morphological reconstruction and extended extremum transform to suppress the minimum and maximum gray regions initially, then the gradient image was modified based on above markers, and finally the island shoreline was extracted based on the improved watershed transform method. The typical islands located in the South China Sea region were selected, and the accuracy and reliability of the proposed method were verified using 2017 GF-2 satellite data. The results show that the accuracy of artificial shorelines is higher than 90% within 1 pixel(4 m) for island shorelines, and the accuracy of sandy shorelines is higher than 90% within 1.5 pixels(6 m), confirmed that the proposed method can be used for segmentation of high-resolution multispectral images and extraction of island shoreline. |
| |
| Keywords: | island shoreline GF-2 extended extremum transform marked watershed |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
| 点击此处可从《海洋学研究》浏览原始摘要信息 |
| 点击此处可从《海洋学研究》下载免费的PDF全文 |
|
