摘要
针对海岸线的影像特征,提出一种基于边缘引导约束的海岸线提取方法。该方法是在遥感影像进行中值滤波处理,去除噪声的基础上,通过交互方式对遥感影像进行灰度拉伸,以增强海岸线特征;再利用梯度算子卷积运算得到梯度影像;然后通过边缘像素点和灰度差阈值引导种子点进行生长,进而提取包括滩涂在内的水域或其他地物;最后通过轮廓跟踪提取生长区域的边界,即海岸线。试验表明,该方法可准确提取出海岸线位置,效果较好。
Aimed at image features of coastlines, a coastline extraction method based on guidance and constraint of edge was presented.On the basis of removing the remote sensing image noises by means of median filtering, the remote sensing image was processed with gray stretching though interacting modes to enhance the coastline feature;The gradient image was obtained by gradient operator convolution operation.Then edge pixel and gray threshold guided the growth of seed points, extracting waters including tidal flat or other surface features.Finally, the boundary of the growing area extracted by means of contour tracking was the coastline. Experiments show that the proposed method could extract coastlines correctly and the results are good.
出处
《海洋测绘》
CSCD
2015年第6期58-61,共4页
Hydrographic Surveying and Charting
关键词
高分辨率遥感影像
边缘引导约束
区域生长
边缘像素点
海岸线提取
high-resolution remote sensing image
edge guidance and constraint
region growing
edge pixel
contour tracking
coastline extraction