计算机与数字工程2017，Vol.45Issue(9)：1853-1858,6.DOI:10.3969/j.issn.1672-9722.2017.09.034
基于改进分水岭算法的高分遥感影像分割方法研究
Segmentation of High Resolution Remote Sensing Image Based on Improved Watershed Transformation
陈会斌 1郑永辉 1张磊2
作者信息
- 1. 解放军理工大学野战工程学院 南京 210007
- 2. 72253部队 济南 250022
- 折叠
摘要
Abstract
Image segmentation is based on the premise and key step of remote sensing image analysis,image processing is a classic problem. Watershed transform is a suitable tool for powerful morphological image segmentation,but its weak point lies in its over-segmentation results. Morphological reconstruction and labeling watershed segmentation scheme is proposed and it uses Sobel gradient to reconstruct the gradient image with the edge information,then uses morphological opening and closing reconstruction to filter gray image,and calculates the binary tag image,on this basis,then tag-based watershed transformation is carried out. Experi?mental results show that this method significantly improve segmentation accuracy,and better suppress over-segmentation.关键词
高分影像/图像分割/改进分水岭算法Key words
high resolution image/segmentation/improved watershed algorithm分类
信息技术与安全科学引用本文复制引用
陈会斌,郑永辉,张磊..基于改进分水岭算法的高分遥感影像分割方法研究[J].计算机与数字工程,2017,45(9):1853-1858,6.
