| 基于改进标记的高分辨率遥感图像分水岭分割方法 |
| |
| 引用本文： | 尹高飞,肖鹏峰,冯学智. 基于改进标记的高分辨率遥感图像分水岭分割方法[J]. 遥感信息, 2010, 0(2): 12-17. DOI: 10.3969/j.issn.1000-3177.2010.02.003 |
| |
| 作者姓名： | 尹高飞 肖鹏峰 冯学智 |
| |
| 作者单位： | 南京大学地理信息科学系,南京,210093 |
| |
| 基金项目： | 国家高技术研究发展计划，国家自然科学基金，国家自然科学基金，教育部高等学校博士学科点专项科研基金 |
| |
| 摘 要： | 扩展最小变换是标记分水岭变换中常见的标记提取方法,但实验证明,基于该方法的图像分割结果在地物边缘处仍会存在破碎多边形。本文采用一种新的标记提取方法,首先利用高斯低通滤波器对梯度图像进行平滑,然后在梯度图像的低频成分中利用扩展最小变换提取标记,最后利用二值标记图像对原始梯度图像进行梯度重建,分水岭变换在重建后的梯度上进行。实验证明,该方法可以有效地去除地物边缘处的破碎多边形。
|
| 关 键 词： | 图像分割 标记提取 分水岭变换 低通滤波 高分辨率遥感图像 |
| 收稿时间： | 2009-04-07 |
Segmentation of High Resolution Remote Sensing Image Via an Improved Marker-Based Watershed Algorithm |
| |
| YIN Gao-fei,XIAO Peng-feng,FENG Xue-zhi. Segmentation of High Resolution Remote Sensing Image Via an Improved Marker-Based Watershed Algorithm[J]. Remote Sensing Information, 2010, 0(2): 12-17. DOI: 10.3969/j.issn.1000-3177.2010.02.003 |
| |
| Authors: | YIN Gao-fei XIAO Peng-feng FENG Xue-zhi |
| |
| Affiliation: | YIN Gao-fei,XIAO Peng-feng,FENG Xue-zhi(Department of Geographic Information Science,Nanjing University,Nanjing 210093) |
| |
| Abstract: | As a popular marker-extraction method in marker-based watershed transform the extended minima transform is proven that it cannot obtain an ideal result: some fragmentized parcels still scatter around the boundary between objects.A novel marker-extraction approach is employed in this paper to eliminate the fragments parcels.Firstly,smoothen the gradients with a Gaussian low-pass filter,and then extract the markers via the extended minima transform from the low frequency components of the gradients.Then recon... |
| |
| Keywords: | image segmentation marker extraction watershed transform low-pass filter high resolution remote sensing image |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
| 点击此处可从《遥感信息》浏览原始摘要信息 |
|
点击此处可从《遥感信息》下载全文 |
|
