基于分水岭区域差异性的多级图像分割

被引:15
作者
赵晨光
宋利伟
庄天戈
机构
[1] 上海交通大学生物医学工程系
关键词
图像分割; 分水岭区域; 数学形态学;
D O I
10.16183/j.cnki.jsjtu.2004.09.014
中图分类号
TP391.41 [];
学科分类号
080203 ;
摘要
分水岭分割算法的不足之处在于过分割,即生成大量的小区域而使目标物淹没其中.文中通过定义分水岭区域的差异性给出一种多级图像分割方案,克服了传统的解决过分割问题时需要预先设置标记点的缺陷,也避免了多分辨率分割中边缘模糊和移位的问题.该算法可以应用于多种类型的图像,并经实验证明具有良好的分割效果.
引用
收藏
页码:1489 / 1491
页数:3
相关论文
共 7 条
  • [1] Robust anisotropic diffusion. Black M J,Sapiro G,Marimont D H,et al. IEEE Transactions on Image Processing . 1998
  • [2] An overview of morphological segmentation. Meyer F. International Journal of Pattern Recognition and Artificial Intelligence . 2001
  • [3] Minimal set of markers for the watershed transform. Lotufo R,Silva W. International Symposium on Mathematical Morphology VI -ISMM 2002 . 2002
  • [4] Image segmentation and analysis via multiscale gradient watershed hierarchies. Gauch J M. IEEE Transactions of Image Processing . 1999
  • [5] Geodesic saliency of watershed contours and hierarchical segmentation. Najman L,Schmitt M. IEEE Transactions on Pattern Analysis and Machine Intelligence . 1998
  • [6] Watersheds in digital spaces: An efficient algorithm based on immersion simulation. Vincent L,Soille P. IEEE Transactions on Pattern Analysis and Machine Intelligence . 1991
  • [7] Watershed-based segmentation and region merging. Bleau A,Leon L J. Computer Vision and Image Understanding . 2000