作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

›› 2004, Vol. 30 ›› Issue (15): 46-47.

• 论文 • 上一篇    下一篇

克服Watershed算法过度分割的方法

王鹏,单保慈,谢千河,曾振柄   

  1. 中国科学院成都计算机应用研究所,中国科学院北京高能物理研究所,中国电子科技大学,中国科学院成都计算机应用研究所 成都 610041
  • 发布日期:2004-08-15

Method of Overcoming the Oversegmentation by Wavelet Transform

WANG Peng, SHAN Baoci, XIE Qianhe, ZENG Zhenbing   

  • Published:2004-08-15

摘要: Water-shed算法是一种形态学的图像分割算法,但由于其对噪声十分敏感,分割结果往往存在过度分割的现象。该文提出一种基于小波分析的Watershed算法,充分利用小波的多分辨率特性有效地解决了Watershed算法的过度分割问题,并大大提高了分割计算速度。

关键词: 小波变换, 分水岭算法, 过度分割, 多分辨率

Abstract: Watershed algorithm is a kind of image segmentation algorithm based on mathematical morphology. The high sensitivity of the watershed algorithm to noise yields a very large number of catchment basins, lead to oversegmentation. A watershed algorithm based on wavelet analysis is proposed to overcome the oversegmentation. It uses the multiresolution of wavelet transform to overcome the oversegmentation and improve the compute speed.

中图分类号: