一种改进的分水岭算法在医学图像分割中的应用研究
pplication of the improved watershed algorithm in medical image segmentation
分水岭算法在图像处理领域得到了广泛的应用，但单独使用分水岭进行图像分割会产生过分割现象，难以得到令人满意的结果。针对分水岭算法存在的过分割问题，提出一种改进的分水岭算法，能有效的抑制过分割现象。首先对对输入图像进行平滑滤波处理并计算该图像的多尺度灰度梯度图像；然后提出基于多阈值分割的方法消除无效的梯度信息；再进行分水岭变换并对相似的区域进行合并。实验结果表明，本文提出的改进的分水岭算法能有效的处理过分割问题，比传统的分水岭算法得到的分割效果要好很多。
he algorithm of watershed has been widely applied in the field of image processing, it will appear over-segmentation phenomenon when we use the traditional watershed only, and it is difficult to get a satisfying result. To address this issue, proposed an improved algorithm of watershed in this paper, it can effectively restrain the phenomenon. Smooth filter processing and computing the multi-scale image intensity gradient image of the input image first. And then implement of the watershed transform and merge the similar regions. Through experiments, the proposed algorithm in this paper can effectively inhibit the over-segmentation than the traditional watershed algorithm.
刘林、黄英、贺振华
医学研究方法生物科学理论、生物科学方法
分水岭算法图像分割大脑MR图多阈值分割区域合并平滑滤波
Watershed algorithmImage segmentationMR image of brainMulti-threshold segmentationRegion mergingSmooth filter
刘林,黄英,贺振华.一种改进的分水岭算法在医学图像分割中的应用研究[EB/OL].(2013-01-22)[2025-07-22].http://www.paper.edu.cn/releasepaper/content/201301-958.点此复制
评论