基于控制标记符分水岭的医学图像分割

提出了一种适合医学图像分割的改进分水岭算法，针对分水岭算法存在的过分割问题，采用数学形态学重构滤波器和Ostu阈值分割方法对图像中感兴趣的目标和背景进行标记。根据标记的二值图像，运用形态学极小值标定技术对原有梯度图像进行修正。最后，使用分水岭算法对修正的梯度图像进行分割。该方法能有效抑制过分割现象，而且计算复杂程度较低、分割效果较好。

Medical Image Segmentation Based on Marker-Controlled Watershed

A modified watershed algorithm was proposed for medical image segmentation, with regard to the drawback of over-segmentation based on traditional watershed algorithm, a new methodology incorporated morphological reconstruction filters and Ostu threshold segmentation was used to obtain the foreground marker image and background marker image.Then the makers are imposed on the original gradients as its minima.Finally, the watershed algorithm is applied to the modified gradients to segment the image.Compared with other methods, the proposed algorithm can more efficiently reduce the over-segmentation and require less computation.