基于梯度修正和区域融合的分水岭分割算法

针对传统分水岭算法中存在的过分割现象，提出了一种基于梯度修正和层次区域融合的分水岭分割方法。该算法首先利用开闭双重建操作以及非线性变换对梯度图像进行修正；然后求取浮点活动图像并作为分水岭算法的输入；最后在区域灰度相似性准则的基础上，结合对比度和边界强度准则对分水岭变换结果进行小区域的合并，得到最终的分割结果。实验结果表明，该算法能有效地解决过分割问题，具有良好的鲁棒性和适应性。

Watershed segmentation algorithm based on gradient modification and region merging

In order to solve the problem of over-segmentation of the traditional watershed algorithm, a new algorithm of watershed segmentation method was proposed based on gradient modification and hierarchical region merging. Firstly, nonlinear transformation and morphological opening by reconstruction and closing by reconstruction were used to modify the gradient image; secondly, the floating point active image was computed as the input of watershed algorithm; finally, some small segmented regions were incorporated into the near bigger regions which was based on criterion of the average gray value of the region, and according to the contrast control and the edge strength criterion. The experimental results show that this method has good robustness and adaptability which can settle the over-segmentation problem effectively.