基于动态合并准则的分水岭算法在医学图像处理中的应用

Application of the dynamic combination rule-based watershed algorithm in medical image processing

针对医学图像的模糊特点和实际应用的要求,提出了一种基于动态合并准则的分水岭分割方法。阐述了在分水岭算法的基础上做的一些改进,通过动态合并准则进行边分割边合并,有效地抑制了过分割现象。结果表明,该方法能够快速、准确地得到医学图像的分割结果,并且具有较强的抗噪声能力。

In view of the fuzziness of medical image and the requirement in practical application, a watershed segmentation method was proposed based on the dynamic combination rule. It restrained the over-segmentation occurrence effectively by combining while segmenting in accordance with the dynamic combination rule. Experiments show that the proposed algorithm can segment medical images quickly and accurately. Furthermore, it is robust under noisy background.