基于改进Canny算子的双水平集医学图像分割

医学图像常常伴有噪声以及灰度不均匀等问题,且图像边缘具有复杂性,因此传统的双水平集模型无法达到理想的分割效果。针对该问题提出了一种基于改进Canny算子的双水平集医学图像分割算法,首先在Canny算子的基础上加入梯度模板,用来解决传统Canny算子无法完全检测出复杂图像边缘的问题,然后将检测后的图像融入到双水平集模型中,通过在水平集模型中加入偏移场来解决灰度不均匀的问题,并且在距离规则化项中使用双肼势函数。实验结果证明,该模型对于复杂边界以及灰度不均匀的图像有较好的分割效果。

Bilevel Set Medical Image Segmentation Based on Improved Canny Operator

Medical image is often accompanied by noise and uneven gray level,and the edge of the image is complex,which are cause difficulties of medical image segmentation.Therefore,a two-level set medical image segmentation algorithm based on improved Canny operator is proposed.Firstly,gradient template is added on the basis of Canny operator,and then the detected image is integrated into the double level set model.By adding migration field to the level set model,the problem of uneven gray level is solved,and the dihydrazine potential function is used in the distance regularization term.Experimental results show that the model in this paper has a good segmentation effect for the image with complex boundary and uneven gray level.