基于Chan-Vese模型与形态学的医学图像分割算法

Medical Image Segmentation Algorithm Based on Chan-Vese Model and Morphology

摘要:

针对传统的多相Chan-Vese模型在进行多区域分割时容易产生空相位的问题,提出了一种改进的新的医学图像分割算法;该算法结合Chan-Vese模型、数学形态学、复合多相水平集分割算法,通过迭代腐蚀操作提取医学图像的轮廓,利用添加了复合多相水平集算法的Chan-Vese模型对医学图像进行分割,通过迭代膨胀操作复原图像;实验结果和分析表明,采用该算法很好地解决了医学图像分割过程中容易出现的多区域分割问题,减少了空相位的产生,而且对图像边缘有很好的分割效果。

Abstract:

The traditional model of multiphase Chan-Vese is easy to cause empty phase problem during the multi-region segmentation,To solve this problem,it proposes a improved Chan-Vese model of medical image segmentation algorithm combined with mathematical morphology in this paper.The algorithm integrates Chan-Vese model, mathematical morphology, complex multiphase level sets segmentation algorithm, first, through iterative etching operation to extract the outline of the medical image, and then the medical image is segmented by the Chan-Vese model based on the complex multiphase level sets, finally the medical image is dilated iteratively by using morphological dilation to restore the image.The experimental results and analysis show that this algorithm solves the multi-regional medical image segmentation problem well, and reduces the empty phase, and has a very good segmentation effect on the edge of the image.