基于改进水平集的脑肿瘤图像分割

Brain tumor image segmentation based on improved level set

针对CV模型分割方法在处理脑肿瘤图像中存在的分割精度和稳定性差的问题,提出了一种基于改进CV模型的水平集分割方法。该算法保留了CV模型处理弱边缘图像的优点,同时加入了图像梯度惩罚项,使其能准确分割灰度不均匀图像。实验结果表明,文中模型相比于CV模型和LBF模型,能更准确地分割MR脑肿瘤图像。

A horizontal set segmentation method based on improved CV model is proposed to solve the problem of the difference of segmentation accuracy and stability in the image of brain tumor.The algorithm retains the advantages of the CV model to deal with weak edge image,and the image gradient penalty is added to enable it to accurately split the gray image.The experimental results show that this model can be more accurate than the CV model and the LBF model for the segmentation of MR brain tumor images.