PDE降噪改进模糊聚类的图像局部分割

The Local Segmentation of Image by Fuzzy Clustering of PDE Noise Reduction to Improve

利用传统的降噪算法对脑MRI图像降噪时,会使脑MRI图像的纹理、边缘和血管等重要信息丢失.而偏微分方程(PDE)降噪算法能够在降低噪声的同时,可以有效缓解上述情形,确保细节的保留.模糊C均值(FCM)聚类算法和图像降噪算法在图像分割中有广泛的应用.本文将这两种方法结合起来,提出一种PDE降噪和模糊C均值聚类算法相结合的图像分割算法.首先利用改进的模糊C均值聚类算法(SKFCM)对图像进行初始分割;然后采用PDE降噪算法对感兴趣的部分进行精确分割,进而提取目标区域;最后通过仿真实验验证了此方法.结果表明该算法对噪声有很强的抑制能力,并且得到很好的分割效果。

Traditional medical image noise reduction algorithm when reduces noise,making some edge medical images,texture detail and some information of vessels is lost. The algorithm of the image noise reduction based on partial differential equations( PDE) can effectively overcome the shortcomings,while the noise is reduced,as much keep as possible the edge and information of the texture details of the image. Algorithm of fuzzy C- Means( FCM) Clustering and image noise reduction algorithm is widely used in image segmentation. In this paper,the two methods are combined,an image segmentation algorithm is proposed of the PDE noise reduction and the algorithm of the fuzzy C-Means Clustering combining. First,we segment original image by using the improved fuzzy c- means clustering algorithm( SKFCM); secondly,the interested part of the image is accurately segmented through the algorithm of PDE,and then extract the target area; finally,this method is verified by simulation experiments. The results show that the algorithm has strong restraint ability against noise,and get a good segmentation effect.