图像分割在医学超声图像的定量、定性分析中均扮演着十分重要的作用,并直接影响到后续的分析、处理工作。针对医学超声图像对比度低和噪声强的特点,提出了一种将超像素和模糊聚类技术相结合的图像分割方法。该方法利用简单线性迭代聚类算法产生多个超像素子区域,通过比较各个子区域间特征向量的相似性,利用模糊C均值(FCM)聚类技术对这些过分割区域进行合并,实现超声图像目标区域的有效分割。和传统的基于单像素的FCM聚类算法相比,该方法具有较强的鲁棒性,有效提高了目标区域的分割精度和分割效率,取得了较好的分割效果。
Image segmentation plays a very important role in medical ultrasound image analysis.In this paper,a novel image segmentation method based on FCM combining with superpixel was proposed to apply in noisy ultrasound images with low contrast.The method uses the algorithm SLIC to generate numerous superpixels firstly,then use FCM algorithm to merge superpixels by comparing their similarities to achieve the goal of object segmentation correctly.Compared with conventional pixel-based FCM algorithm,the proposed method is robust to noise and can improve the segmentation accuracy and efficiency obviously.