Segmentation of Ultrasound Image Based on Texture Feature and Graph Cut

Abstract

This image partition plays an important role in both qualitative and quantitative analysis of medical ultrasound images. But medical ultrasound images have features of poor contrast and strong speckle noise and segmenting result may not be satisfactory with traditional image segmentation method. Medical ultrasound images are segmented using image segmentation method based on texture feature and graph cut in this paper. The texture feature parameters are obtained according to gray level co-occurrence matrix. The similarities matrix is made based on texture feature parameters and gray intensity of pixel. We use the spectral graph theoretic framework of Normalized cuts to find partitions of an image based on the similarities matrix. Experimental results show that the method is an effective segmentation method for medical ultrasonic image.