摘要
针对超声心动图像噪声大,边界模糊,灰度分布不均匀等特点,提出了一种新的图像分割算法。首先对超声图像进行滤波和边缘锐化等预处理,然后再使用GVFSnake模型对预处理后的图像进行分割,从而提取出心室的边界。经试验证明,这种新的超声心动图像分割算法,比传统的超声图像分割算法在图像分割精度上有显著的提高。
A novel image segmentation algorithm is presented because the ultrasound heart image has large noise, blur edge and non-uniform gray distribution. The main idea is firstly to pre-treat the ultrasound image by filtering and edge sharpening and then to segment the image with GVFSnake model so that the ventricle boundary can be obtained. The experiments show that the new segmentation algorithm of the ultrasound heart image possesses much better precision than the traditional one.
出处
《成都信息工程学院学报》
2007年第3期297-300,共4页
Journal of Chengdu University of Information Technology
