基于多尺度多结构元素的肝脏图像分割

提出了一种基于多尺度形态滤波的多结构元素的肝脏CT图像分割方法.首先,利用肝脏先验知识预测肝脏位置,执行多阈值法简化原图.然后,通过递归多尺度形态滤波和区域标记得到肝脏初始轮廓.最后,构造5个不同方向的结构元素并结合它们的多尺度形态学检测肝脏边缘.为验证本算法的正确性和通用性,选用了不同形状的肝脏图像进行实验.结果表明本文算法能够成功地完成肝脏分割.

Liver image segmentation based on multi-scale and multi-structure elements

A new approach to segment the liver of abdominal CT image is proposed based on multi-scale morphological filtering and multi-structure elements.In the preprocessing,the input CT image is simplified by estimating the liver position using prior knowledge about the location of the liver.Then multi-level threshold is performed on the estimated liver position to get the simplified image.The proposed scheme utilizes recursively multi-scale morphological filtering with region-labeling to generate the initial liver...