一种改进的B-Snake肝癌图像分割算法  

An Improved B-spline Snake Model for Liver Cancer Image Segmentation

作　　者：史延新[1] 

机构地区：[1]西安工业大学数理系,陕西西安710032

出　　处：《太原理工大学学报》2011年第1期30-33,共4页Journal of Taiyuan University of Technology

基　　金：国家自然科学基金(60634030);高等学校博士学校点专项科研基金资助项目(20060699032)

摘　　要：针对肝癌CT图像特点,提出了一种改进的B样条曲线的Snake模型图像分割算法,该算法首先对腹部CT图像进行预处理,获得肝脏癌变部分的初始轮廓,再构造闭合B样条Snake模型,最后使用MMSE最小化外力变形模型以实现图像的准确分割。实验结果表明,该方法对于肝癌CT图像取得了肝脏癌变部分目标的良好分割效果。An improved B-spline Snake model for image segmentation was proposed,in view of the features of liver cancer CT images. Firstly the abdominal CT imge was preprocessed to obtain liver cancer part of the initial outline,then closed B-spline Snake model was constructed. Finally, deformed model was used for accurate segmentation of image using the MMSE to minisize external force. Exerimental results show that this algorithm was an effective method for segmenting the interested target of liver CT images.

关 键 词：B样条曲线 SNAKE模型 MMSE 图像分割 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]