厦门大学学报（自然科学版）2012，Vol.51Issue(4)：701-703,3.
改进的区域生长算法及其在肾实质自动分割中的应用
An Improved Region Growing Algorithm and Its Applications in Kidney Segmentation
摘要
Abstract
An improved region growing algorithm for automatic segmentation kidney from abdominal CT image is proposed. Compared to original region growing method, this method automatically selected initial seed-pixels and is robust to the order of region growing. Firstly,computing the neighbor similarity factor CNSF) based on local histogram of each pixel and spatial information of local pixels. Then,building the criteria of initial seed-pixels,region growth and region growing termination based on NSF. Finally,MIC-CAI metrics are adopted to measure the segmentation accuracy. Experimental results demonstrated the performance of method.关键词
图像分割/直方图特征/区域生长/邻域相似性指标Key words
image segmentation/histogram features/region growing/neighbor similarity factor分类
医药卫生引用本文复制引用
高岩,王博亮..改进的区域生长算法及其在肾实质自动分割中的应用[J].厦门大学学报（自然科学版）,2012,51(4):701-703,3.基金项目
国家自然科学基金项目(30770561,61071151,61102137) （30770561,61071151,61102137）
