FGMM_MRF层次模型在图像分割中应用.pdfVIP

第 17 卷 第 12 期 计算机辅助设计与图形学学报 Vol 17 , No12 2005 年 12 月 JOURNAL OF COMPU TERAIDED DESIGN COMPU TER GRAPHICS Dec , 2005 　 FGMMMRF 层次模型在图像分割中的应用 李旭超 　　　朱善安 (浙江大学电气工程学院　杭州 　310027) (lixuchaoliaoning @sinacom) 摘 要 　为了更好地反映图像的区域结构 ,在高层次标记图像中 ,区域内部用各向同性的 MRF 建模 ,区域的边界用 各向异性的 MRF 来建模 ;在低层次灰度图像中 ,用 FGMM 来描述待分割图像的概率分布 采用 Ba yes 方法 ,根据标 记图像的后验分布所对应的 FGMMMRF 模型的条件概率 ,用 ICM 局部优化算法获得 MAP 准则下的分割图像 用 模拟图像和 MR 图像进行实验 ,区域的边界和整体属性具有较好的视觉效果 关键词 　有限通用混合模型马尔可夫随机场层次模型 ;迭代条件模式优化算法 ;MAP 准则 ; 图像分割 中图法分类号 　TP391 Application of FGMMMRF Hierarchical Model to Image Segmentation Li Xuchao 　Zhu Shan ’an ( College of Elect rical Engi neeri ng , Zhejian g U niversity , Hangz hou 　310027) Abstract 　　In order to accurately describe the region structure of a higherlevel label image , the interior region is modeled by isotropic Markov random field (MRF) , while the boundary is modeled by anisotropic MRF For lowerlevel gray image , the prior distribution of segmentation image is modeled by finite general mixture model ( F GMM) According to the posterior distribution of the label image conditioned on the gray image corresponding to the conditional probabilit y of F GMMMRF model , the Bayes formulation and the lo cal iterated conditional modes ( ICM) optimization algorithm are adopted , and based on the MAP (maximum a posterior) criterion the image segmentation result is obtained Numerical simulations demonstrate that the whole propert y and the boundary of image area show better vision effect with a test to synthetic image and real MR brain image Key words 　F GMMMRF hierarchical model ; ICM optimization algorithm ; MAP criterion ; image segmen tation