图像序列处理中的形态分割方法  

Image Sequence Processing Using 3-D Morphological Segmentation

作　　者：钟伟[1] 余松煜[1] 芮雨[1] 

机构地区：[1]上海交通大学图像通信与信息处理研究所,上海200030

出　　处：《上海交通大学学报》2001年第9期1314-1316,1320,共4页Journal of Shanghai Jiaotong University

摘　　要：将数学形态学应用到序列图像分析中 ,提出了一种基于图像序列的 3- D形态分割算法 ,对图像序列简化后进行边缘增强 ,用改进的 3- D分水岭方法提取分割信息 .基于分割信息对几种常用图像序列的处理算子和处理方法进行了研究和改进 .实验结果表明 ,该方法能对序列图像有效地进行时空分割 ,利用分割信息可显著提高各种图像序列处理的性能 ,而且具有算法简单、效率高的特点 .A mathematical morphology method was applied in image sequence analysis. An image sequence segmentation method based on 3-D morphology was presented and the frequently used image sequence processing techniques were improved based on the segmentation information. This algorithm first simplifies the image sequence, and then enhances the edge. The segmentation information is extracted using 3-D water-shed algorithm. Several frequently used image sequence processing operators are studied. Test results demonstrate that the method can effectively segment image sequence and the segmentation information can greatly improve the performance of the image sequence processing. This algorithm is simple and efficient.

关 键 词：形态学 图像序列分割 图像序列处理 形态分割 3-D分水岭方法 图像分析 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]