融合边缘流和归一化割的岩心图像分割.pdfVIP

光 电 子 ·激 光 第 2 1 卷 第 3 期 　20 10 年 3 月 　　　　　　 Vol . 2 1 No . 3 　Mar . 20 10 Journal of Optoelectronics ·Laser 　　　　　　　　 ① 融合边缘流和归一化割的岩心图像分割 吴晓红 , 罗代升 ( 四川大学电子信息学院,四川 成都 610064) 摘要 :提出一种融合多尺度边缘流和归一化割的图像分割方法。首先采用多尺度边缘流 ,检测在大尺度和小尺 度上均存在的边缘 ,并在小尺度上定位边缘 ,再通过边缘连接 ,得到封闭的边缘;然后把多尺度边缘流分割得到 的区域作为归一化割中图的顶点,区域间的相似性度量作为相似性矩阵的元素 ,大大降低了归一化割对内存的 需求 ,提高了特征方程的计算效率 ,使归一化割得到实用 ,同时也降低了边缘流的过分割现象。实验表明,在岩 心扫描图像的分割中,该方法能得到很好的效果。 关键词 :融合; 岩心图像; 多尺度边缘流; 归一化割; 过分割 ( ) 中图分类号: TP391. 41 　　文献标识码 :A 　　文章编号 2010 The core image segmentation based on integration of multiscale edgeflow and normalized cut WU Xiaohong , LUO Daisheng (School of Electronics Information Engineering ,Sichuan University ,Chengdu Sichuan 610064 ,China) Abstract :An image segmentation method which combines Multiscale Edgeflow and Ncut (Normalized cut) was put forward. Fir st ,Multiscale Edgeflow was used to detect the edges existing in both large scale and small scale ,edge position was made in smallscale ,then a closed edge was acquired through a marginal connection . Second ,the region acquired by Multiscale Edgeflow was regarded as the vertices of the graph in Ncut ;the similarity criterion among regions was regarded as element s of similarity matrix . Thus ,the memory requirement for Ncut was greatl