| 基于修正交叉视觉皮质模型的图像分割方法 |
| |
| 引用本文： | 牛建伟,沈思思,童超,高小鹏,汪孔桥. 基于修正交叉视觉皮质模型的图像分割方法[J]. 北京邮电大学学报, 2010, 33(1): 56-60. DOI: 10.3969/j.issn.1007-5321.2010.01.012 |
| |
| 作者姓名： | 牛建伟 沈思思 童超 高小鹏 汪孔桥 |
| |
| 作者单位： | 北京航空航天大学,计算机学院,北京,100191;诺基亚中国研究院,北京,100176 |
| |
| 基金项目： | 国家自然科学基金项目(60702031,60873241);;国家高技术研究发展计划项目(2008AA01Z217,2007AA01Z145,2007AA01A127) |
| |
| 摘 要： | 提出了一种基于修正交叉视觉皮质模型(MICM)的图像自适应分割新方法. 根据待分割图像的自身特性，自适应地设定参数，并以互信息量为目标函数选取最佳分割结果. 该方法解决了针对不同的图像需要人工设定交叉皮质模型(ICM)参数和需要人工选取最佳分割结果的2个问题. 实验结果表明，与通过大量实验获得模型参数的脉冲耦合神经网络(PCNN)基本模型和ICM基本模型相比，MICM与其综合评价函数值相近；与模糊聚类分割算法和最大类间方差(OTSU)算法相比，MICM算法有较明显的视觉优势，并且其综合评价函数值也分别提高了约15%和13%.
|
| 关 键 词： | 图像分割 交叉视觉皮质模型 自适应 互信息量 |
| 收稿时间： | 2009-05-12 |
| 修稿时间： | 2009-08-08 |
A New Image Segmentation Method Based on Modified Intersecting Cortical Model |
| |
| NIU Jian-wei,SHEN Sisi,TONG Chao,GAO Xiao-peng,WANG Kong-qiao. A New Image Segmentation Method Based on Modified Intersecting Cortical Model[J]. Journal of Beijing University of Posts and Telecommunications, 2010, 33(1): 56-60. DOI: 10.3969/j.issn.1007-5321.2010.01.012 |
| |
| Authors: | NIU Jian-wei SHEN Sisi TONG Chao GAO Xiao-peng WANG Kong-qiao |
| |
| Affiliation: | (1.School of Computer Science and Engineering, Beihang University, Beijing 100191, China; 2.Nokia China Research Center, Beijing 100176, China) |
| |
| Abstract: | An image segmentation method based on the modified Intersecting Cortical Model (SICM) was proposed to be able to set the SICM parameters adaptively according to different characteristics of images and choose the optimal segmentation results automatically, which are two main obstacles for Intersecting Cortical Model to be used in practice. The experimental results show that SICM has visually better segmentation, and the comprehensive evaluation value of SICM increases by approximately 15 percent and 13 percent respectively compared with those of the fuzzy C-means algorithm and OSTU algorithm. |
| |
| Keywords: | image segmentation intersecting cortical model (ICM) self-adaptive mutual information |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
| 点击此处可从《北京邮电大学学报》浏览原始摘要信息 |
|
点击此处可从《北京邮电大学学报》下载全文 |
|
