基于免疫聚类算法的MRI膝关节图像分割  

Segmentation of knee structure based on immune clustering in MRI image

作　　者：闫夏[1] 谭光兴[1] 林川[1] 

机构地区：[1]广西科技大学电气与信息工程学院,广西柳州545006

出　　处：《广西科技大学学报》2015年第1期70-74,共5页Journal of Guangxi University of Science and Technology

基　　金：广西自然科学基金项目(2010GXNSFA013024);广西重点实验室建设项目(13-051-38);2012年广西研究生教育创新计划项目(2012105940811 M01)资助

摘　　要：为了解决传统的阈值法无法适用的特征值相差不大的较复杂的医学图像分割,并针对聚类算法在分割图像时对初始值敏感且易陷入局部最优解这一缺点,提出了一种新的图像分割方式—基于免疫聚类的图像分割.实验表明,该方法既有效地克服了聚类算法的缺陷,又减轻了免疫算法在进化后期的波动现象,提高了其收敛速度和局部搜索能力,更具稳定性和准确性,在一定程度上改善了医学图像的分割效果.To solve the segmentation of small characteristic values that traditional threshold method cannot solve, and considering the clustering algorithm in image segmentation is sensitive to initial number and easy to fall into local optimal solutions, a new algorithm for segmentation of knee structure in MRI image, artificial immune clustering segmentation algorithm is provided in the article. Experiments show that this method can effectively overcome the defects of the clustering algorithm, reduce the wave phenomenon in the late of immune algorithm, and improve its convergence speed and local search ability, be more stable and accurate, and to a certain degree improve the image segmentation effect.

关 键 词：MRI膝关节 免疫 聚类 免疫聚类 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]