基于改进遗传模糊聚类和水平集的图像分割算法
|
摘要
针对传统的模糊聚类算法(FCM)容易陷入局部最优,水平集方法 (Level set)容易受初始边界和控制参数的影响等问题,引入具有全局搜索能力的遗传算法(GA)初始化聚类中心,提出改进的模糊聚类算法分割得到目标的粗边缘,利用水平集方法强大的演化能力收敛到目标边缘。该算法可以减少水平集方法控制参数的个数,降低计算的复杂度,提高分割速度。实验在多目标轮廓图像、轮廓不清晰图像上进行,实验结果表明,该方法能够很好地检测出多目标及弱边缘图像的轮廓,在乳腺X线图像中,肿块的分割精度、过分割率和欠分割率分别为98.35%,0.27%和1.12%,优于同类算法。
The traditional fuzzy clustering algorithm(FCM)is easy to fall into the local optimum,the level set method is easy to be affected by the initial boundary and control parameters and so on.To solve the problems,the genetic algorithm(GA)with global search ability was introduced to initialize the cluster center,and an improved fuzzy clustering algorithm was proposed to get the rough edge of the target.The strong evolutionary ability of the level set method was used to get the target edge.The algorithm can reduce the number of control parameters of the level set method,reduce the computational complexity,and improve the segmentation speed,effectively.Experiments were performed on multi-target contour images and unconspicuous contour images.The results show that the proposed method can detect the contour of multi target and weak edge image.The segmentation accuracy,over segmentation rate and under segmentation rate of lump in the mammography are 98.35%,0.27%,and 1.12% respectively,which are better than those of the similar algorithms.
The traditional fuzzy clustering algorithm(FCM)is easy to fall into the local optimum,the level set method is easy to be affected by the initial boundary and control parameters and so on.To solve the problems,the genetic algorithm(GA)with global search ability was introduced to initialize the cluster center,and an improved fuzzy clustering algorithm was proposed to get the rough edge of the target.The strong evolutionary ability of the level set method was used to get the target edge.The algorithm can reduce the number of control parameters of the level set method,reduce the computational complexity,and improve the segmentation speed,effectively.Experiments were performed on multi-target contour images and unconspicuous contour images.The results show that the proposed method can detect the contour of multi target and weak edge image.The segmentation accuracy,over segmentation rate and under segmentation rate of lump in the mammography are 98.35%,0.27%,and 1.12% respectively,which are better than those of the similar algorithms.
引文
[1]ZHOU Lili,JIANG Feng.Survey on image segmentation methods[J].Application Research of Computers,2017,34(7):1921-1928(in Chinese).[周莉莉,姜枫.图像分割方法综述研究[J].计算机应用研究,2017,34(7):1921-1928.]
[2]JIANG Guiping,QIN Wenjian,ZHOU Shoujun,et al.Stateof-the-art in medical image segmentation[J].Chinese Journal of Computers,2015,38(6):1222-1242(in Chinese).[江贵平,秦文健,周寿军,等.医学图像分割及其发展现状[J].计算机学报,2015,38(6):1222-1242.]
[3]Dubey YK,Mushrif MM.FCM clustering algorithms for segmentation of brain MR images[J].Advances in Fuzzy Systems,2016(3):1-14.
[4]Zhang K,Zhang L,Lam KM,et al.A level set approach to image segmentation with intensity inhomogeneity[J].IEEETransactions on Cybernetics,2016,46(2):546-557.
[5]Wang XF,Min H,Zou L,et al.A novel level set method for image segmentation by incorporating local statistical analysis and global similarity measurement[J].Pattern Recognition,2015,48(1):189-204.
[6]Shi N,Pan J.An improved active contours model for image segmentation by level set method[J].Optik-International Journal for Light and Electron Optics,2016,127(3):1037-1042.
[7]Li Z,Zeng L,Xu Y,et al.Level set method for image segmentation based on local variance and improved intensity inhomogeneity model[J].Iet Image Processing,2017,10(12):1007-1016.
[8]Dubrovina-Karni A,Rosman G,Kimmel R.Multi-region active contours with a single level set function[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2015,37(8):1585-1601.
[9]LIANG Bing,XU Hua.Kernel fuzzy C-means clustering based on improved artifical bee colony algorithm[J].Journal of Computer Applications,2017,37(9):2600-2604(in Chinese).[梁冰,徐华.基于改进人工蜂群的核模糊聚类算法[J].计算机应用,2017,37(9):2600-2604.]
[10]ZHANG Hui,ZHU Jiaming,TANG Wenjie.Image segmentation algorithm based on clustering and improved double level set[J].Computer Science,2017,44(S1):198-201(in Chinese).[张辉,朱家明,唐文杰.基于聚类和改进型水平集的图像分割算法[J].计算机科学,2017,44(S1):198-201.]
[11]Zheng Y,Jeon B,Xu D,et al.Image segmentation by generalized hierarchical fuzzy C-means algorithm[J].Journal of Intelligent&Fuzzy Systems,2015,28(2):961-973.
[12]Gong M,Liang Y,Shi J,et al.Fuzzy c-means clustering with local information and kernel metric for image segmentation[J].IEEE Transactions on Image Processing,2013,22(2):573-584.
[13]Nageswari CS,Helenprabha K.Spatially constrained distance regularized level set evolution method for segmentation of hydrops fetalis from ultrasound fetal heart images[J].Design Automation for Embedded Systems,2018(4):1-20.
[14]ZHANG Shifu.Method of segmentation accuracy assessment based on vector-raster overlay analysis[J].Journal of Yantai University(Natural Science and Engineering Edition),2015,28(2):119-124(in Chinese).[张世富.矢量-栅格叠置分析的影像分割精度评价方法[J].烟台大学学报(自然科学与工程版),2015,28(2):119-124.]
[2]JIANG Guiping,QIN Wenjian,ZHOU Shoujun,et al.Stateof-the-art in medical image segmentation[J].Chinese Journal of Computers,2015,38(6):1222-1242(in Chinese).[江贵平,秦文健,周寿军,等.医学图像分割及其发展现状[J].计算机学报,2015,38(6):1222-1242.]
[3]Dubey YK,Mushrif MM.FCM clustering algorithms for segmentation of brain MR images[J].Advances in Fuzzy Systems,2016(3):1-14.
[4]Zhang K,Zhang L,Lam KM,et al.A level set approach to image segmentation with intensity inhomogeneity[J].IEEETransactions on Cybernetics,2016,46(2):546-557.
[5]Wang XF,Min H,Zou L,et al.A novel level set method for image segmentation by incorporating local statistical analysis and global similarity measurement[J].Pattern Recognition,2015,48(1):189-204.
[6]Shi N,Pan J.An improved active contours model for image segmentation by level set method[J].Optik-International Journal for Light and Electron Optics,2016,127(3):1037-1042.
[7]Li Z,Zeng L,Xu Y,et al.Level set method for image segmentation based on local variance and improved intensity inhomogeneity model[J].Iet Image Processing,2017,10(12):1007-1016.
[8]Dubrovina-Karni A,Rosman G,Kimmel R.Multi-region active contours with a single level set function[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2015,37(8):1585-1601.
[9]LIANG Bing,XU Hua.Kernel fuzzy C-means clustering based on improved artifical bee colony algorithm[J].Journal of Computer Applications,2017,37(9):2600-2604(in Chinese).[梁冰,徐华.基于改进人工蜂群的核模糊聚类算法[J].计算机应用,2017,37(9):2600-2604.]
[10]ZHANG Hui,ZHU Jiaming,TANG Wenjie.Image segmentation algorithm based on clustering and improved double level set[J].Computer Science,2017,44(S1):198-201(in Chinese).[张辉,朱家明,唐文杰.基于聚类和改进型水平集的图像分割算法[J].计算机科学,2017,44(S1):198-201.]
[11]Zheng Y,Jeon B,Xu D,et al.Image segmentation by generalized hierarchical fuzzy C-means algorithm[J].Journal of Intelligent&Fuzzy Systems,2015,28(2):961-973.
[12]Gong M,Liang Y,Shi J,et al.Fuzzy c-means clustering with local information and kernel metric for image segmentation[J].IEEE Transactions on Image Processing,2013,22(2):573-584.
[13]Nageswari CS,Helenprabha K.Spatially constrained distance regularized level set evolution method for segmentation of hydrops fetalis from ultrasound fetal heart images[J].Design Automation for Embedded Systems,2018(4):1-20.
[14]ZHANG Shifu.Method of segmentation accuracy assessment based on vector-raster overlay analysis[J].Journal of Yantai University(Natural Science and Engineering Edition),2015,28(2):119-124(in Chinese).[张世富.矢量-栅格叠置分析的影像分割精度评价方法[J].烟台大学学报(自然科学与工程版),2015,28(2):119-124.]