蝙蝠优化的二维Tsallis熵多阈值SAR图像分割
|
摘要
针对智能优化SAR图像分割算法存在计算量大、易陷入局部最优、分割精度不够等问题,融合蝙蝠算法和二维Tsallis熵多阈值,提出了一种蝙蝠优化的二维Tsallis熵多阈值SAR图像分割算法。算法利用立方映射均匀化初始蝙蝠种群,引入Levy飞行特征加强算法跳出局部最优能力,使用Powell局部搜索加快算法收敛等3方面改进蝙蝠算法;同时将二维Tsallis熵单阈值分割方法扩展到多阈值分割,建立基于多阈值的选取方法,并结合改进的蝙蝠算法,将二维Tsallis熵多阈值应用于SAR图像分割中。仿真结果表明,与其他智能优化分割算法相比,本分割算法在边缘处理和分割精度上都有明显优势。
For intelligent optimization of SAR image segmentation algorithm,there are some problems,such as large computation,local peak,and insufficient segmentation accuracy.In this paper,a SAR image segmentation algorithm is proposed,which combines bat algorithm and two-dimensional Tsallis entropy multithreshold.The algorithm homogenizes the initial bat population by using cubic mapping,introduces the Levy flight feature enhancement algorithm to jump out of the local optimum capability,and improves the bat algorithm in three aspects by using Powell local search and accelerating algorithm convergence.The algorithm simultaneously extends the two-dimensional Tsallis entropy single-threshold segmentation method to multi-threshold segmentation and establishes a multi-threshold based selection method.Combined with improved bat algorithm,two-dimensional Tsallis entropy multi-threshold is applied to SAR image segmentation.Simulation results show that the segmentation algorithm has obvious advantages in edge processing and segmentation accuracy compared with other intelligent optimization algorithms.
For intelligent optimization of SAR image segmentation algorithm,there are some problems,such as large computation,local peak,and insufficient segmentation accuracy.In this paper,a SAR image segmentation algorithm is proposed,which combines bat algorithm and two-dimensional Tsallis entropy multithreshold.The algorithm homogenizes the initial bat population by using cubic mapping,introduces the Levy flight feature enhancement algorithm to jump out of the local optimum capability,and improves the bat algorithm in three aspects by using Powell local search and accelerating algorithm convergence.The algorithm simultaneously extends the two-dimensional Tsallis entropy single-threshold segmentation method to multi-threshold segmentation and establishes a multi-threshold based selection method.Combined with improved bat algorithm,two-dimensional Tsallis entropy multi-threshold is applied to SAR image segmentation.Simulation results show that the segmentation algorithm has obvious advantages in edge processing and segmentation accuracy compared with other intelligent optimization algorithms.
引文
[1]何志勇,孙立宁,黄伟国,等.基于Otsu准则和直线截距直方图的阈值分割[J].光学精密工程,2012,20(10):2315-2323.
[2]文政颖,于海鹏.基于多Gamma分布模型的SAR图像直方图分割算法[J].计算机工程与设计,2014,35(6):2104-2108.
[3]陈祥,孙俊,尹奎英,等.基于Otsu与海域统计特性的SAR图像海陆分割算法[J].数据采集与处理,2014,29(4):603-608.
[4]MA Xiaole,LIU Shuaiqi,HU Shaohai,et al.SAR Image Edge Detection via Sparse Representation[J].Soft Computing,2018,22(8):2507-2515.
[5]李淑清,李瑞华,王潇.基于SAR图像的海洋溢油分割方法研究[J].测绘工程,2017,26(2):37-41.
[6]ZHANG Shi,SHE Lihuang,LU Li,et al.A Modified Fuzzy C-Means for Bias Field Estimation and Segmentation of Brain MR Image[C]∥25th Chinese Control and Decision Conference, Guiyang:IEEE,2013:2080-2085.
[7]杨晓波.结合小波变换和K均值聚类的SAR舰船目标分割[J].舰船科学技术,2016,38(18):169-171.
[8]汪华章,宰文姣.改进型脉冲耦合神经网络高分辨率SAR图像分割[J].北京大学学报(自然科学版),2013,49(2):176-182.
[9]CHEN L C,PAPANDREOU G,KOKKINOS I,et al.Semantic Image Segmentation with Deep Convolutional Nets and Fully Connected CRFs[C]∥3rd International Conference on Learning Representations,San Diego:[s.n.],2015:1-14.
[10]MALMGREN-HANSEN D, NOBEL-JRGENSEN M.Convolutional Neural Networks for SAR Image Segmentation[C]∥IEEE International Symposium on Signal Processing and Information Technology,Abu Dhabi:IEEE,2016:231-236.
[11]BASELICE F,FERRAIOLI G.Unsupervised Coastal Line Extraction from SAR Images[J].IEEE Geoscience and Remote Sensing Letters,2013,10(6):1350-1354.
[12]CAO Jiannong.Review on Image Segmentation Based on Entropy[J].Pattern Recognition and Artificial Intelligence,2012,25(6):958-971.
[13]LIN Qianqian,OU Congjie.Application of Two Dimensional Tsallis Entropy in Image Thresholding Segmentation[J].Transducer and Microsystem Technologies,2014,33(7):150-153.
[14]SAHOO P K,ARORA G.Image Thresholding Using Two-Dimensional Tsallis-Havrda-Charvát Entropy[J].Pattern Recognition Letters,2006,27(6):520-528.
[15]陈恺,陈芳,戴敏,等.基于萤火虫算法的二维熵多阈值快速图像分割[J].光学精密工程,2014,22(2):518-523.
[16]崔丽群,黄殿平,宋晓.基于云模型鱼群算法的多阈值图像分割研究[J].计算机工程与应用,2017,53(7):204-208.
[17]杨秋翔,周海芳,贾彩琴,等.基于改进布谷鸟搜索算法的二维Tsallis熵多阈值快速图像分割[J].小型微型计算机系统,2016,37(3):617-621.
[18]马英辉,吴一全.基于改进的指数交叉熵和萤火虫群优化的工业CT图像分割[J].图学学报,2017,38(1):57-62.
[19]尹晓叶,李俊吉.基于增强布谷鸟搜索的图像分割算法[J].控制工程,2017,24(10):2118-2124.
[20]鲍义东,周改云,赵伟艇.自适应蚁群和模糊聚类的SAR图像分割[J].测绘科学,2016,41(8):121-124.
[21]WU Yiquan,SONG Yu,ZHOU Huaichun.State Identification of Boiler Combustion Flame Images Based on Gray Entropy Multiple Thresholding and Support Vector Machine[J].Proceedings of the CSEE,2013,33(20):66-73.
[22]YANG Xinshe.Nature Inspired Meta-Heuristic Algorithms[M].2nd ed.Frome:Luniver Press,2010:97-104.
[23]周燕,刘培玉,赵静,等.基于自适应惯性权重的混沌粒子群算法[J].山东大学学报(理学版),2012,47(3):27-32.
[24]冯辉宗,彭丹,袁荣棣.基于PSO-SVM的发动机故障诊断研究[J].计算机测量与控制,2014,22(2):355-357.
[25]刘笃晋,梁弼,袁小艳,等.基于改进布谷鸟算法的彩色图像多阈值分割[J].计算机工程与设计,2016,37(12):3322-3326.
[26]田小林,焦李成,缑水平.基于PSO优化空间约束聚类的SAR图像分割[J].电子学报,2008,36(3):453-457.
[27]葛亮,杨竣铎.基于蚁群优化多层图划分的彩色图像分割方法[J].计算机应用研究,2015,32(4):1265-1268.
[28]赵文昌,李忠木.融合改进人工蜂群和K均值聚类的图像分割[J].液晶与显示,2017,32(9):726-735.
[2]文政颖,于海鹏.基于多Gamma分布模型的SAR图像直方图分割算法[J].计算机工程与设计,2014,35(6):2104-2108.
[3]陈祥,孙俊,尹奎英,等.基于Otsu与海域统计特性的SAR图像海陆分割算法[J].数据采集与处理,2014,29(4):603-608.
[4]MA Xiaole,LIU Shuaiqi,HU Shaohai,et al.SAR Image Edge Detection via Sparse Representation[J].Soft Computing,2018,22(8):2507-2515.
[5]李淑清,李瑞华,王潇.基于SAR图像的海洋溢油分割方法研究[J].测绘工程,2017,26(2):37-41.
[6]ZHANG Shi,SHE Lihuang,LU Li,et al.A Modified Fuzzy C-Means for Bias Field Estimation and Segmentation of Brain MR Image[C]∥25th Chinese Control and Decision Conference, Guiyang:IEEE,2013:2080-2085.
[7]杨晓波.结合小波变换和K均值聚类的SAR舰船目标分割[J].舰船科学技术,2016,38(18):169-171.
[8]汪华章,宰文姣.改进型脉冲耦合神经网络高分辨率SAR图像分割[J].北京大学学报(自然科学版),2013,49(2):176-182.
[9]CHEN L C,PAPANDREOU G,KOKKINOS I,et al.Semantic Image Segmentation with Deep Convolutional Nets and Fully Connected CRFs[C]∥3rd International Conference on Learning Representations,San Diego:[s.n.],2015:1-14.
[10]MALMGREN-HANSEN D, NOBEL-JRGENSEN M.Convolutional Neural Networks for SAR Image Segmentation[C]∥IEEE International Symposium on Signal Processing and Information Technology,Abu Dhabi:IEEE,2016:231-236.
[11]BASELICE F,FERRAIOLI G.Unsupervised Coastal Line Extraction from SAR Images[J].IEEE Geoscience and Remote Sensing Letters,2013,10(6):1350-1354.
[12]CAO Jiannong.Review on Image Segmentation Based on Entropy[J].Pattern Recognition and Artificial Intelligence,2012,25(6):958-971.
[13]LIN Qianqian,OU Congjie.Application of Two Dimensional Tsallis Entropy in Image Thresholding Segmentation[J].Transducer and Microsystem Technologies,2014,33(7):150-153.
[14]SAHOO P K,ARORA G.Image Thresholding Using Two-Dimensional Tsallis-Havrda-Charvát Entropy[J].Pattern Recognition Letters,2006,27(6):520-528.
[15]陈恺,陈芳,戴敏,等.基于萤火虫算法的二维熵多阈值快速图像分割[J].光学精密工程,2014,22(2):518-523.
[16]崔丽群,黄殿平,宋晓.基于云模型鱼群算法的多阈值图像分割研究[J].计算机工程与应用,2017,53(7):204-208.
[17]杨秋翔,周海芳,贾彩琴,等.基于改进布谷鸟搜索算法的二维Tsallis熵多阈值快速图像分割[J].小型微型计算机系统,2016,37(3):617-621.
[18]马英辉,吴一全.基于改进的指数交叉熵和萤火虫群优化的工业CT图像分割[J].图学学报,2017,38(1):57-62.
[19]尹晓叶,李俊吉.基于增强布谷鸟搜索的图像分割算法[J].控制工程,2017,24(10):2118-2124.
[20]鲍义东,周改云,赵伟艇.自适应蚁群和模糊聚类的SAR图像分割[J].测绘科学,2016,41(8):121-124.
[21]WU Yiquan,SONG Yu,ZHOU Huaichun.State Identification of Boiler Combustion Flame Images Based on Gray Entropy Multiple Thresholding and Support Vector Machine[J].Proceedings of the CSEE,2013,33(20):66-73.
[22]YANG Xinshe.Nature Inspired Meta-Heuristic Algorithms[M].2nd ed.Frome:Luniver Press,2010:97-104.
[23]周燕,刘培玉,赵静,等.基于自适应惯性权重的混沌粒子群算法[J].山东大学学报(理学版),2012,47(3):27-32.
[24]冯辉宗,彭丹,袁荣棣.基于PSO-SVM的发动机故障诊断研究[J].计算机测量与控制,2014,22(2):355-357.
[25]刘笃晋,梁弼,袁小艳,等.基于改进布谷鸟算法的彩色图像多阈值分割[J].计算机工程与设计,2016,37(12):3322-3326.
[26]田小林,焦李成,缑水平.基于PSO优化空间约束聚类的SAR图像分割[J].电子学报,2008,36(3):453-457.
[27]葛亮,杨竣铎.基于蚁群优化多层图划分的彩色图像分割方法[J].计算机应用研究,2015,32(4):1265-1268.
[28]赵文昌,李忠木.融合改进人工蜂群和K均值聚类的图像分割[J].液晶与显示,2017,32(9):726-735.