不同粒度遥感信息的非线性优化Otsu分割算法
|
摘要
遥感影像具有覆盖面广、光谱信息丰富及不同粒度的遥感信息应用需求特点,传统的图像分割算法不能较好地适用于这类图像的信息提取。针对遥感影像多波段的特性和遥感信息不同粒度需求的特点,基于最大类间方差算法(Otsu算法),设计了不同粒度遥感信息的非线性优化Otsu分割算法:(1)引入PCA算法对遥感影像的多波段数据进行降维,降低了遥感影像的信息冗余度;(2)基于最小值判断,添加分割算法的终止条件,提高了不同粒度遥感信息分割的计算效率。最后,以舟山海域的空间地物信息提取为例,比较了非线性优化Otsu分割算法与传统的Otsu、2D-Otsu、K-means、FCM分割算法的优劣性。结果表明,非线性优化Otsu分割算法:(1)兼顾了遥感影像的波谱信息,降低了遥感信息的错分率;(2)通过降低遥感影像的信息冗余,提高了计算效率;以两类地物类别提取为例,与Otsu算法相比,时间效率提高了59.88%;(3)通过添加计算约束条件,求解不同地物类别的分割阈值,满足了不同粒度遥感信息的应用需求。
Remote sensing image has the characteristics of large-scale coverage,rich spectral information and different granularity requirements.But the traditional segmentation theory is not suitable for remote sensing image.Aiming at the characteristic of multi-band remote sensing image and different granularity requirement,this paper improved the maximum inner variance algorithm(Otsu),designed a nonlinear optimization of multi-dimension Otsu algorithm:(1)using PCA algorithm reduced the multi-band information of remote sensing image,which reduced the information redundancy of remote sensing image;(2)adding termination condition of segmentation algorithm,by calculating minimum value to determine the calculation termination,improved the efficiency of image segmentation.Finally,taking the information extraction of islands as an example,it analyzed the advantage and disadvantage of nonlinear optimization Otsu segmentation algorithm and Otsu,2 D-Otsu,K-means and FCM segmentation algorithm in detail.The result showed that nonlinear optimization Otsu segmentation algorithm:(1)taking into account the spectral information of remote sensing image,reduces misclassification rate of remote sensing information;(2)the redundancy of remote sensing image is reduced and the computational efficiency is improved;taking two classes feature extraction as an example,the time efficiency is improved by 59.88% compared with Otsu algorithm;(3)add computing constraints conditions to solve the segmentation thresholds of different objects,which satisfies condition of multi granularity demand for remote sensing information.
Remote sensing image has the characteristics of large-scale coverage,rich spectral information and different granularity requirements.But the traditional segmentation theory is not suitable for remote sensing image.Aiming at the characteristic of multi-band remote sensing image and different granularity requirement,this paper improved the maximum inner variance algorithm(Otsu),designed a nonlinear optimization of multi-dimension Otsu algorithm:(1)using PCA algorithm reduced the multi-band information of remote sensing image,which reduced the information redundancy of remote sensing image;(2)adding termination condition of segmentation algorithm,by calculating minimum value to determine the calculation termination,improved the efficiency of image segmentation.Finally,taking the information extraction of islands as an example,it analyzed the advantage and disadvantage of nonlinear optimization Otsu segmentation algorithm and Otsu,2 D-Otsu,K-means and FCM segmentation algorithm in detail.The result showed that nonlinear optimization Otsu segmentation algorithm:(1)taking into account the spectral information of remote sensing image,reduces misclassification rate of remote sensing information;(2)the redundancy of remote sensing image is reduced and the computational efficiency is improved;taking two classes feature extraction as an example,the time efficiency is improved by 59.88% compared with Otsu algorithm;(3)add computing constraints conditions to solve the segmentation thresholds of different objects,which satisfies condition of multi granularity demand for remote sensing information.
引文
[1]刘永学,李满春,毛亮.基于边缘的多光谱遥感图像分割方法[J].遥感学报,2006,10(3):350-356.
[2]MULLER-KARGER F,ROFFER M,WALKER N,et al.Satellite remote sensing in support of an integrated ocean observing system[J].IEEE Geoscience&Remote Sensing Magazine,2013,1(4):8-18.
[3]COHEN W B,GOWARD S N.Landsat’s role in ecological applications of remote sensing[J].Bioscience,2004,54(6):535-545.
[4]GIARDINO C,BRESCIANI M,VILLA P,et al.Application of remote sensing in water resource management:the case study of lake trasimeno,Italy[J].Water Resources Management,2010,24(14):3885-3899.
[5]路京选,宋文龙,曲伟,等.跨时空观下的遥感应用新视野[J].遥感学报,2015,19(6):873-881.
[6]史舟,梁宗正,杨媛媛,等.农业遥感研究现状与展望[J].农业机械学报,2015,46(2):247-260.
[7]陈仲新,任建强,唐华俊,等.农业遥感研究应用进展与展望[J].遥感学报,2016,20(5):748-767.
[8]李刚,万幼川,管玉娟.应用商空间理论的遥感影像多粒度合成分割[J].应用科学学报,2011,29(4):390-396.
[9]刘超,蔡文华,陆玲.图像阈值法分割综述[J].电脑知识与技术,2015(1):140-142.
[10]吴一全,孟天亮,吴诗婳.图像阈值分割方法研究进展20年(1994-2014)[J].数据采集与处理,2015,30(1):1-23.
[11]谢勰,王辉,张雪锋.图像阈值分割技术中的部分和算法综述[J].西安邮电大学学报,2011,16(3):1-5.
[12]Otsu N.A threshold selection method from gray-level histograms[J].IEEE transactions on systems,man,and cybernetics,1979,9(1):62-66.
[13]REDDI S S,RUDIN S F,KESHAVAN H R.An optimal multiple threshold scheme for image segmentation[J].Systems Man&Cybernetics IEEE Transactions on,1984,SMC-14(4):661-665.
[14]LIAO P S,CHEN T S,CHUNG P C.A Fast algorithm for multilevel thresholding[J].Journal of Information Science&Engineering,2001,17(5):713-727.
[15]LIU J,LI W,TIAN Y.Automatic thresholding of gray-level pictures using two-dimension Otsu method[C]∥International Conference on Circuits and Systems,China:IEEE,1991:325-327.
[16]ZHOU C,TIAN L,ZHAO H,et al.A method of two-dimensional Otsu image threshold segmentation based on improved firefly algorithm[C]∥Cyber Technology in Automation,Control,and Intelligent Systems,China:IEEE,2015:1420-1424.
[17]BHANDARI A K,SINGH G K,SINGH G K.Modified artificial bee colony based computationally efficient multilevel thresholding for satellite image segmentation using Kapur’s,Otsu and Tsallis functions[J].Expert Systems with Applications,2015,42(3):1573-1601.
[18]GUO W Y,WANG X F,XIA X Z.Two-dimensional Otsu’s thresholding segmentation method based on grid box filter[J].Optic-International Journal for Light and Electron Optics,2014,125(18):5234-5240.
[19]BHANDARI A K,SINGH V K,KUMAR A,et al.Cuckoo search algorithm and wind driven optimization based study of satellite image segmentation for multilevel thresholding using Kapur’s entropy[J].Expert Systems with Applications,2014,41(7):3538-3560.
[20]王玉,李玉,赵泉华.可变类多光谱遥感图像分割[J].遥感学报,2016,20(6):1381-1390.
[21]张金静,李玉,赵泉华.多主体框架下结合最大期望值和遗传算法的SAR图像分割[J].中国图象图形学报,2016,21(1):86-94.
[22]张海涛,李雅男.阈值标记的分水岭彩色图像分割[J].中国图象图形学报,2015,20(12):1602-1611.
[23]陈鹏翔,杨晟院.区域拟合的背景去除图像分割模型[J].中国图象图形学报,2016,21(6):683-690.
[24]邓敏,李志林,程涛.多粒度的GIS数据不确定性粗集表达[J].测绘学报,2006,35(1):68-74.
[25]李然,干宗良,朱秀昌.基于PCA硬阈值收缩的平滑投影Landweber图像压缩感知重构[J].中国图象图形学报,2013,18(5):504-514.
[26]杜洪,夏欣,琚生根,等.基于PCA图像压缩算法研究与实现[J].四川大学学报(自然科学版),2014,51(5):910-914.
[27]陈善学,胡灿,屈龙瑶.基于自适应波段聚类PCA的高光谱图像压缩[J].科学技术与工程,2015,15(12):86-91.
[2]MULLER-KARGER F,ROFFER M,WALKER N,et al.Satellite remote sensing in support of an integrated ocean observing system[J].IEEE Geoscience&Remote Sensing Magazine,2013,1(4):8-18.
[3]COHEN W B,GOWARD S N.Landsat’s role in ecological applications of remote sensing[J].Bioscience,2004,54(6):535-545.
[4]GIARDINO C,BRESCIANI M,VILLA P,et al.Application of remote sensing in water resource management:the case study of lake trasimeno,Italy[J].Water Resources Management,2010,24(14):3885-3899.
[5]路京选,宋文龙,曲伟,等.跨时空观下的遥感应用新视野[J].遥感学报,2015,19(6):873-881.
[6]史舟,梁宗正,杨媛媛,等.农业遥感研究现状与展望[J].农业机械学报,2015,46(2):247-260.
[7]陈仲新,任建强,唐华俊,等.农业遥感研究应用进展与展望[J].遥感学报,2016,20(5):748-767.
[8]李刚,万幼川,管玉娟.应用商空间理论的遥感影像多粒度合成分割[J].应用科学学报,2011,29(4):390-396.
[9]刘超,蔡文华,陆玲.图像阈值法分割综述[J].电脑知识与技术,2015(1):140-142.
[10]吴一全,孟天亮,吴诗婳.图像阈值分割方法研究进展20年(1994-2014)[J].数据采集与处理,2015,30(1):1-23.
[11]谢勰,王辉,张雪锋.图像阈值分割技术中的部分和算法综述[J].西安邮电大学学报,2011,16(3):1-5.
[12]Otsu N.A threshold selection method from gray-level histograms[J].IEEE transactions on systems,man,and cybernetics,1979,9(1):62-66.
[13]REDDI S S,RUDIN S F,KESHAVAN H R.An optimal multiple threshold scheme for image segmentation[J].Systems Man&Cybernetics IEEE Transactions on,1984,SMC-14(4):661-665.
[14]LIAO P S,CHEN T S,CHUNG P C.A Fast algorithm for multilevel thresholding[J].Journal of Information Science&Engineering,2001,17(5):713-727.
[15]LIU J,LI W,TIAN Y.Automatic thresholding of gray-level pictures using two-dimension Otsu method[C]∥International Conference on Circuits and Systems,China:IEEE,1991:325-327.
[16]ZHOU C,TIAN L,ZHAO H,et al.A method of two-dimensional Otsu image threshold segmentation based on improved firefly algorithm[C]∥Cyber Technology in Automation,Control,and Intelligent Systems,China:IEEE,2015:1420-1424.
[17]BHANDARI A K,SINGH G K,SINGH G K.Modified artificial bee colony based computationally efficient multilevel thresholding for satellite image segmentation using Kapur’s,Otsu and Tsallis functions[J].Expert Systems with Applications,2015,42(3):1573-1601.
[18]GUO W Y,WANG X F,XIA X Z.Two-dimensional Otsu’s thresholding segmentation method based on grid box filter[J].Optic-International Journal for Light and Electron Optics,2014,125(18):5234-5240.
[19]BHANDARI A K,SINGH V K,KUMAR A,et al.Cuckoo search algorithm and wind driven optimization based study of satellite image segmentation for multilevel thresholding using Kapur’s entropy[J].Expert Systems with Applications,2014,41(7):3538-3560.
[20]王玉,李玉,赵泉华.可变类多光谱遥感图像分割[J].遥感学报,2016,20(6):1381-1390.
[21]张金静,李玉,赵泉华.多主体框架下结合最大期望值和遗传算法的SAR图像分割[J].中国图象图形学报,2016,21(1):86-94.
[22]张海涛,李雅男.阈值标记的分水岭彩色图像分割[J].中国图象图形学报,2015,20(12):1602-1611.
[23]陈鹏翔,杨晟院.区域拟合的背景去除图像分割模型[J].中国图象图形学报,2016,21(6):683-690.
[24]邓敏,李志林,程涛.多粒度的GIS数据不确定性粗集表达[J].测绘学报,2006,35(1):68-74.
[25]李然,干宗良,朱秀昌.基于PCA硬阈值收缩的平滑投影Landweber图像压缩感知重构[J].中国图象图形学报,2013,18(5):504-514.
[26]杜洪,夏欣,琚生根,等.基于PCA图像压缩算法研究与实现[J].四川大学学报(自然科学版),2014,51(5):910-914.
[27]陈善学,胡灿,屈龙瑶.基于自适应波段聚类PCA的高光谱图像压缩[J].科学技术与工程,2015,15(12):86-91.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |