首页 | 本学科首页   官方微博 | 高级检索  
     

基于蚁群优化算法的遥感器波段设置探究
引用本文:张路,刘良云,高建威,焦全军,贾建华. 基于蚁群优化算法的遥感器波段设置探究[J]. 遥感技术与应用, 2016, 31(3): 488-496. DOI: 10.11873/j.issn.1004-0323.2016.3.0488
作者姓名:张路  刘良云  高建威  焦全军  贾建华
作者单位:(1.西安科技大学测绘科学与技术学院,陕西西安 710054;2.中国科学院遥感与数字地球研究所,北京 100094)
基金项目:国家自然科学基金项目“中国科学院新型对地观测系统科技创新交叉合作团队”(41325004),国家科技支撑计划课题(2013BAC03B02).
摘    要:卫星载荷研制发射后其光谱和空间观测模式固定,无法根据复杂地表的多样化需求进行实时灵活调整,且目前遥感器波段设置尚不完善还存在优化空间.引进基于蚁群优化算法的波段选择方法(AntColonyOptimization basedBandSelection,ACOBS),结合北美区域33景AVIRIS航空高光谱图像,开展了不同区域、不同地表覆盖类型的高光谱波段优选研究,发现各地表类型优选波段组合存在一定差异,其中4波段组合中红光、近红外波段为2个共同入选波段,6波段组合中绿光、红光、短波红外波段为3个共有波段,8波段组合中紫光、绿光、红光、红边、近红外1、近红外2、短波红外1、短波红外2为8个共有入选波段,其他入选波段与地表覆盖类型有关.在此基础上,进一步开展了多光谱卫星波段设置评价研究,发现:4波段优化方案中,绿光、红光、近红外波段1 (770~895nm)、近红外波段2(900~1350nm)为最优波段组合;6波段优化方案中,绿、红、红边、近红外1(770~895nm)、近红外2(900~1350nm)、短波红外1(1560~1660nm)为最优波段组合;8波段优化方案中,蓝、绿、红、红边、近红外1(770~895nm)、近红外2(900~1350nm)、短波红外1(1560~1660nm)和短波红外2(2100~2300nm)为最优波段组合.研究结果表明Land satTM OLI、SPOT等陆地资源遥感器波段设置还存在一定优化调整空间,特别是红边波段在目前传感器波段设置中没有得到足够重视.

关 键 词:波段设置  蚁群优化算法  波段选择  高光谱数据  
收稿时间:2015-04-28

The Explore of Band Set based on Ant Colony Optimization Algorithm for Remote Sensor
Zhang Lu,Liu Liangyun,Gao Jianwei,Jiao Quanjun,Jia Jianhua. The Explore of Band Set based on Ant Colony Optimization Algorithm for Remote Sensor[J]. Remote Sensing Technology and Application, 2016, 31(3): 488-496. DOI: 10.11873/j.issn.1004-0323.2016.3.0488
Authors:Zhang Lu  Liu Liangyun  Gao Jianwei  Jiao Quanjun  Jia Jianhua
Affiliation:(1.College of Geometrics,Xi’an University of science and Technology ,Xi’an 710054,China;;2.Institute of Remote Sensing and Digital Earth,Chinese Academy of Sciences,Beijing 100094,China)
Abstract:Currently,both spectral parameters settings and spatial observation mode of on|orbit satellite sensors are fixed and cannot be flexibly adjusted in real|time according to the diversified needs of observing complex ground surface.However,the spectral band setting of current remote sensing sensor is still needed to be optimized.In this paper,using the band selection method based on the ant colony optimization algorithm (Ant Colony Optimization|based Band Selection,ACOBS),Combining 33 views AVIRIS Airborne Hyperspectral image in North America,carried out optimum bands combination research of hyperspectral in different regions,different land cover types.The result shows that optimum combinations of spectral bands are different for different land cover types.Red and near infrared bands are the candidates in optimal 4|band combination,green、red and short|wave infrared bands are the candidates in optimal 6|band combination,purple,green,red,red edge,near infrared1,near infrared2,SWIR1and SWIR 2 are the candidates in optimal 8|band combination,and other selected bands in different band combinations are different depend on different land cover types.Optimum spectral bands for multi|spectral satellite sensor was further analyzed,the result shows that optimal 4|band combination is green,red,near|infrared 1 (770~895 nm) and near|infrared 2 (900~1 350 nm); the optimal 6|band combination is green,red,red edge,near infrared 1(770~895 nm) and near infrared 2(900~1 350 nm),SWIR1(1 560~1 660 nm),the optimal 8|band combination is blue,green,red,red edge,near|infrared 1(770~895 nm),near|infrared 2(900~1 350 nm),SWIR1(1 560~1 660 nm) and SWIR 2 (2 100~2 300 nm).It is indicated that spectral bands setting of Landsat TM/OLI,SPOT is not optimum for observing earth surface,and red edge band is neglected in current optical multi|spectral satellite sensor.
Keywords:Sensor  Band setting  Ant colony optimization  Band selection  Hyperspectral data  
本文献已被 CNKI 等数据库收录!
点击此处可从《遥感技术与应用》浏览原始摘要信息
点击此处可从《遥感技术与应用》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号