结合波段相关性的FY-3C MERSI异常条带检测方法  

Automatic Detection of Strip Defects for FY-3C MERSI Imagery by Band Correlation Theory

作　　者：阎宏波 樊宪磊 瞿瑛 YAN Hongbo;FAN Xianlei;QU Ying(School of Geographical Sciences,Northeast Normal University,Changchun 130024,China)

机构地区：[1]东北师范大学地理科学学院,长春130024

出　　处：《遥感信息》2020年第3期50-56,共7页Remote Sensing Information

基　　金：国家自然科学基金青年项目(41601349);国家自然科学基金面上项目(41971287);吉林省科技发展计划优秀青年人才基金项目(20180520220JH);中央高校基本科研业务费专项资金项目(2412019FZ003)。

摘　　要：针对风云三号C星(FY-3C)搭载的中分辨率光谱成像仪(medium resolution spectral imager,MERSI)数据存在异常条带的问题,提出了一种结合波段相关性的异常条带检测方法。算法结合了FY-3C MERSI数据波段相关性和异常条带分布规律,通过构建波段差矩阵,快速准确地得到影像数据中存在的异常条带。定量实验分析结果表明,由于阈值选取和确定分割界线起止顺序等原因,基于数学形态法的检测结果会出现异常行条带定位不准确、存在大量误检和漏检的现象(行检测率64.5%,误检率45.2%,漏检率35.5%);而本文算法可以准确定位异常条带的行列号,能够有效地检测出FY-3C MERSI数据的异常行列条带(行检测率98.1%,误检率0.0%,漏检率1.9%),更适用于对大数据量的FY-3C MERSI影像数据异常行列条带进行批量检测处理。The strip defects are annoying noises which widely exist in remote sensing imagery.It has been reported that the validity and reliability of results derived from remote sensing data can be seriously reduced by the strip defects.Therefore,it is of great significance to detect strip defects of remote sensing imagery before quantitative processing and analysis.Herein,we propose an automatic detection method for medium resolution spectral imager(MERSI)data onboard Fengyun-3C(FY-3C)satellite based on the band correlation theory(BCT).In this study,67 scenes of FY-3C MERSI data are carried out for validation.The strip defects are detected with the BCT and mathematical morphological method,separately.The results show that large number of strip defects are omitted or erroneously detected(overall accuracy is 64.5%,commission error is 45.2%,and omission error is 35.5%)mainly due to the determination of thresholds and starting/ending positions of the segmentation boundary,while the method proposed by this paper can accurately and efficiently locate the positions of row/column strip defects of FY-3C MERSI data(overall accuracy is 98.1%,commission error is 0.0%,and omission error is 1.9%).The results indicate that the BCT method is more suitable for automatic detection of strip defects for FY-3C MERSI data,which lays a foundation for subsequent processing and analysis.

关 键 词：FY-3C MERSI 异常条带检测 波段相关性 数学形态法 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]