基于GF-1影像面向对象分类方法的水稻种植信息提取研究

Planting Information Extraction of Rice by Object-oriented Classification Method based on GF-1 Images

应用遥感技术提取水稻种植信息是农业遥感的重要内容。GF-1卫星WFV数据为农业信息提取提供了新的途径,面向对象的分类方法是遥感解译的重要方法。本研究以扬州市为研究区域,基于GF-1影像WFV数据,采用面向对象的分类方法,提取水稻种植信息,并实地调查验证试验结果,试图探讨GF-1数据面向对象分类方法在水稻种植信息提取中的可行性与影响提取精度的因素。结果表明,应用GF-1数据,采用面向对象的分类方法能够很好地完成扬州市水稻种植信息的提取,2016年扬州市有水稻种植面积214 524 hm^2,总体精度达到98.5%,Kappa系数0.95,面积精度达97.5%;实地考察能够提高提取精度,地形破碎程度越低,提取精度越高。

Rice planting information extraction by remote sensing is an important part of agricultural remote sensing.GF-1 satellite WFV data provides a new way for agricultural information extraction,object-oriented classification method is an important method of remote sensing interpretation.This research takes Yangzhou as the research area,based on the GF-1 image data,uses the object-ori-ented classification method,extracts the rice planting information,and carries on the field investigation verification test result.The feasibility of GF-1 data oriented object classification in extracting rice planting information and the factors affecting extraction preci-sion are discussed.The results showed that GF-1 data can be used to extract rice planting information in Yangzhou by object-oriented classification method.Rice planting area was 214 524 hm^2 in Yangzhou City,the overall accuracy of rice was 98.5%,Kappa coefficient was 0.95,area accuracy was 97.5%.Field investigation can improve the extraction accuracy.The degree of terrain fragmentation affects the extraction accuracy,with the decrease of terrain fragmentation,the extraction accuracy is increased.