基于GF2号卫星影像的农业信息提取方法对比分析

发布时间：2018-03-04 13:41

本文选题：农业信息　切入点：GF卫星　出处：《测绘通报》2017年01期 　论文类型：期刊论文

【摘要】：以GF2卫星0.8 m全色/3.2 m多光谱分辨率遥感影像为基础数据源,对基于GF2号卫星影像的农业信息提取流程和方法进行了研究与对比分析。首先对GF2号卫星影像进行波谱分析;其次对GF2号影像进行融合,并对多种融合方法进行质量评价;最后选择阈值法、波谱间关系法、非监督分类法和面向对象法分别对GF2号影像数据进行农业信息提取试验,并对信息提取结果进行精度验证和结果分析。试验表明,面向农业信息提取的GF2号卫星影像融合方法中,Pansharp融合算法融合影像色彩正常,无虚影,清晰度高,地类对比度正常,纹理清晰,熵值及与原始多光谱影像的相关系数高。阈值法和谱间关系法适用于提取单要素农业信息,非监督分类法能够初步获取研究区土地利用情况,面向对象法提取研究区全要素信息精度高。总体来说,不同信息提取方法具有各自的优势,在具体实际应用中,可以根据目标地类的波谱特性,选择适宜的遥感影像处理和信息提取方法。
[Abstract]:Based on the GF2 image of 0.8 m panchromatic / 3.2m multispectral resolution, the extraction process and method of agricultural information based on GF2 satellite image are studied and compared. Firstly, the spectral analysis of GF2 satellite image is carried out. Secondly, we fuse GF2 image and evaluate the quality of many fusion methods. Finally, we select threshold method, spectral relationship method, unsupervised classification method and object oriented method to carry out agricultural information extraction test of GF2 image data. The experiment shows that the GF2 satellite image fusion method for agricultural information extraction has normal color, no virtual image, high definition and normal ground contrast. The texture is clear, the entropy value and the correlation coefficient with the original multispectral image are high. The threshold method and the spectral relation method are suitable for extracting agricultural information of single element, and the unsupervised classification method can obtain the land use situation of the study area preliminarily. In general, different information extraction methods have their own advantages. In practical applications, they can be based on the spectral characteristics of target ground classes. Select appropriate remote sensing image processing and information extraction methods.
【作者单位】：辽宁工程技术大学测绘与地理科学学院;中国科学院地理科学与资源研究所;山西家豪测绘集团有限公司;太原理工大学矿业工程学院;
【基金】：国家科技支撑计划(2014BAL07B02)
【分类号】：S127

【相似文献】