期刊文献+

基于水深权重的面向对象精度评价方法

Object-oriented Accuracy Evaluation Method based on Water Depth Weight
下载PDF
导出
摘要 对地表覆盖产品进行精度验证是保证其推广应用的首要前提,但用于精度评价的参考数据受气候因素影响,存在水体对象边界不确定性和图斑破碎问题,造成现有方法难以准确评价水体、湿地等地类的精度。因此,提出了基于水深加权的几何相似度算法和基于纠正权重中心的位置相似度算法,该算法将地类作为一个整体看待,采用水深加权法获得对象的位置和几何相似度,将几何相似度和位置相似度相加计算得到对象精度,并将对象的面积作为权重计算该地类的总体精度。实验选用GLC_FCS和Landsat8遥感卫星影像作为参考数据,对GlobeLand30水体地类进行验证,得到2015年水体地类的总体精度为87.53%,修正了其他方法得到的评价结果与实际不符的情况。 The accuracy verification of surface coverage products is the first prerequisite to ensure their popularization and application.However,the reference data used for accuracy evaluation is af-fected by climate and other factors,and there are uncertainties in the boundary of water objects and the fragmentation of object patches,which makes it difficult for existing methods to accurately evalu-ate the accuracy of water bodies,wetlands and other areas.Therefore,this paper proposes a geomet-ric similarity algorithm based on water depth weighting and a location similarity algorithm based on the correction of weight center.This algorithm takes the land category as a whole,uses the water depth weighting method to obtain the location and geometric similarity of the object,adds the geo-metric similarity and location similarity to calculate the object accuracy,and uses the area of the ob-ject as the weight to calculate the overall accuracy of the land category.GLC_FCS is selected for the experiment and Landsat8 remote sensing satellite images were used as reference data to verify GlobeLand30 water and land types,and the overall accuracy of water and land types in 2015 was 87.53%,which corrected the discrepancy between the evaluation results obtained by other methods and the actual situation.
作者 仲尹文 陈斐 ZHONG Yinwen;CHEN Fei(East China University of Technology,330013,Nanchang,PRC)
出处 《江西科学》 2023年第3期559-564,共6页 Jiangxi Science
基金 国家自然科学基金项目(41801228)。
关键词 地表覆盖 精度评价 面向对象 水深权重 surface coverage precision evaluation object-oriented water depth weight
  • 相关文献

参考文献8

二级参考文献90

共引文献565

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部