文章基本信息

标题：Information-theoretic Perspectives of Image-based Categorical Mapping
本地全文：下载
作者：Zongjian Lin ; Yan Chen ; Jingxiong Zhang 等
期刊名称：ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
印刷版ISSN：2194-9042
电子版ISSN：2194-9050
出版年度：2008
卷号：XXXVII Part B2
页码：885-890
出版社：Copernicus Publications
摘要：This paper discusses quantitative evaluation of information in remote sensing and GIS data. The formula for calculating information contents of remote sensing imagery are developed, taking into account of signal-noise-ratio (SNR), inter-pixel and cross-band correlations, while information amounts in GIS data are calculated by referring to geometric position, attributes, and the equivocation in both. Tests were carried out to compute and analyze information contents in remote sensing imagery and area-class maps produced by four different methods. It was found that amounts of information in resultant area-class maps are consistent with classifiers' performances, and are decreased in comparison with the source imagery. Further research will be directed towards the relationships between information contents and uncertainty
关键词：Measures of Information; Remote Sensing Imagery; Area-class Maps; Uncertainty; Correlation; Equivocation