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：2003
卷号：XXXIV-3/W13
出版社：Copernicus Publications
摘要：Laser scanning produces a set of irregularly spaced data points, the values of which indicate the height of the surface above datum. These data are frequently used for creating digital surface models (DSM) which represent the upper surface of the earth, or features on that surface. As the first stage in this modelling process, many commercial software packages require that the irregularly spaced raw data points are interpolated onto a regular grid before a DSM can be created. This process of interpolation introduces error into the surface model. Errors within height models have, in the past, been communicated in terms of global measures of accuracy for the model. Such quantification ignores the spatial structure of errors across the surface, hindering subsequent analysis. This paper investigates the structure, spatial patterns, and magnitude of these errors, with the aim of aiding future decision-making based on the derived DSM. This paper focuses on a number of processes which affect the level and the location of errors within a model, including comparing a variety of interpolation methods and altering the size of grid spacings for the resampling. Understanding both the quantities and the spatial structure of the error is important where models are to be used for subsequent detailed analysis, such as for telecommunications planners and flood modellers. The ability to anticipate where the largest errors are likely to occur is of paramount importance in order to assess the confidence value which may be attached to any subsequent analysis based upon the models
