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：2006
卷号：XXXVI-2/W40
页码：65-70
出版社：Copernicus Publications
摘要：A methodology for the automatic adaptation of models for the extraction of landscape objects with a variable number of similar object parts to a coarser image resolution is proposed in this paper. A given object model, consisting of parallel line-type object parts and represented as a semantic net, can with these methods be fully automatically adapted for the extraction of that same object in another coarser resolution. The methods are developed exemplarily for the landscape object road. Roads can be modelled with a beforehand unknown number of lanes, allowing thereby the description of different types of roads by a single model. This paper also introduces dynamic relations for realizing the variability in the models. The adaptation utilizes the linear scale-space for the prediction of the appearance of the road in the lower resolution and also considers the possible occurrence of scale-space events during scale change. The developed automatic adaptation methods are demonstrated using an example for a variable road model. Intermediate steps and results of the adaptation process are presented and discussed
