| In the context of digital city construction, most current3D building reconstructing methods lacks explicit description of space and spatial relation, while this paper holds that a proper3D building reconstructing method should concentrate on not only the information of building solid but also the interrelated space and spatial relation information, which are the essential to fully express architectures, and to support various related spatial analysis. For the situation that the architecture industry in China retains a considerable amount of construction drawing data, this paper focuses on the information extraction as well as the solid modeling from the established data.Firstly, the relation network between building components including aggregation, containing and adjacency information are obtained, aided by a classification of building components and based on the prior work of component extraction from drawings.Then, under a unified parametric3D solid modeling technology tweaked for the differences and similarities in architectures, each building component could acquire its proper solid model representation, which is the fundamental of building space, by combining the component relation network and the components’information.Finally, considering the space as the essence of buildings, the information of building spaces are calculated from disorderly organized components, including their semantics, properties, locations, geometries and interrelated components. The spatial relation between spaces including adjacency, connectivity and aggregation are then obtained.The result of the study shows that the above stages of3D modeling progress based on CAAD construction drawing data are not isolated but closely related, where solid modeling, space extraction and relation reconstruction are forming an integrated systemic task. The conclusion is consistent with the building data model achievement of our subject group, and has certain reference value for upcoming research. |
PDF Full Text Request