With the rise of the wisdom of urban construction, the intelligent city requirements expression to fine the geometric structure of spatial entity, so as to realize the refined analysis of urban geographic information, to achieve the purpose of intelligent management. But the traditional geographic entity geometric data organization method ignores the express geographic entity's internal geometry information,, and can not meet the requirements of the fine analysis. Therefore, it is a problem to be solved that how to make a thorough physical and geometrical data of its spatial geometry. In this paper, we propose a method of organizing the Tetrahedral Network in the geographic entity and combining with hierarchical representation for B-rep model to solve the problem.
In view of monitoring the changes of buildings on Earth's surface ,by analyzing the distribution characteristics of building in remote sensing image, combined with multi-scale in image segmentation and the advantages of mathematical morphology, this paper proposes a multi-scale combined with mathematical morphology of high resolution remote sensing image segmentation method, and uses the multiple fuzzy classification method and the shadow of auxiliary method to extract information building, With the comparison of k-means classification, and the traditional maximum likelihood classification method, the results of experiment object based on multi-scale combined with mathematical morphology of image segmentation and extraction method, can accurately extract the structure of the information is more clear classification data, provide the basis for the intelligent monitoring of earth data and theoretical support.