| In the field of GIS 3D modeling,the demand for automatically constructing 3D models by using architectural engineering views is rapidly increasing.The reading,analysis and 3D construction of two-dimensional architectural engineering views are an important research topic involving computer graphics and computer-aided design.After decades of research and practice,many kinds of construction algorithms have been proposed at home and abroad,but there is still a big gap in practical application,which needs further research in-depth.Therefore,it is of great theoretical significance and application value to study the automatically constructing method of 3D models by using architectural engineering multi-view.In this paper,the 3D model constructing methods based on B-Rep and based on CSG are compared and analyzed.Combining the principle of topological geometry,a method for 3D model construction is studied based on the spatial topology relationship in multi-view.The main research contents and innovations include:(1)Aiming at the problems that the design standards of architectural engineering views are not fixed and it is difficult to be automatically recognized by computer,the drawing method of architectural multi-views is regulated and the position distribution of multi-views is adjusted,which lays the foundation for improving the efficiency of multi-view recognition and 3D model reconstruction.(2)In order to effectively store the extracted primitive data and satisfy the input environment of the 3D construction algorithm,the structure design and preprocessing methods of the primitive data are given.That is,the six basic views contained in the multi-view are quickly and efficiently separated with the moving dividing line method,and the coordinate transformation is used to convert the multi-view from the plane coordinates to the corresponding 3D spatial coordinates.(3)In order to construct the topological structure of the 3D base entity,firstly,a contour ring searching algorithm is proposed in this paper based on two branch points to identify basic contour rings in the view;then,a contour ring matching algorithm is proposed based on the base ring in the top view,to extract the projected base ring constituting the base entity from multi-view;finally,the basic contour ring is combined to construct the topological structure of the base entity.(4)In order to construct the geometric structure of the 3D base entity,a 3D vertex search algorithm based on projection points is proposed,and a calculation method of the 3D vertex coordinates is given.Finally,the projection points in the base ring are replaced with 3D vertices,and the geometric structure of the base entity is successfully constructed.(5)The data containing the topological relations of points,surfaces,and volumes is exported to the obj file,which is the general format of 3D model suitable for all kinds of applications.In this paper,the object-oriented programming ideas is used to develop a function library for building 3D models in multiple views,and the 3D model construction experiments are performed by using multiple views of an actual building as an case.With this function library,two-dimensional multi-view can be converted to 3D model in obj format,and the modeling design effect is demonstrated in the 3D WebGIS system based on Cesium.The research results in this paper provide an effective theory and method for automatically constructing a 3D model from a two-dimensional view. |