基于 AISI 网络的 BIM 三维重建方法研究

自动从点云数据生成建筑信息模型(BIM)一直是建筑自动化领域的研究热点。基于传统算法的建筑自动三维重建的缺点包括人工设计特征，识别过程复杂，应用场景有限等。随着三维机器学习领域的不断成熟，处理点云便有了新的手段。通过引入实例分割中的 ASIS 网络框架对点云进行处理，即从扫描点云场景中自动分割和分类建筑构建元素并得到实例分割矩阵。接着，基于包围盒假设从得到的实例分割矩阵中提取建筑构件外轮廓参数，并将外轮廓参数和分割的语义分类结果作为 BIM 建模的构件参数。最后，将这些提取的构件参数输入到自制的 IFC 生成器中，自动生成基于工业基础类(IFC)标准的 BIM 模型。实验表明，利用无噪点点云方法，可实现基于曼哈顿世界假设下的室内单房间的三维重建。

Research on 3D reconstruction method of BIM based on ASIS network

Automatic generation of building information model (BIM) from point cloud data has beena hot topic in building automation. The disadvantages of automatic 3D building reconstruction basedon traditional algorithms include manual design features, complex identification process and limitedapplication scenes. As 3D machine learning matures, there are new ways to deal with point clouds.The point cloud was processed by introducing the network ASIS in instance segmentation, that is, thearchitectural construction elements were automatically segmented and classified from the scanning ofpoint cloud scene and the instance segmentation matrix was obtained. Then, the external contourparameters of building components were extracted from the obtained instance segmentation matrixbased on the bounding box hypothesis, and the external contour parameters and the semanticclassification results of segmentation were taken as component parameters of BIM. Finally theseextracted component parameters were input into the self-made IFC generator to automaticallygenerate the BIM model based on the Industry Foundation Class (IFC) standard. Experiments showthat the method of noise-free point cloud can realize 3D reconstruction of indoor single room basedon the hypothesis of Manhattan world.