Journal of Geo-information Science >
Analysis of Methods for Building Components Information Extraction Based on Engineering Data
Received date: 2015-01-07
Request revised date: 2015-03-25
Online published: 2015-09-07
Copyright
Building is an important part of city. With the development of Smart City, 3D building model and its applications are going further from the external into the internal environment. Building engineering data contain rich information about building components throughout architecture design and construction process. They are good sources for building modeling and related applications. In this paper, building engineering data are divided into unstructured data, semi-structured data and structured data according to the data development status. The research status of information extraction from unstructured data and structured data are summarized from different aspects, including object recognition methods, extracted contents and application aims. Furthermore, the specification of layers and geometric primitive features for semi-structured data are studied based on the national and industry standards. In conclusion, the standardized layer information reduces the complexity of semi-structured data and assists the process of object recognition and information extraction. With its abundant sources and lower complexity, semi-structured data is becoming an important data source for the construction of building database in Smart City.
JIA Mingyuan , ZHOU Liangchen , LV Guonian , WAN Qing . Analysis of Methods for Building Components Information Extraction Based on Engineering Data[J]. Journal of Geo-information Science, 2015 , 17(9) : 1022 -1028 . DOI: 10.3724/SP.J.1047.2015.01022
Fig. 1 Building engineering data types图1 建筑工程数据类型 |
Fig. 2 Building elements with different geometric features图2 具有不同几何特征的建筑对象 |
Fig. 3 Geometric features of walls图3 墙体几何特征 |
Tab. 1 Information extraction methods based on unstructured data表1 非结构化数据的信息提取方法 |
研究团队 | 识别方法 | 识别范围 | 提取内容 | 应用目标 |
---|---|---|---|---|
Ah-Soon等 | 约束网络 | 墙、门、窗等 | 语义、几何、定位 | 3D重建 |
Lladós等 | 图匹配 | 门、窗、家具等 | 语义 | 手绘图纸识别 |
朱俊芳等 | 样例驱动符号识别 | 墙、门、窗、柱 | 语义、几何、定位 | 3D重建 |
Zhi等 | 拓扑特征 | 门、窗、房间等 | 语义、关系 | 拓扑分析、疏散模拟 |
路通等 | 拓扑特征和层次推进 | 墙、门、窗、柱、梁等 | 语义、几何、定位 | 3D重建和工程算量 |
Tab. 2 Information extraction methods based on structured data表2 结构化数据的信息提取方法 |
研究团队 | 数据源 | 识别范围 | 提取内容 | 应用目标 |
---|---|---|---|---|
万小飞等 | CAAD文件 | 主要建筑构件类型 | 语义、几何、定位、关系 | 3D重建 |
Isikdag等 | IFC数据 | BIM支持的建筑构件类型 | 语义、几何、定位、属性 | 生成LOD |
El-Mekawy等 | IFC数据 | BIM支持的建筑构件类型 | 语义、几何、定位、属性 | 与CityGML转换 |
Rueppel等 | IFC、gbXML | BIM支持的建筑构件类型 | 语义、几何、定位、关系 | 室内导航、火灾模拟 |
Tab. 3 Specification of architectural drawing’s layers表3 建筑图纸分层规范 |
图层关键字 | GBT18112-2000 | 天正建筑软件 | 注释 |
---|---|---|---|
轴线 | 建筑-轴网 | DOTE | 平面轴线 |
阳台 | 建筑-阳台 | BALCONY | 阳台,雨篷 |
柱 | 建筑-柱子 | COLUMN | 柱子 |
门窗 | 建筑-门窗 | WINDOW | 门窗对象 |
轴标 | 建筑-轴标 | AXIS | 轴号与轴线尺寸标注 |
地面 | 建筑-地面 | GROUND | 地面与散水 |
墙线 | 建筑-墙体 | WALL | 墙体 |
屋顶 | 建筑-屋顶 | ROOF | 屋顶、老虎窗等 |
雨水管线 | 建筑-雨水管 | PUB_STRM | 雨水管线 |
楼梯 | 建筑-楼梯 | STAIR | 楼梯、坡道、台阶和休息平台等 |
电梯 | 建筑-电梯 | EVTR | 电梯轿厢和平衡块 |
扶手 | 建筑-楼梯-扶手 | HANDRAIL | 扶手 |
公标 | 公用-尺寸 | PUB_DIM | 尺寸标注 |
标高 | 公用-标高 | DIM_ELEV | 标高标注 |
符号 | 公用-符号 | DIM_SYMB | 其它各种符号标注 |
公文 | 公用-说明 | PUB_TEXT | 各种说明文字 |
窗名 | 建筑-门窗-编号 | WINDOW_TEXT | 门窗编号 |
公框 | 公用-图框 | PUB_TITLE | 图框线与标题栏、会签栏图块等 |
Fig. 4 Reducing the drawing’s complexity by using layers图4 分层降低图纸复杂性 |
Tab. 4 Comparison of different types of building engineering data表4 不同类型建筑工程数据对比 |
数据类型 | 数据来源 | 数据组织方式 | 提取方法 |
---|---|---|---|
非结构化数据 | CAD历史数据/扫描矢量化数据 | 由基本图元组成 | 通过符号识别的方法进行对象识别与提取 |
半结构化数据 | 按照制图标准生产的数据/分解后的CAAD数据 | 在图元、文字等基础上,按照规范化的图层进行组织 | 结合图层信息进行对象识别与 提取 |
结构化数据 | 完整的CAAD数据/BIM数据 | 具有数据模型,按照结构化方式组织 | 通过数据模型进行对象信息理解与抽取 |
The authors have declared that no competing interests exist.
[1] |
|
[2] |
|
[3] |
|
[4] |
|
[5] |
|
[6] |
|
[7] |
|
[8] |
|
[9] |
|
[10] |
|
[11] |
|
[12] |
|
[13] |
|
[14] |
|
[15] |
|
[16] |
|
[17] |
|
[18] |
|
[19] |
|
[20] |
|
[21] |
|
[22] |
|
[23] |
|
[24] |
|
[25] |
|
[26] |
|
[27] |
|
[28] |
|
[29] |
|
[30] |
|
[31] |
|
[32] |
|
[33] |
|
[34] |
|
[35] |
|
[36] |
|
/
〈 | 〉 |