Research On Indoor Component Extraction Method For Multi Story Buildings Based On 3D Laser Point Cloud

With the continuous development and iteration of 3D point cloud data collection devices and algorithms,the acquisition and processing of indoor 3D data has become easier.Nowadays,indoor navigation is widely used in public places,making it crucial to use indoor 3D point clouds to establish models.However,in general,the indoor environment is complex,with diverse structures,and there are a large number of unrelated indoor furniture,which blocks the required components,resulting in more noise and redundant information in the scanning point cloud.Moreover,3D point clouds have the characteristics of large data volume,more redundant information,no connectivity between point clouds,no structural attributes,and no Semantic information.3D model reconstruction using point clouds has also become a research hotspot and difficulty.In the research work of extracting indoor components from point clouds,the extracted elements can basically include walls,floors,ceilings,doors and windows,but most studies do not include beams and columns.For the extraction of the staircase part,most scholars are only limited to image based extraction and have not conducted data extraction modeling research using more accurate data such as point clouds.In the digital modeling of buildings,accurate extraction of indoor beams is crucial.Cross beams are an important component of building structures,which are of great significance for understanding the structural characteristics and performance of buildings.However,in current domestic and international research,the main focus is on wall lines,doors and windows,and there is a lack of extraction of beam elements.Doors and windows are the communicating vessels between indoor space and indoor space,indoor space and outdoor space,which determine the connectivity between spaces.As the basic component connecting floors,stairs also play a crucial role in buildings.The main research work and results of this article are as follows:A point cloud extraction method based on layered slicing and overlay analysis is proposed.This method utilizes component information of different heights to obtain complete point cloud data of walls,beam elements,door elements,and other information of indoor components,and realizes the detection of beam and door component positions.This method can simplify the complexity of building geometric feature extraction,obtain information about indoor components such as beam elements and door elements,and provide data and model support for the implementation of various indoor navigation functions such as indoor positioning,path analysis,and real-time navigation.A wall line extraction method based on slice point cloud has been proposed to solve the problems of incomplete wall surfaces,imprecise and missing wall data scanning in actual data.Firstly,this method slices the point cloud data,and then extracts wall line features through fine denoising and fitting straight lines using the RANSAC algorithm.Subsequently,the extracted wall line features are normalized to obtain the actual wall line.In addition,to support the construction of the IFC building information model,this study also proposes some common methods for extracting wall centerline.With these methods,the wall and centerline features of buildings can be more accurately extracted,providing effective technical support for building digital modeling.Studied how to determine the location of indoor stairs through slicing point clouds and segment the indoor staircase space.In order to achieve the segmentation of indoor stairs,a region growth algorithm was introduced.Through this algorithm,indoor stair data was segmented and extracted through vertical point cloud filtering.And studied the impact of different segmentation parameters on stair region segmentation.Based on the centroid parameters of the stair tread and riser data,obtain the actual stair tread and riser parameters,and compare them with the actual stair data to verify the feasibility of this method.Design experiments for analysis and verification.Based on the measured point cloud data of the rooms on the first and second floors of Building F of the School of Surveying and Mapping at Beijing University of Architecture(Daxing Campus),experiments and analyses were conducted on the methods in the second,third,and fourth chapters to verify the feasibility of these methods in practical applications.By conducting preprocessing and floor partitioning experiments on point clouds,the foundation has been laid for subsequent experiments on extracting walls,stairs,doors,and beams.In the experiment of extracting walls,stairs,doors,and beams from the processed point cloud,the experimental results were analyzed in detail,and the advantages and disadvantages of the experimental method were evaluated and analyzed,providing reference for subsequent research.