Research On Scan Point Clouds Data Preprocessing Techniques Based On Reverse Engineering

The rapid development of the three-dimensional laser scanning technology has solved the difficulties in the traditional sense of modeling modeling.The three-dimensional point cloud data obtained by using the three-dimensional laser scanner includes a large amount of information.The information includes not only the three-dimensional size information of the target object but also the geodetic coordinates of the position points and other physical information such as color texture,and can be digitally stored.With the advancement of three-dimensional measurement technology and the expansion of the scanning measurement range,the acquired point cloud data has assumed a massive trend.However,there are many technical difficulties in the collection and processing of these massive point cloud data.The first is the selection of acquisition methods.The second problem is the processing of collecting point cloud data.These problems include the pre-processing of point cloud data and data noise reduction,reduction,segmentation,and data visualization.This paper takes a boiler room system as the research object,uses a three-dimensional laser scanner to collect the point cloud data,takes into account the layout of the scene of the boiler room and the complexity of the equipment,and uses the multi station layout scanning mode based on the target target to scan the scene and collect the original scene data.In the preliminary processing of the original point cloud data,The supporting software FARO SCENE is used for pre-processing,including point cloud generation,target-based multi-site cloud data splicing,and point cloud color information addition.In the multi-site cloud data mosaic of the boiler room based on target,the method of automatic target recognition and manual target recognition was used for multi-station data splicing.Finally,the boiler room system point cloud data that meet the needs of follow-up research was obtained.In the preprocessing stage of point cloud data,the problem of denoising,streamlining,and image segmentation of point cloud data needs to be solved: For the noise reduction of point cloud data,several common noise reduction algorithms were analyzed,and the interactive noise reduction of point cloud data was completed using FARO SCENE software.In the compression of scanned data,the commonly used scanning data reduction algorithm is introduced.Completion of point cloud datareduction and analysis based on the bounding box method,and using Geomagic software to complete interactive streamlining of point cloud data and streamlined precision comparison analysis.In terms of visualization of point cloud data,several common spatial index structures of massive point cloud data are introduced,organization and management methods of massive point cloud data,and key technologies for visualization of three-dimensional spatial data models.Combined with MeshLab system,the boiler room cloud data visualization was completed.This paper uses the cloud data of the boiler room system as the main line,and conducts data collection,data processing and data visualization respectively,and puts forward many constructive research methods and ideas.It has certain reference value and practical significance.