Research On Filtering Processing And Reconstruction Of Point Cloud Of Hui-style Building Components

As a treasure with a long history and rich cultural rainfall,the unique charm of Huizhou architecture has attracted an increasing number of scholars to dedicate themselves to the protection of ancient buildings and to the development of culture.By establishing a threedimensional model through three-dimensional digital means,people are able to visit the cloud without actually visiting the site,which is conducive to the further development of the protection of cultural relics and the promotion of China’s traditional culture.However,the three-dimensional modelling of ancient buildings is inseparable from the representation of the individual components within the building.Following the field investigation of Hui-style architecture and observation of its three dimensional model,finally,this paper selects the three main architectural components of Hui-style architecture as the objects of research: the drum stone,the figure-eight wall,and the gateway.First of all,Based on the data collection of Huizhou architecture components,this paper presents the process of 3D scanning,establishing 3D point cloud models of building components,pre-processing point clouds,and reconstructing the surface of point clouds.The purpose of this paper is to investigate how to solve the problems of surface filtering and reconstruction in point cloud processing in Huizhou architecture.Secondly,For point cloud filtering,in order to solve the problem of a large number of scattered noise and block noise next to the 3D point cloud data caused by 3D scanning,and improve the problem of large calculation and low accuracy of the traditional 3D space filtering algorithm,we improve the image processing-based filtering algorithm,propose a filtering algorithm based on finding the maximal connecting domain of the point cloud map image,a detailed analysis of the implementation process of the proposed algorithm is presented,and the point cloud data of Huizhou architecture components in Yiren hall are tested in order to verify the efficacy of the algorithm.It achieves an accuracy of 99.381%,providing a benchmark for point cloud filtering of Huizhou architecture components.Finally,For research on three-dimensional point cloud surface reconstruction of Huizhou architecture elements,to address the problem of poor reconstruction accuracy due to hole generation during reconstruction,this paper adopts surface reconstruction based on Alpha Shape algorithm,and adds KD-tree algorithm,according to the average point spacing of point clouds,proposes an Alpha Shape algorithm with automatic threshold update,analyzes the implementation process of the proposed algorithm in detail,and verifies the effectiveness of the proposed algorithm through experimental results.In order to improve the accuracy of surface reconstruction of the building components in Yiren Hall,which offers a guiding role in establishing 3D models of cultural relics.Figure[36] table[5] reference[69]...