Research On Virtual Geometry Restoration Of Historical Building Point Cloud Model

At present,the protection of historical buildings is imminent.The Party Central Committee and the State Council have attached great importance to and issued a number of policies to support the protection planning and repair of immovable cultural relics,and to improve the protection of historical and cultural architectural heritage.With the development of 3D Li DAR scanning technology and photogrammetry technology,the role of 3D point cloud data and image data in the virtual restoration of historical buildings has become more prominent.However,how to realize the virtual geometric restoration of historical building point cloud models based on different data sources is a meaningful and challenging research topic.This paper combines UAV photogrammetry technology and 3D Li DAR scanning technology to obtain complete historical building point cloud data,through image dense point cloud and ground Li DAR point cloud data registration,and based on the context of missing point cloud completion method to achieve historical building points Virtual geometry repair of cloud model.The main research contents and results are as follows:(1)UAV image and ground Li DAR point cloud data processingUse Photoscan to perform aerial three-dimensional processing on image data to generate dense image point clouds.Aiming at the characteristics of the studied ancient building ground Li DAR point clouds,the station coordinate stitching method is used to carry out the ground Li DAR point cloud data registration,and the statistical outlier removal algorithm is used.Point cloud filtering and denoising,and using an octree-based algorithm to compress the point cloud data.After processing,the point cloud data provides high-quality data support for subsequent experiments.(2)Scale-ICP point cloud matching based on SVDA SVD-based Scale-ICP point cloud matching algorithm is used to realize the registration of dense point clouds of images and ground Li DAR point cloud data.Compared with the traditional ICP algorithm,the registration accuracy of this algorithm is improved.The scaling and registration between the two reflect the advantages of the algorithm.(3)Edge detection based on improved PCA normal vector estimationEdge detection is an important task of point cloud completion.Aiming at the problem of the ambiguity of the normal vector direction of the point cloud in the traditional principal component analysis boundary extraction algorithm and the poor detection effect,an improved PCA normal vector estimation edge detection algorithm is proposed.,Improve the accuracy and anti-noise of the edge inspection results,and complete the detection and judgment of the inner and outer boundaries of the point cloud data.(4)Context-based point cloud completion of missing partsAiming at the traditional context-based point cloud completion,this paper proposes a point cloud completion method of segmentation and then block-based iterative completion,which improves the efficiency of point cloud completion by searching for the optimal block in a local context.The ancient building is an example to realize the point cloud completion and analysis and research of the completion result.The results show that the completion method in this paper can better repair the point cloud geometric model of the ancient building.