| Cultural relics are significant carriers of cultural heritage.However,affected by the natural environment and human activities,the cultural relics are damaged by varying degrees.The restoration of the cultural relics has been the imperative task of the restoration bases.Traditional restoration of cultural relics is performed by manual or clay molding.However,these contact restoration methods will inevitably bring incalculable losses on cultural relics.With the development of three-dimensional digitization technology,non-destructive and efficient digital restoration methods have entered into the sight of cultural relics scholars.Reverse engineering and rapid prototyping have provided technical support for the digital restoration of cultural relics.In the reverse engineering,point clouds simplification is employed to reduce the compression of point cloud data and to improve the efficiency of data processing.However,existing studies often ignore the extraction of feature points,which results in the loss of some feature points.In addition,the presence of holes in the surface leads to the more loss of the model’s feature points,and seriously undermines the integrity of the model.Therefore,this work mainly focuses on the extraction of feature point and the hole-filling of point cloud model.Based on this,the digital restoration of cultural relics is realized.The main research of this work is as follows:(1)An algorithm for detection of feature points based on the inverse distance weight and the density is proposed.The algorithm is based on the idea of threshold detection,and can be applied to the detection of feature points of various types.The inverse distance weight and the density are calculated according to the initial distribution of point clouds.The weighted equivalent force is evaluated based on inverse distance weights.The comprehensive detection parameter consists of the weighted equivalent force and the density.Using the comprehensive detection parameter,the target point is determined whether it is a feature point.After extracting the feature points,the average curvature is utilized to divide the non-feature points,and then to set the different regions with different reduction rates,finally fulfilled the point clouds simplification.The experimental results show that,feature point extraction algorithm that integrates with the point cloud simplification can effectively retain the original features of the model while eliminating some redundant points.(2)A hole-filling algorithm based on improved advancing front method is proposed.Firstly,the boundaries of the hole are identified,and then the preprocessing of these boundaries is performed.The insertions of vertices and triangles are completed by the improved advancing front method.These new vertices are checked to avoid the meshes intersecting.The Delaunay criterion is adopted to test these new triangles in the same plane.Lastly,the new vertices are adjusted according to the normal vector and the curvature information,and the closed holes are repaired effectively.Considering the differences between the island hole and the closed hole,these hole boundaries and these island boundaries are employed as the advancing front.A multi-directional advancing method that suitable for the filling of the island holes is proposed.(3)Taking the ancient bronze sword as an example,the effectiveness of the proposed algorithm is verified,the entire process of digital cultural relics restoration is fulfilled.Firstly,the point cloud data of cultural relics are obtained by the Artec Space Spider 3D scanner.Then,denoising and registration of point cloud is accomplished by the reserve software Geomagic Studio.The simplification and hole-filling of the point cloud is completed by employing the method mentioned above,and then the Stereolithography(STL)grid model of the ancient bronze sword is exported.Finally,Magics is employed to slice the sword STL model,Stereo Lithography Apparatus(SLA)technology is chosen to achieve the processing of the cultural relics. |
PDF Full Text Request