Font Size: a A A

Research On Feature Extraction Method Of Cultural Relics Model Based On Multi-scale And Clustering

Posted on:2020-11-01Degree:MasterType:Thesis
Country:ChinaCandidate:L F ZhangFull Text:PDF
GTID:2415330590982223Subject:Computer technology
Abstract/Summary:PDF Full Text Request
In the field of cultural relics restoration,archaeologists often need to extract features from cultural relics models in the process of dealing with cultural relics fragments,in order to obtain shape information and lay the foundation for the subsequent splicing work.In this paper,aiming at the problem of feature redundancy and high time complexity in previous feature extraction methods,a feature extraction method based on multi-scale covariance matrix is proposed,and the feature line is fitted by Gauss mapping and K-Means.The method is applied to the feature extraction of terracotta warriors and horses fragments model,and related research is carried out.The research contents of this paper include:(1)Due to the large number of terracotta warriors and horses debris models,irregular shape and high noise.Previous feature extraction algorithms used in terracotta warriors and horses fragments feature extraction have high time complexity,poor extraction accuracy and poor anti-noise performance.To solve the above problems,this paper proposes a feature point extraction method based on multi-scale covariance matrix,and applies this method to feature point extraction of terracotta warriors and horses fragments model.Experiments verify the advantages of this method in extracting the number of feature points of terracotta warriors and horses model fragments and in anti-noise performance compared with traditional methods.(2)Aiming at the problems of low accuracy of fitting results and breakpoints of feature lines in the previous methods of feature line fitting,a feature line fitting method based on Gauss mapping and K-means clustering is proposed in this paper.Meanwhile,the moving least square method is used to smooth the feature curves to extract the feature curves of terracotta warriors and horses fragments model with better quality.The experimental results verify the proposed method.The advantages of this method in the process of feature line extraction.(3)The feature points and feature line extraction algorithm are applied to the actual work,and the feature extraction system of cultural relics model is designed and implemented.The system uses the terracotta warrior model as input to verify the effect of feature extraction in this method,and can display the extracted effect.
Keywords/Search Tags:Point cloud data, characteristic points, Gauss map, K-means clustering
PDF Full Text Request
Related items