DocumentCode
2726547
Title
The Research of 3D Model Retrieval Algorithms Based on Normal Plane
Author
Hongtao Xie ; Fuhua Shang ; Yadong Zhou ; Xincheng Gao
Author_Institution
Comput. & Inf. Technol. Inst., Northeast Pet. Univ., Daqing, China
fYear
2012
fDate
6-8 Nov. 2012
Firstpage
207
Lastpage
210
Abstract
According to the characteristics of the model number is relatively small, similarity, complexity is not high in the small professional model library, and the existing 3D model retrieval method can not effectively improve the efficiency. This paper puts forward the retrieval algorithms based on normal plane, extract normal vector of each triangular facet in models, in the surrounded icosahedron of model set every surface as the reference surface, calculating the normal number which go through each reference surface and the ratio of the face in line accounts for the whole area, using the Needleman/Wunsch algorithm in the dynamic programming calculating the similarity for matching. Experiments show that in the small professional model library the model retrieval algorithm based on normal plane can greatly improve the efficiency of the model retrieval.
Keywords
dynamic programming; feature extraction; image matching; image retrieval; solid modelling; 3D model retrieval algorithm; Needleman-Wunsch algorithm; dynamic programming; icosahedron; matching similarity; model number; normal plane; normal vector extraction; reference surface; small professional model library; triangular facet; Analytical models; Computational modeling; Feature extraction; Heuristic algorithms; Libraries; Solid modeling; Vectors; 3D model library; model retrieval; normal plane; normal vector;
fLanguage
English
Publisher
ieee
Conference_Titel
Software Engineering (WCSE), 2012 Third World Congress on
Conference_Location
Wuhan
Print_ISBN
978-1-4673-4546-0
Type
conf
DOI
10.1109/WCSE.2012.50
Filename
6394955
Link To Document