DocumentCode
2290605
Title
Content-based retrieval of 3D models in distributed Web databases by visual shape information
Author
Loffler, J.
Author_Institution
Nat. Res. Center for Inf. Technol., Sankt Augustin
fYear
2000
fDate
2000
Firstpage
82
Lastpage
87
Abstract
Digital documents assembled from different components, like text, images, video, audio and 3D models, are supported today by many distributed applications. This paper presents a new approach to retrieving 3D models as part of digital documents using content-based image retrieval functionality. To find a 3D model best-fitting the user´s query, shape information is extracted from the retrieved models and a score is generated by comparing shape images with the graphical user input. A distributed database environment with a WWW interface is presented and used as a generic implementation platform for this technique. The new retrieval method for 3D geometry is then described by extending the generic architecture in two steps, with a server-side module and an interactive client display. I outline how this method can be used for partial geometry downloading of 3D models in digital libraries
Keywords
client-server systems; content-based retrieval; distributed databases; geometry; image segmentation; information resources; interactive systems; multimedia databases; online front-ends; solid modelling; 3D geometry; 3D model retrieval; World Wide Web interface; content-based image retrieval; digital libraries; distributed Web databases; generic implementation platform; graphical user input; interactive client display; multimedia digital documents; partial geometry downloading; score generation; server-side module; shape image comparison; shape segmentation; user query matching; visual shape information; Assembly; Content based retrieval; Data mining; Distributed databases; Geometry; Image databases; Image retrieval; Information retrieval; Shape; Spatial databases;
fLanguage
English
Publisher
ieee
Conference_Titel
Information Visualization, 2000. Proceedings. IEEE International Conference on
Conference_Location
London
ISSN
1093-9547
Print_ISBN
0-7695-0743-3
Type
conf
DOI
10.1109/IV.2000.859741
Filename
859741
Link To Document