Font Size: a A A

Sketch-based3d Model Retrieval

Posted on:2014-05-21Degree:MasterType:Thesis
Country:ChinaCandidate:Y LinFull Text:PDF
GTID:2268330395989037Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
Nowadays along with the rapid development of computer graphics, large numbers of3d models are applied in animation,3d games, advertisement, etc. A lot of public3d model databases are already appeared. How to make it fast and cheap to let users accurately locate the objective3d model, and construct the virtual scene becomes a hot topic in computer graphics field. The method which can try to achieve this goal is sketch-based3d model retrieval. Sketch-based3d model retrieval method uses a sketch to retrieval the objective3d model, and receives extensive attention due to the advantage of convenience, low-lost and good practicability.This paper works mainly on the problem of sketch-based3d model retrieval with feature matching method. As a result we integrate a set of scheme on sketch-based3d model retrieval with feature matching method. In this paper, we have finished three tasks. Firstly, against the problem that occluding contour is lack of the object’s details, we add crease line to show the details of object to improve the ability of expression. Secondly, in order to research the retrieval results of3d model retrieval with feature matching method, we analyze the retrieval results of several image features. Thirdly, previous papers on sketch-based3d model retrieval only propose a new feature and declare their retrieval result is better than others. But these features are very sensitive with parameter settings and have limited scope. We propose a multi-feature fusion matching method based on the existing features. And this method can reduce the dependency of parameter setting.In the last part of this paper we integrate our sketch-based3d model retrieval method to scene construction system and unstructured database system, and they have been used in projects.
Keywords/Search Tags:3d model retrieval, sketch-based retrieval, multi-feature fusion matching, feature matching
PDF Full Text Request
Related items