GPU 기반 곡률 추정 알고리즘
GPU Based Algorithm for Estimating Curvatures
1
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
본 논문에서는 3차원 물체의 외형 특징이 되는 곡률을 추정해내는 GPU (Graphics Processing Unit) 기반 알고리즘을 제안한다. 곡률은 3차원 물체의 간략화, 메쉬 분할 등과 같은 기하학적 모델링 분야에서 매우 유용한 수치이다. 그러나 그래픽스 기술의 발달로 대용량의 3차원 물체들이 응용 프로그램에서 사용되고 데이터 용량에 비례하여 곡률을 추정하는 계산 시간도 증가하게 되었다. 만약 실시간으로 곡률을 추정해야 하는 응용 프로그램의 경우, 3차원 물체의 기하학 정보를 모두 메인 메모리에 올려놓고 CPU에서 계산하게 되면 심각한 속도 저하를 가져올 수 있다. 본 논문은 곡률 추정을 위한 계산 시간을 단축시키기 위해, CUDA (Compute Unified Device Architecture)를 사용하여 꼭지점마다 수행되는 주요 계산 과정을 GPU에서 병렬로 수행하도록 구현하였다. 기존의 곡률 계산 방법보다 훨씬 빠른 수행 결과를 보인다.
Abstract
This paper proposes the GPU (Graphics Processing Unit) based algorithm for estimating curvatures which is one of properties of the shape of 3D objects. In geometric modeling, curvatures are very useful metrics for mesh simplification, mesh segmentation, and so on. Nowadays the advanced techniques for computer graphics let many applications use huge amounts of graphic data, so that the computing time for estimating curvatures of such a huge graphic object becomes to increase. If an application which requires real-time estimation of curvatures will perform the computation in the CPU with the main memory where are all geometry information of a 3D object, it must make the system to slow down. In this paper, to decrease the computing time for estimating curvatures, some crucial computing parts for each vertex are implemented in the GPU using CUDA (Compute Unified Device Architecture). The results show that the proposed algorithm is definitely faster than previous methods.
참고문헌 (12)
- P. Heckbert, "Optimal Triangulation and Quadric-Based Surface Simplification", Journal of Computational Geometry: Theory and Applications, 14(1-3), pp.49~65 (1999)
- G. Gorla, "Texture Synthesis for 3D Shape Representation", IEEE Transactions on Visualization and Computer Graphics, 9(4), pp.512~524 (2003)
- D. DeCarlo, "Suggestive Contours for Conveying Shape", ACM Transactions on Graphics, 22(3), pp.848~855 (2003)
- T. Judd, "Apparent Ridges for Line Drawing", ACM Transactions on Graphics, 26(3), pp.19:1~19:7 (2007)
- F. Cazals, "Estimating Differential Quantities Using Polynomial Fiiting of Osculating Jets", In Proc. of Symposium on Geometry Processing, pp.177~187 (2003)
- J. Goldfeather, "A Novel Cubic-Order Algorithm for Approximating Principal Direction Vectors", ACM Transactions on Graphics, 23(1), pp.45~63 (2004)
- D. L. Page, "Robust Crease Detection and Curvature Estimation of Piecewise Smooth Surface from Triangle Mesh Approximations Using Normal Voting", In Proc. of Conference on Computer Vision and Pattern Recognition, pp.162~167 (2001)
- E. Hameiri, "Estimating the Principal Curvatures and the Darboux Frame from Real 3D Range Data", In Proc. of International Symposium on 3D Data Processing Visualization and Transmission(3DPVT), pp.258~ (2002)
- M. Meyer, "Discrete Differential-Geometry Operators for Triangulated 2-Manifolds, In Visualization and Mathematics III", pp.35~57 (2003), Springer-Verlag
- P. Alliez, "Aniosotropic Polygonal Remeshing", ACM Transactions on Graphics, 22(3), pp.485~493 (2003)
- S. Rusinkiewicz, "Estimating Curvatures and Their Derivatives on Triangle Meshes", In Proc. of 3D Data Processing, Visualization, & Transmission 2004, pp.486~493 (2004)
- W. Griffin, "GPU Curvature Estimation on Deformable Meshes", In Proc. of Interactive 3D Graphics and Games 2011, pp.159~166 (2011)