온라인을 기반한 제품의 구매가 활성화 됨에 따라, 소비자들의 제품 디자인에 대한 사실적이고 정확한 정보를 요구하고 있다. 본 연구에서는 제품의 형상정보(3D mesh)와 색/질감정보(image)를 이용하여 텍스쳐 맵핑된 실사적 3차원 모델을 생성하는 효율적인 방법을 제안한다. 3 차원 형상정보에 대응하는 이미지 상의 텍스쳐 좌표 관계를 구하기 위해 오브젝트 좌표계와 카메라 좌표계 사이의 변환행렬, 카메라의 초점거리, 카메라 CCD 와 프레임상의 이미지 사이의 종횡비를 파라미터로 하는 2D-3D 정합을 수행한다. 이러한 2D-3D 정합에 있어 발생하는 연산의 복잡도와 비선형도를 낮추기 위하여, 카메라 내부파라미터 검정단계, 임의의 회전행렬에 대한 임시적 최적 이동 벡터 (TOTV), 회전행렬에 대한 비선형 최적화 단계로 접근한다. 제안하는 방법의 유용성을 시현하기 위해 3차원 컬러 측정기로는 색외관을 획득하기 힘든 메탈릭 페인트 재질로 이루어진 제품에 적용한 구현결과를 보인다.
온라인을 기반한 제품의 구매가 활성화 됨에 따라, 소비자들의 제품 디자인에 대한 사실적이고 정확한 정보를 요구하고 있다. 본 연구에서는 제품의 형상정보(3D mesh)와 색/질감정보(image)를 이용하여 텍스쳐 맵핑된 실사적 3차원 모델을 생성하는 효율적인 방법을 제안한다. 3 차원 형상정보에 대응하는 이미지 상의 텍스쳐 좌표 관계를 구하기 위해 오브젝트 좌표계와 카메라 좌표계 사이의 변환행렬, 카메라의 초점거리, 카메라 CCD 와 프레임상의 이미지 사이의 종횡비를 파라미터로 하는 2D-3D 정합을 수행한다. 이러한 2D-3D 정합에 있어 발생하는 연산의 복잡도와 비선형도를 낮추기 위하여, 카메라 내부파라미터 검정단계, 임의의 회전행렬에 대한 임시적 최적 이동 벡터 (TOTV), 회전행렬에 대한 비선형 최적화 단계로 접근한다. 제안하는 방법의 유용성을 시현하기 위해 3차원 컬러 측정기로는 색외관을 획득하기 힘든 메탈릭 페인트 재질로 이루어진 제품에 적용한 구현결과를 보인다.
As on-line purchases is activated, customers' demand increases for the realistic and accurate digital information of a product design. In this paper, we propose a practical method that can generate a realistic 3D model of a real product using a 3D geometry obtained by a 3D scanner and its photograph...
As on-line purchases is activated, customers' demand increases for the realistic and accurate digital information of a product design. In this paper, we propose a practical method that can generate a realistic 3D model of a real product using a 3D geometry obtained by a 3D scanner and its photographic images. In order to register images to the 3D geometry, the camera focal length, the CCD scanning aspect ratio and the transformation matrix between the camera coordinate and the 3D object coordinate must be determined. To perform this 2D-3D registration with consideration of computational complexity, a three-step method is applied, which consists of camera calibration, determination of a temporary optimum translation vector (TOTV) and nonlinear optimization for three rotational angles. A case study for a metallic coated industrial part, of which the colour appearance is hardly obtained by a 3D colour scanner has performed to demonstrate the effectiveness of the proposed method.
As on-line purchases is activated, customers' demand increases for the realistic and accurate digital information of a product design. In this paper, we propose a practical method that can generate a realistic 3D model of a real product using a 3D geometry obtained by a 3D scanner and its photographic images. In order to register images to the 3D geometry, the camera focal length, the CCD scanning aspect ratio and the transformation matrix between the camera coordinate and the 3D object coordinate must be determined. To perform this 2D-3D registration with consideration of computational complexity, a three-step method is applied, which consists of camera calibration, determination of a temporary optimum translation vector (TOTV) and nonlinear optimization for three rotational angles. A case study for a metallic coated industrial part, of which the colour appearance is hardly obtained by a 3D colour scanner has performed to demonstrate the effectiveness of the proposed method.
* AI 자동 식별 결과로 적합하지 않은 문장이 있을 수 있으니, 이용에 유의하시기 바랍니다.
제안 방법
1. A practical method of 2D-3D registration is developed for generating a realistic 3D model of a real product using 3D geometry obtained by a 3D scanner and its photographic images. The 2D-3D registration is robustly performed by three step approach, which consists of camera calibration, determination of the temporary optimum translation vector and nonlinear optimization for three rotational angles.
A MP3 player made of shining materials is chosen for a demonstration of the proposed method. 3D points data of the MP3 player is obtained using the 3D scanner of Optoscan HE-100 developed by Breuckman Inc.
In order to connect 3D geometry and 2D image information, it is required to know the mapping relationship between a given point on the surface and its projected image point in the photographic view. For this work, CCD scanning aspect ratio, the camera focal length and the transformation matrix between the image coordinate and the 3D object coordinate must be determined.
In this paper, we present a practical method of 2D-3D registration that can generate a texture mapped 3D model of a real object. To reduce non-linearity and complexity of 2D-3D registration, a practical approach that separately solves intrinsic parameters and extrinsic parameters of the camera is applied.
성능/효과
2. The effectiveness of the our method is demonstrated by applying to a MP3 Player, of which the colour appearance data is hardly obtained by a 3D colour scanner. Test result shows that our method can be applied with acceptable level of errors for the registration.
후속연구
3. Our further studies will consider multipe-view to generate a full realistic 3D model.
참고문헌 (11)
S. S. Suh, C. C. Lee "The Emotional Sensibility Estimation System for Front-load Washer", J. of the Korea Academia-Industrial cooperation Society, Vol. 11, No.3, pp. 821-826, 2010. DOI: http://dx.doi.org/10.5762/KAIS.2010.11.3.821
B. Guenter, C. Grimm, D. Wood, H. Malvar, F. Pighin, "Making Faces", Proc. of 25th Conference on Computer Graphics and Interactive Techniques, pp. 55-66, July, 1998.
P. E. Debevec, C. J. Taylor, J. Malik, "Modeling and Rendering Architecture from Photographs: A Hybrid Geometry-and Image-based Approach", Proc. of 23th Conference on Computer Graphics and Interactive Techniques, pp. 11-20, August, 1996.
R. Y. Tsai, "A Versatile Camera Calibration Technique for High-Accuracy 3D Machine Vision Metrology Using Off-the-Shelf TV Cameras and Lenses", IEEE Journal of Robotics and Automation, Vol. RA-3, No. 4, pp. 221-244, August, 1987.
R. Hou, K. S. Jeong "3D Reconstruction and Self-calibration based on Binocular Stereo Vision", J. of the Korea Academia-Industrial cooperation Society, Vol. 13, No.9, pp. 3856-3863, 2012. DOI: http://dx.doi.org/10.5762/KAIS.2012.13.9.3856
D. G. Lowe, "Fitting Parameterized Three-Dimensional Models to Images", IEEE Transaction on Pattern Analysis and Machine Intelligence, Vol. 13, No.5, pp. 441-450, May, 1991. DOI: http://dx.doi.org/10.1109/34.134043
H. P. A. Lench, W. Heidrich, H. P. Seidel, "Automated Texture Registration and Stitching for Real World Models", Proc. of the 8th Pacific Conference on Computer Graphics and Application, pp. 317-452, October, 2000. DOI: http://dx.doi.org/10.1109/PCCGA.2000.883955
K. Matsushita, T. Kaneko, "Efficient and Handy Texture Mapping on 3D Surfaces", Computer Graphics Forum, Vol. 18, No. 3, pp. 349-358, 1999. DOI: http://dx.doi.org/10.1111/1467-8659.00355
T. Franken, M. Dellepiane, F. Ganovelli, P. Cignoni, C. Montani, R. Scopigno, "Minimizing User Intervention in Registering 2D Images to 3D Models", The Visual Computer, Vol. 21, No. 8-10, pp. 619-628, 2005. DOI: http://dx.doi.org/10.1007/s00371-005-0309-z
M. Corsini, M. Dellepiane, F. Ponchio, R. Scopigno, "Image-to-Geometry Registration: a mutual Information Method exploiting Illumination-related Geometric Properties", Computer Graphics Forum, Vol. 28, No. 7, pp. 1755-1764, 2009. DOI: http://dx.doi.org/10.1111/j.1467-8659.2009.01552.x
W. H. Press, S. A. Teukolsky, W. T. Vetterling, B. P. Flannery, "Numerical Recipes: The Art of Scientific Computing", 3rd edition Cambridge University Press, 2007.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.