逆向工程為一種由實體模型的三維量測資料重建CAD模型的技術，點資料處理則是逆向工程裡最重要的一環。因大量的點資料會提高CAD重建的困難度，因此，本研究的主要目的是發展一系列的網格資料特徵分離技術，依照工件的幾何特性將網格資料進行區域分離。對於二次曲面及自由曲面有不同的處理方式，針對簡單的二次曲面如平面、圓柱與圓錐等網格資料，發展一套完整的分離流程，只針對種子點附近的區域進行處理，以減少分離網格資料的時間；針對幾何型態較為複雜的自由曲面，則以多種基礎幾何之運算進行特徵辨識，並將特徵區域進行最佳化處理，再以曲率方向的運用，刪除不為封閉的特徵區域，以達到較佳的分離結果。

Reverse engineering is a technique to reconstruct the CAD model from the three-dimensional measurement points of a physical object. The pre-processing of the data points is most important because a large number of data points may lead difficulty to reconstruct the CAD model. The main objective of this research is to develop a series of feature segmentation technique for the triangular meshes in terms of the geometric characteristics of the model. There are different ways to deal with the quadratic surfaces and freeform surfaces. For quadratic surfaces such as plane, cylinder and cone, we develop a procedure for the segmentation of the triangular meshes corresponding to such features, and only operate in the nearby area of the seed point so as to reduce the time required in processing the data. For the freeform surfaces with complicated geometric characteristics, we apply many kinds of operations to recognize and optimize different feature regions from triangular meshes. Then we delete redundant vertices in each feature regions in terms of the curvature information so as to yield clean and un-noised segmentation results.

[1]G. Taubin, “Estimating the Tensor of Curvature of a Surface from a Polyhedral Approximation”, Proceedings of the 5th International Conference on Computer Vision , pp. 902-907, 1995.
[2]P. Krsek, G. Lukacs and R. R. Martin,“Algorithms for Computing Curvatures from Range Data”, Mathematics of Surfaces VIII, pp. 1-16, 1998.
[3]J. A. Thorpe, Elementary Topics in Differential Geometry, Springer-Verlag, 1978.
[4]T. Surazhsky, E. Magid, O. Soldea, G. Elber, E. Rivlin, “A Comparison of Gaussion and Mean Curvatures Estimation Methods on Triangular Meshes”, Proceedings of The IEEE International Conference on Robotics and Automation, 2003.
[5]A. F. Koschan, D. L. Page and M. A. Abidi, “Perception-based 3D Triangle Mesh Segmentation Using Fast Marching Watersheds”, 2003 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, Vol. 2, pp. 27-32, 2003.
[6]蔡耀震, 賴景義, 鍾永彬, “網格資料特徵分離”, 第十三屆自動化科技研討會論文集, 1998.
[7]J. Huang and C.H. Menq, “Automatic Data Segmentation for Geometric Feature Extraction from Unorganized 3-D Coordinate Points”, IEEE Transactions on Robotics and Automation, Vol. 17, No. 3, 2001.
[8]G. Lavoue, F. Dupont and A. Baskurt, “Curvature Tensor Based Triangle Mesh Segmentation with Boundary Rectification”, Proceedings of Computer Graphics International, pp. 10-17, 2004.
[9]A. Razdan and M.S. Bae, “A Hybrid Approach to Feature Segmentation of Triangle Meshes”, Computer-Aided Design, Vol. 35, pp. 783-789, 2003.
[10]H. C. Kim, S. M. Hur and S. H. Lee, “Segmentation of the Measured Point Data in Reverse Engineering. International Journal of Advanced Manufacturing Technology, Vol. 20, pp. 571-580, 2002.
[11]M. Vieira and K. Shimada, “Surface Mesh Segmentation and Smooth Surface Extraction Through Region Growing”, Computer Aided Geometric Design, Vol. 22, pp. 771-792, 2005.
[12]N. Gelfand and L.J. Guibas, “Shape Segmentation Using Local Slippage Analysis”, Eurographics Symposium on Geometry Processing, 2004.
[13]張義宏, “光學掃描量測資料之二次曲面特徵分離”, 國立中央大 學機械工程研究所碩士論文, 2006.
[14]P. Benko and T. Varady, “Segmenting Large Point Clouds in Reverse Engineering Conventional Engineering Objects”, Computer and Automation Research Institute, Budapest.
[15]P. Benko, R. R. Martin and T. Varady, “Algorithms for Reverse Engineering Boundary Representation Models”, Computer-Aided Design, Vol. 33, No. 11, pp. 839-851, 2001.
[16]M. Vanco, “A Direct Approach for the Segmentation of Unorganized Points and Recognition of Simple Algebraic Surfaces”, Ph.D. Dissertation, University of Technology Chemnitz, 2002.
[17]翁文德,逆向工程之曲面模型重建技術發展, 國立中央大學機械 工程研究所博士論文, 1999.
[18]A. Hubeli and M. Gross, “Multiresolution Feature Extraction for Unstructured Meshes”, Proceedings of the Conference on Visualization ''01, pp. 287-294, 2001.