3D object reconstruction and representation using neural networks

ABSTRACT

3D object reconstruction is frequent used in various fields such as product design, engineering, medical and artistic applications. Numerous reconstruction techniques and software were introduced and developed. However, the purpose of this paper is to fully integrate an adaptive artificial neural network (ANN) based method in reconstructing and representing 3D objects. This study explores the ability of neural networks in learning through experience when reconstructing an object by estimating it's z-coordinate. Neural networks' capability in representing most classes of 3D objects used in computer graphics is also proven. Simple affined transformation is applied on different objects using this approach and compared with the real objects. The results show that neural network is a promising approach for reconstruction and representation of 3D objects.

REFERENCES

• 1: MICHAEL JUSCHKE. 1996. True Colour Holography. Deg Thesis, Department of Electrical and Electronic Engineering: University of Western Australia.
• 2: W. Niem , J. Wingbermuhle, Automatic reconstruction of 3D objects using a mobile monoscopic camera, Proceedings of the International Conference on Recent Advances in 3-D Digital Imaging and Modeling, p.173, May 12-15, 1997
• 3: VIT ZYKA, RADIM SARA. 1997. Scale Model Refinement For 3D Reconstruction. Center of Machine Perception: Czech Technical University.
• 4: RUIZ, OSCAR EDUARDO, CADAVID, CARLOS ALBERTO, MIGUEL. 2001. Evaluation of 2D Shape Likeness for Surface Reconstruction. XIII International Congress on Graphics Engineering.
• 5: Michael Negnevitsky, Artificial Intelligence: A Guide to Intelligent Systems, Addison-Wesley Longman Publishing Co., Inc., Boston, MA, 2001
• 6: YING WU. 2001. An Introduction to ECE510 Computer Vision. Electrical and Computer Engineering: Northwestern University, Evanston.
• 7: INGE SODERKVIST. 1999. Introductory Overview of Surface Reconstruction Methods." Department of Mathematics: Lulea University, Sweden.
• 8: EMMANOUIL PIPERAKIS, ITSUO KUMAZAWA. 2001. Affine Transformations of 3D Objects Represented with Neural Networks. 3-D Digital Imaging and Modeling, Proceedings. 213--223.
• 9: HOD LIPSON, MOSHE SHPITALNI. 2001. Correlation-Based Reconstruction of a 3D Object from a Single Freehand Sketch. Cornell University, University of Michigan, USA.
• 10: VOLKER KRUGER, GERALD SOMMER. 2000. Gabor Wavelet Networks for Object Representation and Face Recognition. Kiel, Germany.
• 11: Gary Bradski , Stephen Grossberg, Fast-learning VIEWNET architectures for recognizing three-dimensional objects from multiple two-dimensional views, Neural Networks, v.8 n.7-8, p.1053-1080, 1995 [doi>10.1016/0893-6080(95)00053-4]
• 12: YOSHIHIRO KAWAI, TOSHIO UESHIBA, TAKASHI YOSHIMI, MASAKI OSHIMA. 1992 Reconstruction of 3D Objects by Integration of Multiple Range Data. Japan.
• 13: JOSE GASPAR, ETIENNE GROSSMANN, JOSE SANTOS-VICTOR. 2001. INTERACTIVE RECONSTRUCTION FROM AN OMNIDIRECTIONAL IMAGE. 9th International Symposium on Intelligent Robotic Systems.
• 14: ANDERS HEYDEN. 1995. Geometry and Algebra of Multiple Projective Transformations. Dept of Mathematics, Lund University, Sweden.
• 15: MIKLOS HOFFMANN, LAJOS VARADY. 1998. Free-form Surfaces for Scattered Data by Neural Networks. Journal for Geometry and Graphics. Vol.2. 1--6.
• 16: JOARDER KAMRUZZAMAN, S. M. AZIZ. 2002. A Note on Activation Function in Multilayer Feedforward Learning. Monash University, University of South Australia, Australia.
• 17: CS. SZEPESVARI, A. LORINCZ. 1994. Self-Organized Learning of 3 Dimensions. Hungary.
• 18: TOHRU NITTA. 1994. Generalization Ability of the Three-Dimensional Back-Propagation Network. Ibaraki, Japan.
• 19: AC EVANS, NA THACKER, JEW MAYHEW. 1993. A Practical View Based 3D Object Recognition System. University of Sheffield.
• 20: RICHARD S. YOON, DON S. BORRETT, HON C. KWAN. 1995. Three-Dimensional Object Recognition Using a Recurrent Attractor Neural Network. IEEE-EMBC and CMBEC Theme 2: Imaging.

INDEX TERMS

Keywords:
affined transformation, back propagation, multilayer feed-forward neural networks, object space, reconstruction, representation, third order polynomial