3D Object Classification Using a Two-Dimensional Hidden Markov Model

Abstract:

In this paper, we develop a novel method of 3D object classification based on a Two-Dimensional Hidden Markov Model (2D HMM). Hidden Markov Models are a widely used methodology for sequential data modeling, of growing importance in the last years. In the proposed approach, each object is decomposed by a spiderweb model and a shape function D2 is computed for each bin. These feature vectors are then arranged in a sequential fashion to compose a sequence vector, which is used to train HMMs. In 2D HMM, we assume that feature vectors are statistically dependent on an underlying state process which has transition probabilities conditioning the states of two neighboring bins. Thus the dependency of two dimensions is reflected simultaneously. To classify an object, the maximized posteriori probability is calculated by a given model and the observed sequence of an unknown object. Comparing with 1D HMM, the 2D HMM gets more information from the neighboring bins. Analysis and experimental results show that the proposed approach performs better than existing ones in database.

• Periodical: Applied Mechanics and Materials (Volumes 411-414)
• Pages: 2041-2046
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.411-414.2041
• Online since: September 2013
• Authors: Jing Guo, Ming Quan Zhou, Chao Li, Zhe Shi
• Keywords: 3D Model Classification, Expectation Maximization (EM) Algorithm, Hidden Markov Model (HMM), Two-Dimensional Hidden Markov Model

References

[1] 3D model search engine, http: /shape. cs. princeton. edu.

[2] P. Min, J. A. Halderman, M. Kazhdan, T. A. Funkhouser, Early experiences with a 3D model search engine, Web 3D Symposium, 2003: 7–18.

DOI: 10.1145/636593.636595

[3] 3D model retrieval system, http: /3d. csie. ntu. edu. tw/˜dynamic.

[4] D.Y. Chen, X. - P. Tian, Y. -T. Shen, M. Ouhyoung, On visual similarity based 3D model retrieval, Computer Graphics Forum (EG 2003 Proceedings), 2003, 22(3).

DOI: 10.1111/1467-8659.00669

[5] 3D model similarity search engine, http: /merkur01. inf. uni-konstanz. de/CCCC.

[6] D. V. Vrani'c., An improvement of rotation invariant 3D shape descriptor based on functions on concentric spheres, In ICIP 2003, (2003).

[7] Csakany P, Wallace A M., Representation and classification of 3-D objects, IEEE Transactions on Systems, Man, and Cybenetics-Part B: Cybemetics, 2003, 33(4): 638-647.

DOI: 10.1109/tsmcb.2003.814302

[8] Huber D, Kapuria A, Donamukkala R R, Parts-based 3D object classification , Proceedings of the IEEE International Conference on Computer Vision and Pattern Recogniton, 2004, 82- 89.

DOI: 10.1109/cvpr.2004.1315148

[9] Osada, R. Funkhouser, T. Chazelle B, Shape distribution, ACM Transactions on Graphics, 2002, 21(4): 807-832.

DOI: 10.1145/571647.571648

[10] Vandeborre J P, A practical approach for 3D model indexing by combining local and global invariants, 3D Data Processing Visualization and Transmission, 2002: 644-647.

DOI: 10.1109/tdpvt.2002.1024132

[11] Liu yi , Wang xulei , Zha hong bin, Acta Scientiarum Naturalium Universitatis Pekinensis , 2009(6) 45: 967-968. (In Chinese).

[12] Lawrence R. Rabiner, A Tutorial on Hidden Markov Models and Selected Applications in Speech Recognition, Process of the IEEE, 1989, 7(2): 257–285.

DOI: 10.1109/5.18626

[13] Yujian Li et al, Hidden Markov models with states depending on observations, PatternRecognition Letters, 2004, 26 (2005) : 977–984.

DOI: 10.1016/j.patrec.2004.09.050

[14] Liu xiaoming, Yin jian wei, Feng zhilin , Dong Jinxiang, Journal of Zhejiang University ( Engineering Science) , 2006 , 40(8) : 1300-1305. (In Chinese).