Extraction of Inertial Parameters Based on 3D-Voxel Model of Equal Volume

Abstract:

Because of 3D model in continuous quality can be expressed by a lot of voxels with same size and small enough dimensions, a detection method of inertial parameters based on principal direction of voxel model is proposed. The method first searches three-spindle for voxel model by employing famous principal component analysis (PCA) method and establishes a new coordinates system by three-spindle above. Next, inertial parameters curve of the model can be obtained by calculating principal moment of inertia before translating the origin of the new coordinate system to the minimum point of the model and center of mass after that. The curve gives a geometric descriptor of the mode. Storing and matching the descriptor, the method calculates similarity between query model and database model, and then displays consequence according to the similarity from good to bad. Experiences show that the inertial parameters can be used as a kind of feature descriptor on the equal volume model and can be calculated easily.

• Periodical: Applied Mechanics and Materials (Volume 733)
• Pages: 889-892
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.733.889
• Online since: February 2015
• Authors: Ning Sun, Jin Min Wei
• Keywords: Equal Volume, Inertial Parameters, Similarity Measure, Voxel Model

References

[1] Liu Wei. Feature extraction and retrieval of 3D model [D]. Shanghai Jiao Tong University, Shanghai. (2008).

[2] Wang Jiale. The key techniques of 3D model retrieval based on the content [D]. Shanghai Jiao Tong University, Shanghai. (2008).

[3] Wang Jin, Sun Ning. Measurement of the 9 inertia parameters for the rigid body based on moment of inertia [J]. Acta Metrologica Sinica, 2012, 33(2): pp.135-139.

[4] D. V. Vranic, D. Saupe, J. Richter. Multimedia Signal Processing, 2001 IEEE Fourth Workshop on, October3-52001[C]. Cannes: IEEE, (2001).

DOI: 10.1109/mmsp.2001.962749