Classifying 3D Models Based on Transcending Local Features
Classifying 3D models is an essential task for 3D industry as it organizes objects according to categories, which helps searching 3D models perform more quickly. Many features and a suitable classifier have been used to improve classifying. However, it takes a long time for both extracting features and classifying 3D models. In this paper, a small set of transcending local features which are maximum distances from points in local regions to the center of 3D model is proposed. Then a Support Vector Machine is selected to classify 3D models based on data type of features and advantages of Support Vector Machine. The experiments are conducted on benchmark databases in Shape Retrieval Contest 2010. The results show that the approach employed to classify 3D models in an acceptable responding time of real applications is effective.
KeywordsClassifying 3D model Extract feature Support vector machine 3D model Computer vision
This work has received support from the project “T2019-07-09” funded by Thai Nguyen University of Information Technology and Communication, Vietnam.
- 1.Cao, Z., Huang, Q., Karthik, R.: 3D object classification via spherical projections. In: International Conference on 3D Vision, pp. 566–574 (2017)Google Scholar
- 7.Kassimi, A., Beqqali, O.E.: 3D model classification and retrieval based on semantic and ontology. Int. J. Comput. Sci. 8(5), 108 (2011)Google Scholar
- 8.Li, Z., Wang, D., Li, B., Zhong, L.: 3D model classification using salient features for content representation. In: Sixth International Conference on Natural Computation, vol. 7, pp. 3541–3545 (2010)Google Scholar
- 9.He, Y., Tang, Y.Y.: Classification of 3D models for the 3D animation environments. In: IEEE International Conference on Systems, Man and Cybernetics, pp. 3786–3791 (2009)Google Scholar