Combining shape and color for retrieval of 3D models
College
College of Computer Studies
Department/Unit
Software Technology
Document Type
Conference Proceeding
Source Title
NCM 2009 - 5th International Joint Conference on INC, IMS, and IDC
First Page
1295
Last Page
1300
Publication Date
12-1-2009
Abstract
Current research on the retrieval systems for 3D models focuses on using the shape of the models to facilitate search and retrieval. This paper explores the possibility of augmenting the existing 3D shape-based similarity measures by combining shape and color. First, a new descriptor was developed based on the D2 shape descriptor. In our method, N pairs of faces are randomly chosen from a 3D model, with probability proportional to the area of the face. The ratio of the smaller area over the larger area is computed and its frequency stored, generating a frequency distribution of N ratios which is stored as the second dimension of a 2D array, while the first dimension contains the frequency distribution of distances of randomly generated point pairs (the D2 distribution). Second, this research introduces the use of the color features of a 3D model in combination with the shape features to determine similarity. The research involves the study and adoption of an existing 2D color-based similarity measure for 3D models. The analysis of the results is based on the precision and recall of both approaches. © 2009 IEEE.
html
Digitial Object Identifier (DOI)
10.1109/NCM.2009.140
Recommended Citation
Ruiz, C. R., Cabredo, R., Monteverde, L., & Huang, Z. (2009). Combining shape and color for retrieval of 3D models. NCM 2009 - 5th International Joint Conference on INC, IMS, and IDC, 1295-1300. https://doi.org/10.1109/NCM.2009.140
Disciplines
Computer Sciences | Software Engineering
Keywords
Three-dimensional modeling
Upload File
wf_yes