A New Descriptor of Slices Histogram from 3D Shape Models

3-dimensional (3D) model retrieval is one of an important research fields to retrieve a query for models with matching shape from the database. This paper proposes a new shape descriptor using histograms of 2D slice images to cut in the x-, y-, and z-coordinates, respectively for solving the similarity measure from 3D models. Our approach is to compute the slices of 3D model for the x-, y-, and z-coordinates, respectively for cutting three axes of a model and to set by the principal axis based on the principal component analysis (PCA). Experimental results present the performances of the proposed method are better than those of the previous, and demonstrate our 3D retrieval system.