A DOE Approach for Sensitivity Analysis of a Shape Partitioning Algorithm

In the field of geometrical product specification and verification, one of the main problems is classification and segmentation of 3D shapes. Shape recognition and segmentation is a widespread research area with different application fields (image processing, shape searching, pattern recognition, reverse engineering, etc.). Many methodologies and algorithms have been developed within such different fields, each one exhibiting optimized performances with respect to the set of objects and targets in each application [1, 11, 12]. Nevertheless, for manufactured parts a unique description of shape during the whole product lifecycle is still envisaged, and GPS (“Geometrical Product Specification and Verification”) project seems to be the most promising approach, but it should be stated that the partitioning process is still to be improved both theoretically and operationally. The ISO Technical Committee 213 (TC213), entrusted to develop the GPS project, founded the partitioning process on the classification of shapes based on symmetrical properties of surfaces [5, 6]. The aim of this paper is to describe the method proposed by Gelfand and Guibas [4] and analyze its performances on sampled surfaces by varying parameters of the method that basically affect its efficiency. In fact, the ISO research is currently devoted to identify a segmentation method characterized by efficiency, reliability, robustness and applicability with the aim to standardize the methodology for the verification phase of the manufacturing process. In this paper, a DOE analysis has been performed, in order to search an optimal parameter configuration, necessary to consider the method as a standard for shape partitioning.

An Engineering Shape Benchmark for 3D Models

Three dimensional shape searching is a problem of current interest in several different fields, especially in the mechanical engineering domain. There has been a large body of work in developing representations for 3D shapes. However, there has been limited work done in developing domain dependent benchmark databases for 3D shape searching. In this paper, we propose a benchmark database for evaluating shape based search methods relevant to the mechanical engineering domain. Twelve feature vector based representations are compared using the benchmark database. The main contributions of this paper are development of an engineering shape benchmark and an understanding of the effectiveness of different shape representations for classes of engineering parts.