This preview shows page 1. Sign up to view the full content.
Unformatted text preview: s under the action of rigid motions and the Parzen’s method for the non-parametric estimation of Probability Density Functions from a finite number of sampled points. This paper provides an extensive experimental test on the original algorithms developed by Mario Costa by analyzing a number of cases which differ on size and surface classes. From a conceptual point of view, the non-parametric, model-independent estimation of the Probability Density Function of a set of points closes the loop composed of design, manufacturing and inspection activities along the product development process. It seems therefore possible to adopt the proposed methodology to provide an unambiguous description of product morphology. Keywords: geometrical invariant set, statistical analysis, shape recognition 1. GEOMETRICAL PRODUCT SPECIFICATIONS Since 1995 researchers involved in geometric dimensioning and tolerancing declared the need of parallel investigations on specification and verification activities performed on the geom...
View Full Document
- Spring '10
- The Land