%0 Journal Article
%T Hybrid Functional-Neural Approach for Surface Reconstruction
%A Andr¨¦s Iglesias
%A Akemi G¨˘lvez
%J Mathematical Problems in Engineering
%D 2014
%I Hindawi Publishing Corporation
%R 10.1155/2014/351648
%X This paper introduces a new hybrid functional-neural approach for surface reconstruction. Our approach is based on the combination of two powerful artificial intelligence paradigms: on one hand, we apply the popular Kohonen neural network to address the data parameterization problem. On the other hand, we introduce a new functional network, called NURBS functional network, whose topology is aimed at reproducing faithfully the functional structure of the NURBS surfaces. These neural and functional networks are applied in an iterative fashion for further surface refinement. The hybridization of these two networks provides us with a powerful computational approach to obtain a NURBS fitting surface to a set of irregularly sampled noisy data points within a prescribed error threshold. The method has been applied to two illustrative examples. The experimental results confirm the good performance of our approach. 1. Introduction Manufacturing industries are constantly evolving in response to the new challenges of the globalization and the growing competition in this global market. Product design is playing a central role in this process, as current customers are increasingly demanding a mass customization of the products. As a result, the geometric and aesthetic properties of the manufactured goods (shape, color, and dimensions) have to be modified frequently in order to meet the new market demands. A major step in this process is the generation of real prototypes with different materials to explore and analyze their geometric properties and the feedback of potential customers when exposed to different variations of the final product. Prototype generation and customization can be dramatically improved by using digital technologies, in which the physical model is digitized, stored, and manipulated by computer, a process called reverse engineering [1, 2]. Typically, this process begins with data sampling by using 3D laser scanning and other digitizing devices. This technology is intensively used for the construction of car bodies, ship hulls, airplane fuselage, and other free-form objects [2¨C7]. The resulting data points are then fitted to mathematical entities such as curves and surfaces, usually in parametric form. The output is a very accurate digital version of the real product, which is also simpler and easier to store, analyze, and manipulate. It also simplifies the transfer and communication processes among designers, manufacturers, and providers, making the model available in just a few seconds all over the world, a key aspect in our ubiquitously
%U http://www.hindawi.com/journals/mpe/2014/351648/