Kohonen-swarm algorithm for unstructured data in surface reconstruction

Abstract

This work introduces a new method for surface reconstruction based on hybrid soft computing techniques: Kohonen Network and Particle Swarm Optimization (PSO). Kohonen network learns the sample data through mapping grid that can grow. The implementation is executed by generating Kohonen mapping framework of the data subsequent to the learning process. Consequently, the learned and well-represented data become the input for surface fitting procedure, and in this study, PSO is proposed to probe the optimum fitting points on the surfaces. The proposed algorithms are applied on different types of curve and surfaces to observe its ability in reconstructing the objects while preserving the original shapes. The experimental results have shown that the proposed algorithm have succeeded in producing the reconstructed surfaces with minimum errors generated.