Local-genetic slicing of point clouds for rapid prototyping

Purpose - The purpose of this paper is to propose to overcome the limitations of polygonization of point clouds for rapid prototyping purposes, by using a direct slicing approach, based on a hybrid local-genetic algorithm to achieve a robust direct slicing system. Design/methodology/approach - At first a volume analysis is performed on the point cloud and a space decomposition is realized using elementary voxels. Then, considering each Z level of the voxelized point cloud, the external non-void voxels are linked togheter using an hybrid local and genetic approach, to generate the contour of the object with an automatic process. The contour of the object is finally converted into slice files suitable for the rapid prototyping machine. Findings - The genetic algorithm (GA) is very effective in detecting those slices where an optimal solution is not achieved with the local approach, and in finding the minimum path that connects the points belonging to the slice contour. Research limitations/implications - Further studies must be conducted to improve the efficiency of the approach to the travelling salesman problem (TSP) and to the relation between the cell dimension and the point cloud density. In this context, the use of adaptive slicing will be considered, in order to improve time performances. Practical implications - The approach is fully automated and enables the direct creation of layered manufactured copies of 3D scanned products directly from the point clouds, avoiding the tessellation phase that is often time consuming and characterized by errors in the STL file. Originalitylvalue - The use of TSP problem to solve the direct slicing of point clouds is more effective than simple spline fitting techniques, avoiding self-crossing curves. This approach solves the TSP problem for each slice, exploiting the volumetric space decomposition to fasten the achievement of the solution. In fact the local knowledge is used by the nearest neighbours local search and the partial solution achieved is the starting point of the GA. The GA is effective in finding global minima and results to be fastened by the local approach.