Generating NC tool paths from random scanned data using point-based models

This paper presents a new approach for the generation of NC tool paths from random scanned data. Instead of using smooth or triangulated surfaces reconstructed from raw data, which is usually a time-consuming reverse engineering approach, the point-based surfel models computed by a GPU (graphics processing unit) are used to generate NC tool paths. The tool-path generation is highly efficient and still maintains the advantage of having accurate and smooth machining result. The word “surfel” itself is the combination of the two words “surface” and “element”. It is originally applied to the rendering of scanned data. In this paper, the point-based model is created using an elliptical Gaussian re-sampling filter that is based on a signal re-sampling algorithm. Since the input scanned data is of discrete and random nature, the warping process is utilized to transform the input data into a continuous surface and then re-sample the continuous surface by using GPU. Because the re-sampled data can accurately represent the original surface, tool paths can be generated based on the point data set. For cutting tools with various sizes, adaptive re-sampling schemes are employed to generate sufficient sampled points for the generation of accurate and smooth tool-paths.