Research And Implementation Of A 3D Mapping Point Generation Algorithm For Laser Etching

Laser etching processing is one of cutting-edge processing technologies in the field of modern machining.Especially in the etching of precision patterns on the surface of parts,laser etching technology is superior to other traditional processing techniques in terms of processing effect and efficiency.The input information of the three-dimensional laser etching mainly includes three-dimensional model information and etching pattern information.Since the pattern information for etching is generally stored in a small texture texture,an algorithm is required to automatically and gauge a large amount of small texture information on the surface of the three-dimensional model.The position of each of the textures used for etching on the three-dimensional model called the mapping point of the model.How to generate the mapping points according to the specified texture information and model information is an important problem in laser etching processing.The traditional research is mainly divided into three directions: one is the method commonly used in industrial processing,which firstly etch the processing result on a two-dimensional soft surface and then adsorb it on the surface of the machined part.For high requirements on the etching material,this method is not suitable in most cases.The second is to decompose a surface model into a plurality of expandable sub-surfaces by UVW expansion and expand them to a two-dimensional plane.Since the decomposition process usually tears the model,this makes the expansion results difficult to use for the final point generation.The third is to transform the model into large-scale discrete point cloud data and expand the point cloud data to a two-dimensional plane.Since traditional point cloud data generation usually relies on random sampling of each triangular surface,the final expanded point cloud has no correlation,so it is difficult to generate mapping points by this data.In this paper,a new algorithm for 3D texture point generation in laser etching based on point cloud is proposed.The algorithm complete the discrete point cloudization of the surface of the 3D model by intersecting 3D model surface and a serials of rays.With the specification of point cloud data obtained by this method,the algorithm can use the fast parallel prefix summation algorithm on the unfolded 2D point cloud texture data to replace the traditional discrete summation algorithm to obtain the surface distance of any two points on the 3D surface and generate map points.In addition,the paper optimizes the intersection of ray and 3D model by using RT core of GPU.This method is superior to the previous GPU ray intersection algorithm based on CUDA core in the efficiency of intersection.At the same time,the paper also replaces the prefix and algorithm on the traditional CPU by the GPUbased CUDA core parallel prefix and algorithm to complete the efficiency optimization of the 3D model surface distance calculation.From the analysis of the experimental results,it can be seen that the algorithm proposed in the paper completes the expansion of the 100,000-dollar triangular surface model and the point cloud generation work within 2-4 seconds,which is 80-120 times better than traditional CPU algorithm.It was also 3-4 times better than traditional GPU implementations(depending on the CUDA core).The resulting point data has an error of less than 0.02 mm per meter.Both the local error and the global error(the variance of the local error)are much smaller than the traditional UV expansion method and the traditional point cloud method based on triangular surface random sampling.Therefore,the algorithm proposed in this paper has a very significant improvement in efficiency and effect compared to traditional algorithms.