Design And Optimization Of 3D Printing Tree Support Structure Based On Instance Segmentation

3D printing technology is a relatively mature processing technology.In actual production,external support needs to be added to ensure the smooth progress of the processing process.Tree support is a new type of external support method,combined with the instance segmentation technology in artificial intelligence,can better solve the problems of collapse,deformation,and lack of precision in the 3D printing process,and improve production efficiency.This article focuses on the current situation of the collapse and lack of precision of tree support in production and processing.The specific contents are as follows:(1)For the problem of feature loss in the printing model,a new instance segmentation algorithm is adopted,including deepening the feature extraction architecture,using the nonlinear activation function Mish,using CIoU loss in the regression loss function of target detection,and using Ranger optimizer training.The accuracy of recognition of local features of the object model is improved by 13.2%,and the instance segmentation of feature regions is performed.(2)For the feature reconstruction after instance segmentation,the classical hiding method is optimized and improved,including Z-Buffer algorithm of spatial range hiding method and area scan line algorithm of plane range hiding method.It realized the projection mapping transformation of 3D object model by combining coordinate refinement and Z-Buffer algorithm.Adaboost ensemble learning and reverse hidden elimination method were used to realize 3D feature reconstruction.(3)For the multi-featured three-dimensional object model,based on the umbrella search algorithm,completed the improvement and generation of multiple types of tree supports,including tree supports from the model itself,tree supports with different growth angles,and different growth Proportional tree supports and tree supports with different branch diameters refine the generation criteria of tree supports.(4)Based on the improved instance segmentation and 3D reconstruction algorithm,the performance analysis of the processed parts with different tree supports through orthogonal experiments is aimed at the recognition of local features of the model,the selection of support points and the multi-type tree support during the 3D printing process.Conduct comprehensive experimental research on issues such as improvement and generation.Experiments show that the use of the new tree support generation algorithm can save more than 10%of production time,can improve the surface finish of the model by about 7%,and increase the material utilization rate by about 10%.