增材制造用的三维静态模型重建方法  

A 3D Static Model Reconstruction Method for Additive Manufacturing

作　　者：徐悦 阎汉生[1] 徐勇军[1] 林欣 郭晓莹 XU Yue;YAN Hansheng;XU Yongjun;LIN Xin;GUO Xiaoying(Department of Mechanical and Electrical Engineering,Guangdong Polytechnic of Industry&Commerce,Guangzhou,China,510510)

机构地区：[1]广东工贸职业技术学院机电学院,广州510510

出　　处：《福建电脑》2022年第7期27-29,共3页Journal of Fujian Computer

基　　金：中国职业技术教育学会分支机构、研究院(中心)2021课题资助(No.2021B162);第十三届“挑战杯”广东大学生创业大赛暨2022年广东省科技创新战略专项资金(攀登计划)校级培育项目资助(No.2022SCPY019);广东大学生科技创新培育专项资金资助项目(No.pdjh2021a0741)资助。

摘　　要：针对增材制造中的静态三维模型,本文采取三个步骤来建立模型,即三维点坐标的分层提取、三维点的融合和三维点的三角面片化。本文所用的重建算法采用多幅来自不同视角的图像,每个图像来源于各个同步采集的摄像机。这些图像都是在抠像幕布下拍摄得到,所有对象都可以被自动地分割,可通过EPVH和CVH方法依据轮廓图像建立得到可视凸壳。完成可视凸壳的建立后,重建过程按点云抽取、点云融合和点云三角化三个步骤完成。与传统算法相比,本文算法在鲁棒性、低成本和精确度方面具有显著优势。For the static 3D model in additive manufacturing, this paper takes three steps to establish the model,that is, the hierarchical extraction of 3D point coordinates, the fusion of 3D points and the triangulation of 3D points. The reconstruction algorithm used in this paper uses multiple images from different perspectives, and each image comes from each synchronous camera. The images are captured under the matting screen, all objects can be automatically segmented, and the visual convex shell can be established according to the contour image by EPVH and CVH methods. After the visual convex shell is established, the reconstruction process is completed in three steps: point cloud extraction, point cloud fusion and point cloud triangulation. Compared with traditional algorithms, this algorithm has significant advantages in robustness, low cost and accuracy.

关 键 词：增材制造 三维重建 点云 

分 类 号：TP391.7[自动化与计算机技术—计算机应用技术]