基于散乱点云的飞机蒙皮缝隙识别算法  

Aircraft skin seam recognition based on scattered point cloud

作　　者：李红卫 戴佳佳 谢乾 汪俊[2] LI Hong-wei;DAI Jia-jia;XIE Qian;WANG Jun(AVIC Xi’an Aircraft Industry(Group)Company LTD.,Xi’an 710089,China;College of Mechanical and Electrical Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China)

机构地区：[1]航空工业西安飞机工业(集团)有限责任公司,陕西西安710089 [2]南京航空航天大学机电学院,江苏南京210016

出　　处：《计算机工程与设计》2022年第2期465-471,共7页Computer Engineering and Design

基　　金：国家自然科学基金面上基金项目(61772267)。

摘　　要：针对数字化测量技术,研究基于散乱点云的飞机蒙皮对缝特征识别方法,力求快速定位对缝点并提取其空间特征;建立基于张量投票的点云特征模型,对缝隙点的识别方法进行研究,以多模型拟合技术获取多条拟合直线,进一步通过最小二乘法得到最终拟合直线。分析点云数据的几何信息,利用强化策略加强缝隙特征。实验结果表明,该方法能够有效提取飞机蒙皮对缝的缝隙点,并将其转换为直线模型,高效地进行对缝点的识别。Aiming at the digital measurement technology,the recognition method was proposed to extract the seam points efficiently of the aircraft skins,and to locate the joints quickly and extract their spatial features.The tensor-voting algorithm was used to establish the seam region of the point cloud,and the seam points recognition was learned using multi-model fitting technology,and the final lines were obtained using the least square method.The geometric information of point cloud data was analyzed,and the seam features were strengthened with the reinforcement strategy.The proposed algorithm can efficiently extract the seam point of the aircraft skin and obtain the seam feature model by converting it into the line model through the effective proof of the experiment.

关 键 词：飞机蒙皮 缝隙识别 张量投票 直线检测 模型拟合 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]