基于改进Canny-FAST算法的折叠翼几何尺寸自动化测量  

Automatic measurement of folding wing geometry size based on improved Canny-FAST

作　　者：蔡引娣 孙梓盟 李世雄 牛超 葛太坤[2] 高鑫巍 CAI Yindi;SUN Zimeng;LI Shixiong;NIU Chao;GE Taikun;GAO Xinwei(College of Mechanical Engineering,Dalian University of Technology,Dalian 116081,China;China Airborne Missile Academy,Luoyang 471009,China)

机构地区：[1]大连理工大学机械工程学院,辽宁大连116081 [2]中国空空导弹研究院,河南洛阳471009

出　　处：《光学精密工程》2025年第3期466-475,共10页Optics and Precision Engineering

基　　金：国家重点研发计划资助项目(No.2022YFB4600900);国家自然科学基金联合基金集成项目(No.U23B6005)。

摘　　要：为了实现非标复杂构件折叠翼几何尺寸的高精度、高效率测量,提出一种基于改进Canny-FAST算法的折叠翼几何尺寸自动化测量方法,解决了传统人工测量方法存在的测量精度低、重复性差及效率低等问题。通过在传统CannyFAST算法中引入Douglas-Peucker算法的方式以简化折叠翼轮廓特征,解决了传统Canny-FAST算法在提取特征复杂轮廓角点时提取精度低、稳定性不好等问题,提高了折叠翼几何尺寸的测量精度。基于提出方法研制了折叠翼几何尺寸自动化测量系统,通过建立该系统中测量模块和运动模块的自动交互通讯,实现折叠翼几何尺寸高精度、高稳定性和高效率的自动化测量。经实验表明,自研的折叠翼几何尺寸自动化测量系统长度测量精度优于65μm,角度测量精度优于0.049°,满足折叠翼几何尺寸测量精度要求。In order to achieve high-precision and high-efficiency measurements of the folding wing geome⁃try of non-standard complex components,an automatic measurement method based on an improved Can⁃ny-FAST algorithm was proposed.This method addresses the challenges of low measurement accuracy,poor repeatability,and reduced efficiency associated with traditional manual measurement techniques.The Douglas-Peucker algorithm was integrated into the conventional Canny-FAST approach to enhance con⁃tour feature simplification,thereby improving extraction accuracy and robustness.An automatic measure⁃ment system for folding wing geometry was developed based on this method.By establishing seamless in⁃teractive communication between the measurement module and the motion module,measurement that is automatic,high-precision,stable,and efficient was realized.The performance of the self-developed auto⁃matic measurement system is verified,with length measurement accuracy exceeding 65μm and angle measurement accuracy better than 0.049°,thus meeting the requirements for folding wing geometry as⁃sessments.

关 键 词：折叠翼 图像拼接 自动化 Canny-FAST算法 DOUGLAS-PEUCKER算法 

分 类 号：TP394.1[自动化与计算机技术—计算机应用技术] TH691.9[自动化与计算机技术—计算机科学与技术]