摘要
利用计算机视觉技术对蒙皮接缝进行测量是该领域的主流发展方向,但单一的视觉测量方法具有自身的局限性。为进一步优化蒙皮接缝特征测量技术,提出了将线结构光视觉测量与面结构光视觉测量相结合的方法检测接缝特征,探索一种基于数据融合的蒙皮接缝特征测量算法,修正接缝的阶差值与间隙值以提高测量精度。依据两种结构光视觉测量技术的特点,利用分布式二级数据融合算法,通过数据对准、剔除异常值以及自适应加权融合相结合的方式,对两种结构光视觉测量数据进行融合处理,并应用于实际蒙皮接缝测量。试验结果表明,经过分布式二级数据融合处理后,阶差测量值最大均值误差为0.0319 mm,间隙测量值最大均值误差为0.0344 mm,阶差和间隙的平均误差分别降低了4%和30%。此融合算法均衡了两种结构光视觉测量技术的优势,降低了系统整体误差,为高效、准确测量蒙皮接缝提供了新思路。
Computer vision is becoming a common trend in aircraft skin seams feature research,but the simplex computer vision meas-urement method has limitations.To further optimize the measurement of air craft skin seam features.A mixed measurement method is employed.Line laser and FPP are used to measure the seam area of aircraft skin.Data from two measurement methods are fused to cor-rect for gaps and flushes.Firstly,according to different characteristics of line laser and FPP,the distributed two-level fusion algorithm is adopted to eliminate outliers through data alignment.Then,the two kinds of measurement data are fused through adaptive weighted fu-sion.Finally,it is applied to the measurement of skin seams for actual aircraft.A comparison control design with four groups of standard seams and one group of actual aircraft skin seams is utilized.The experimental data suggest that the accuracy of gap and flush is signifi-cantly improved after the measurement results are optimized through distributed two-level data fusion.In comparison with the measure-ment data before optimization,the maximum mean error of flush measurement value is 0.0319 mm,and the maximum mean error of gap measurement value is 0.0344 mm,the gap measurement accuracy is improved by nearly 30%,and the flush measurement accuracy is improved by nearly 4%.This fusion algorithm balances the advantages of two structured light vision measurement technologies,reducing overall error of the system,which provides a new idea for efficient and accurate measurement of skin joints.
作者
温志辉
夏桂锁
刘芳
刘孙强
刘万栋
WEN Zhihui;XIA Guisuo;LIU Fang;LIU Sunqiang;LIU Wandong(School of Testing and Optoelectronic Engineering,Nanchang Hangkong University,Nanchang Jiangxi 330063,China;AVIC Shanghai Aero Measurement Controlling Technology Research Institute,Shanghai 201601,China)
出处
《传感技术学报》
CAS
CSCD
北大核心
2024年第4期629-638,共10页
Chinese Journal of Sensors and Actuators
基金
国家自然科学基金项目(51765049)
航空科学基金项目(2019ZE056001)。
关键词
数据融合
自适应加权
结构光
飞机检查
接缝测量
data fusion
adaptive weighting
structured light
aircraft inspection
seam measurement
