摘要
为了解决飞机大部件表面缺陷检测,提出了基于多维度光立体的飞机大部件表面缺陷检测应用方法。首先在图像数据采集中设计一种多维数据采集硬件模块,利用Photometric相机采集现场光线强度信息,Photoneo相机采集空间坐标信息;其次利用多传感器融合技术构建几何约束,找出传感器测量之间的关联,进而计算出点光源光线的入射方向;接着,根据朗伯余弦定律(Lambert's cosine law)计算出表面法向量,形成表面法向图;最后计算法向量图的梯度以及最小包围盒,进而获得表面缺陷。该方法能够实现多类型多尺度的缺陷检测,缺陷检出率为94.9%,检出效率10s m^(2),有效的解决了飞机大部件表面缺陷检测的问题。
In order to solve the surface defect detection of aircraft large components,an application method for surface defect detection of aircraft large components based on multi-dimensional light stereo is proposed.Firstly,a multidimensional data acquisition hardware module is designed in the image data collection set.Photometric cameras are used to collect on-site light intensity information,while Photoneo cameras collect spatial coordinate information;Secondly,using multi-sensor fusion technology to construct geometric constraints,identify the correlation between sensor measurements,and then calculate the incident direction of the point light source;Next,based on Lambert's cosine law,the surface normal vector is calculated to form a surface normal graph;Finally,calculate the gradient of the normal vector graph and the minimum bounding box to obtain surface defects.The proposed method can achieve defect detection of multiple types and scales,the defect detection rate is 94.9%,and the detection efficiency is 10s/m^(2).The method effectively solves the problem of surface defect detection for large components of civil aircraft.
作者
张宪法
冯静璇
卜晓燕
曹嘉玲
苗笑冬
ZHANG Xian-fa;FENG Jing-xuan;BU Xiao-yan(Shanghai Aircraft Manufacturing Company,Shanghai 200093)
出处
《航空精密制造技术》
2025年第4期22-26,共5页
Aviation Precision Manufacturing Technology
关键词
飞机大部件
表面缺陷检测
多维度光立体
法向图
最小包围盒
aircraft major components
surface defects detecting
multi-dimensional light stereoscopic
normal graph
minimum bounding box
