基于图像差分及邻域特性的航空磁环缺陷检测

航空设备的零部件高达数十万,零器件的可靠性是保障安全飞行的前提条件。在航空零器件生产线的终端环节实现实时性高精度的检测是航空制造业实现品质控制的核心环节。传统的人工检测方法早已无法适应自动化生产流水线的速度,且该方法存在检测精度低、劳动强度大等缺点。该文基于图像差分及邻域特性对航空磁环的表面缺陷进行自动检测,通过3维块匹配滤波算法对原始图像进行去噪,并联合灰度及梯度进行阈值分割。然后对去噪图及阈值分割图进行差分运算,提取出边缘及缺陷区域。最后基于缺陷像素及边缘像素的邻域差异性提取缺陷点并检测出缺陷区域。该方法作为一种无接触、无损伤的自动检测技术,无须人工介入,通过对外环、内环及表层共3类磁环图像进行实验仿真,仿真结果说明该方法具有检测精度高、鲁棒性强的优点。

The parts and components of aviation equipment are up to hundreds of thousands,and the reliability of zero components is a prerequisite for safe flight.The realization of real-time and high-precision detection in the terminal link of aviation zero device production line is the core link of quality control in aviation manufacturing industry.The traditional manual detection method has long been unable to adapt to the speed of automatic production line,and this method has some shortcomings such as low detection accuracy,high labor intensity and so on.In this paper,the surface defects of aeromagnetic rings are detected automatically based on image difference and neighborhood characteristics.The original image is denoised by Blockmatching and 3D filtering,and threshold segmentation is carried out by combining grayscale and gradient.Then the difference operation is carried out on the denoising image and threshold segmentation image,and the edges and defect areas can be extracted.Finally,the defect points are extracted and the defect areas are detected based on the neighborhood differences between defect pixels and edge pixels.As a contactless and damage-free automatic detection technology,this method does not need human intervention.The simulation results show that the method has the advantages of high detection accuracy and strong robustness through the experimental simulation of three kinds of magnetic ring images:outer ring,inner ring and surface layer.