基于图像识别的飞机铆钉尺寸检测系统开发

发布时间：2018-07-01 16:16

本文选题：图像识别 + 中值滤波　；参考：《南京航空航天大学》2016年硕士论文

【摘要】：铆钉作为重要的飞机紧固件,其尺寸精度及质量的检测对飞机装配有重要意义。针对传统铆钉检测手段存在的检测效率低、精度不易控制、人为因素影响较大、容易损坏铆钉等弊端,本文开发了一种基于图像识别的铆钉尺寸检测系统,并对此进行了相关的研究工作。主要研究内容包括以下几个方面:1.通过对铆钉尺寸检测系统功能需求分析,结合铆钉自身的特性,对图像识别检测系统进行了理论分析,完成了整个铆钉尺寸检测系统的总体方案设计,确定了该检测系统的最终实现形式;2.根据检测要求,完成了对黑白工业相机、远心镜头、光源等关键元器件的选型以及对铆钉的输送机构和夹持机构进行了设计,完成了该检测系统的机械结构装配图;3.通过对图像处理算法进行分析和研究,然后使用两种经典的滤波算法对含噪声的铆钉图像进行处理与分析;最后提出了一种改进的中值滤波算法,该方法在进行去噪处理时加入了信号点和噪声点的判断,能有效地提高去噪效果;4.通过研究相机针孔成像模型和相机标定方法,完成了系统的标定。采用模块化设计思路开发出便于人为操作的软件系统界面。该铆钉尺寸检测系统作为自动化检测设备,结构简单,成本低,工作效率高,测量精度可以达到0.01mm,能够满足现代飞机装配制造在线测量及实时控制的要求,提高了飞机装配的自动化水平及装配质量。
[Abstract]:As an important aircraft fastener, rivets are of great significance to the assembly of aircraft because of their dimensional accuracy and quality. In view of the shortcomings of traditional rivet detection methods, such as low detection efficiency, difficult to control precision, great influence of human factors and easy to damage rivets, a rivet size detection system based on image recognition is developed in this paper. And the related research work is carried out. The main research contents include the following aspects: 1. By analyzing the functional requirements of the rivet size detection system and combining the characteristics of the rivet itself, the paper makes a theoretical analysis of the image recognition and detection system, and completes the overall scheme design of the whole rivet size detection system. The final realization form of the detection system is determined. According to the testing requirements, the selection of key components, such as black and white industrial camera, telecentric lens, light source and so on, as well as the design of rivet conveying mechanism and clamping mechanism are completed, and the mechanical structure assembly drawing of the inspection system is completed. Through the analysis and research of image processing algorithm, two classical filtering algorithms are used to process and analyze the riveted image with noise. Finally, an improved median filter algorithm is proposed. The method adds signal points and noise points to the denoising process, which can effectively improve the denoising effect. The system is calibrated by studying the pinhole imaging model and camera calibration method. The software system interface which is convenient for human operation is developed by modularization design idea. As an automatic testing equipment, the rivet size detection system has the advantages of simple structure, low cost, high working efficiency and measuring precision of 0.01mm. it can meet the requirements of on-line measurement and real-time control of modern aircraft assembly and manufacture. The automation level and assembly quality of aircraft assembly are improved.
【学位授予单位】：南京航空航天大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：V262;TP391.41

【参考文献】