Study And Implementation Of Coronal Rivet Based On Image Detection Technology

With the development of computer science,electronics and artificial intelligence,image detection technology is becoming more and more matured.Object recognition and detection are widely used by image detection technology.Compared with traditional detection methods such as manual detection and three-coordinate detection,image detection technology has advantages of high efficiency,rapidity and low cost,and has become a mainstream development direction in the field of detection technology.Riveting is a more widely used connection on the aircraft,used in the aircraft rivet can reach millions.Coronal rivet is one of the rivet types,which has the characteristics of high dimensional accuracy,and its dimensional consistency directly affects the quality and production cycle of aircraft riveting.Due to the cold heading process used in the manufacture of coronal rivets,the existing means cannot meet the requirements of consistency due to the influence of die size and friction factors.Therefore,it is necessary to realize the detection of each rivet size by means of automatic image detection of rivets to meet the requirements of size consistency.In this paper,the method of coronal rivet image processing and size detection is studied.The main research contents include the following aspects.(1)Firstly,the basic method of coronary rivet image preprocessing is studied.A variety of image processing methods are adopted to perform experimental simulation and comparative analysis on the filtering,enhancement,segmentation threshold and boundary extraction of rivet images,and the basic methods of rivet image processing are determined to lay a foundation for subsequent image processing.(2)Secondly,the image processing technology of coronal rivet is studied.The preprocessed rivet image still has the noise and defects caused by the automatic rivet detection system and the characteristics of cold heading process.In this paper,a rivet image processing method based on mathematical morphology,local histogram equalization and global threshold segmentation is proposed to improve the image quality and realize the rivet boundary extraction.(3)Finally,the rivet boundary image and size detection algorithm are studied.In order to further improve the accuracy of size detection,the boundary image features are analyzed and studied,and the solutions to the existing detection difficulties are proposed.On this basis,the coronal rivet size detection algorithm is proposed,and the experimental verification of the algorithm is carried out.Through the research in this paper,the processing and size detection of coronal rivet images made by cold heading technology are realized,and the detection accuracy reaches the level of existing image detection equipment,and the size detection with prominent characteristics of cold heading is better than the existing detection equipment.The research in this paper lays a foundation for the development of automatic rivet detection system.