| Nowadays, most inspections of manufacturer on quality of tiny component are mainly depended on manual means. The manual inspection ways cannot reach the market requirement for the quality and quantity of precision components. In this paper, the machine vision based inspection on precision small electronics components is investigated. According to the user's requirements, a visual inspection system for component's surface is designed, therefore, the auto inspection of component's surface quality is realized and the efficiency of manufacturer is improved greatly.Firstly, the background and meaning of this study are analyzed, and then the current state of inspection about precision small components is introduced. Thereafter, the method based on machine vision for the components inspection and the advantages of machine vision are analyzed. Moreover, the development status and trends of machine vision are introduced.Secondly, the development indicators of visual inspection system and the difficulties in detecting objects are introduced, and a visual inspection hardware platform for surface defect of precision tiny components is then designed. According to characteristics of the components, an image acquisition project is designed. Assisted by using mechanical transmission and robot, a control method base on PC+PLC+ROBOT is proposed. In this way, the resource utilization rate and the detection efficiency are both enhanced. Moreover, two simple and practical separation methods of blow-out model and two-rail system are designed and explained.Thirdly, the camera calibration theory as the reason of robot tracing and catching is studied. In this part, some current calibration means are briefly introduced and a camera calibration way on 2D target is completely illustrated, and the experiment results are given in the end.Fourthly, the image detection algorithm about surface defects of precision small components is investigated. In this part, pretreatment of image processing is discussed, and the enhancement algorithm and the denoising algorithm of image are studied in detail. Then, the way of area positioning in the defect inspection of precision components is introduced and the detection means base on edge detection and threshold Segmentation are described in detail. In addition, some specific methods about surface defects feature extraction and recognition are described.Finally, the software of visual detection system is given, and the detection results and some reasons of misdetection and false detection are analyzed. |
PDF Full Text Request