Research On Online Real-time Detection System Of PCB Components

Defect inspection of printed circuit board components is a necessary link in its production process.However,at present,the automatic optical inspection equipment for PCB board component defects in foreign countries is expensive,and such equipment in China is not mature enough.Therefore,most electronic companies still use manual visual inspection to perform inspections.It is easy to cause missed detection and false detection,and the efficiency is low.Therefore,the development of a low-cost,high-efficiency PCB board component defect detection equipment has become an urgent need for electronics companies.Based on this,this paper studies the PCB board component online real-time inspection system,obtains the PCB image on the plug-in line in real time,and processes and analyzes the acquired image to obtain the PCB board component defect detection result The main work of this article is as follows:(1)Study the PCB board stitching algorithm.In order to solve the problem of obvious patching in the images obtained by the fast matching stitching algorithm,by studying the ORB feature point matching algorithm and analyzing its advantages and disadvantages,this paper proposes an improved ORB stitching algorithm based on limited areas.Reduced splicing errors.The algorithm improves the efficiency of feature point collection by limiting the sampling area,and constrains the feature point matching distance to accelerate the feature point matching and improve the stitching efficiency;(2)By studying the characteristics of PCB board components,the components are classified into two categories:polar components and non-polar components.In order to solve the problem of strong dependency on the information of template components during threshold segmentation of color components,this paper proposes a threshold segmentation algorithm based on the color channel difference,which only uses the three-channel gray-level mean difference to differentiate to reduce template dependency;Aiming at the problem that white and highlight components are not easy to be separated from the white silk screen on the bottom plate,this paper proposes a threshold segmentation algorithm based on BCG,which improves the contrast between components and silk screen to achieve effective segmentation;Aiming at the problem of poor gray template matching during diode polarity detection,this paper proposes a geometric matching algorithm based on the shape of the target.This algorithm combines the characteristics of edge curve extraction and geometric feature extraction,and has good matching effect;In the detection of electrolytic capacitor polarity,in view of the problem of poor fitting of the defect circle by the traditional least square method,this paper proposes a least square circle fitting algorithm based on the subpixel edge to solve the problem of fitting convex and concave defect circles;(3)Based on the study of DM codes and QR codes,a DM code positioning algorithm based on the corner point density is designed based on the rich black and white blocks of the DM codes.In combination with the dark channel principle,the problem of overexposure in the imaging process is solved and the image contrast is enhanced To improve the recognition rate;For QR code locators,a QR code localization algorithm based on anamorphic locators is designed,and a multi-threshold segmentation algorithm is used to solve the problem of poor threshold segmentation caused by shadows of QR codes caused by high components;(4)Test result analysis and data recording,test and verify the function of this system,calculate the accuracy rate of each module,and record the test data locally in the form of Excel,and transmit the data to the database in the server through the local area network.Organize and analyze.