Study On AOI System Based On Image Processing Technology

In the process of printed circuit board,components are generally placed on the circuit board by surface mount equipment,plug-in machine or manual way.There will be some errors,such as the component is missing,wrong or the direction is opposite,etc.At present,the automatic optical inspecting equipment for PCB is basically for the solder face,the inspecting effect is good for solder joint defect or position fixed component defect.The inspecting accuracy for the plug-in components on component surface is poor and the result is easily to be false.Therefore,the components surface is almost by artificial inspection.The inspecting process is repetitive and is also very prone to errors.Based on present situation above,the research of AOI system based on the image processing technology in this thesis can be able to achieve the function of detecting the defects on PCB component surface.The system consists of four parts,including image acquisition module,image processing module,program interface module and data storage module.And this thesis proves that the AOI system based on image processing technology is feasible with the analysis of the experimental results of this system.The main contents and results of this thesis are as follows:(1)The hardware part of the image acquisition module is composed of VT-EX1400 CPS industrial camera,VT-LEM0814CB-MP8 lens,box-type shadowless light source;The software part is based on the LabVIEW language and VAS visual acquisition module.The industrial camera with 14 million pixels ensures a high resolution of the captured image.The box-type shadowless light source ensures that the light on the PCB is even and the shadow of the component comes down to the minimum.This module achieves the function of high-quality and low noisy PCB image real time acquisition.(2)There are three kinds of methods to detect the existence of components,which are color extraction,computing similarity,template matching.A variety of detection methods to meet the needs of the vast majority of component testing.(3)Diode polarity detection algorithm aimed at the uncertain location of diode body,using two methods of diode positioning which are threshold segmentation and the method of finding pin;Analyzing the diode body to judge the polarity of the diode.(4)Electrolytic capacitor polarity detection algorithm uses the improved Hough circle detection method to determine the location of electrolytic capacitor;Then intercepting and expanding the ring with polarity indicator of electrolytic capacitor,and threshold segmentating for the ring using the method of OTSU.And then judge the polarity.(5)According to fixed pin position and the different distance between both sides and the pin,socket direction detection algorithm use a method of calculating the distance from the edge of the socket to the pin,in order to judge the direction of socket.