Research On Algorithm Of Detecting The Polarized Electronic Component Of PCB-board Based On Machine Vision

The result of detecting the flawed electronic component on the PCB-board will influence the production cost of circuit board directly,among which the lack of the component,reverse connection and misconnection of polarized electronic component will do extremely huge harm to the circuit board.Compared to the traditional manual detecting method,automated optical inspection(AOI)has higher reliability,continuity and accuracy.Therefore,it is of great importance to invent an AOI system applying to the polarized electronic component of PCB-board.This paper mainly solve the problem of detecting the flawed polarized electronic component of PCB-board,focuses on the research of the fast algorithm for detecting the electrolytic capacity and the polarity of the diode,establishes a particular AOI detecting system and realizes the detection for the flawed polarized electronic component of the power panel of LCD displayer.The research on this paper includes:1?PCB-board flaw detecting system establishment.The detecting system is divided into hardware system and software system.The hardware system consists of lighting module and image gathering module and the software system consists of template configuration module,component information module and component programming module.2?The research on the preprocessing of images and the circuit board location algorithm.Firstly,the image is rotated and the noise is removed with smoothing processing adopting the Gauss filter algorithm.Secondly,the target and background of the image are segmented through image segmentation algorithm,and the edge of the target image will be extracted with canny edge detecting algorithm.In circuit board location algorithm,the Hough line detection will be conducted towards the circuit edge image,and the conjunction of the line will be selected as reference point,on which the locating hole will be searched.And then the location of the circuit board will be realized by locating the location hole with Hough detecting algorithm.3?A generalized Hough detecting algorithm is presented for detecting the electrolytic capacity fast.Firstly,the canny edge detecting algorithm is optimized by adding the user-defined sensibility a and selecting the threshold value,thereby improving its resistance to interference.Secondly,the voting pints is reduced by lessening the involvement of pixels on the edge,and the space of accumulator is shrunk via pyramid hierarchical searching,thereby increasing the speed of the algorithm.The result of the experiment suggests that the speed of detecting electrolytic capacity with the algorithm presented in this paper is 80%higher than that with traditional generalized Hough algorithm.4.The research on detecting algorithm of flawed diode,slot and IC.Firstly,the diode is located through the shape-based template matching algorithm.Secondly,an improved Otsu algorithm is presented to realize the detection of polarity of diode,with the interclass variance maximized and the distance between pixels minimized.The improved algorithm segments the target and background relatively well and detects the polarity of the diode relatively accurately under the outside interference.At last,the detection for the flawed slot and IC is thereby realized.