Research On The General Location Identification And Detection System Of SMT

Mounting Machine is the Core equipment of(SMT)production Line.Improving the mounting efficiency of the mounting machine is the key to improve the production efficiency of the SMT system,and the mounting efficiency of the mounting machine is mainly determined by the speed of positioning and testing of the mounting components in the process of mounting,so it is necessary to improve the detecting speed of the mounting components.In the process of actual mounting,there are different types of components with or without pins,protruding pins and the shape of pins,so the algorithms for locating and correcting errors of various components will be different.Therefore,the purpose of this paper is to study the location identification and detection of various components,to improve the speed of positioning and detection of chip elements,to increase the scope of mountable components,and thus to increase the production benefit of the mounting machine.This subject is based on the improvement of LED special double rotating head arch mounting machine which is developed by the laboratory and transformed it into a multi-function mounting machine which can mount many kinds of components.The following major tasks were completed:(1)Through the analysis of the composition of the mounting machine vision system and the performance of the high-speed mounting machine in the market,the performance index of the mounting machine studied in this paper is formulated,and the hardware composed of the mounting machine visual system is selected according to the performance index.Set up the visual system that this subject research needs;(2)By analyzing the process of image pre-processing such as image filtering,image segmentation,image edge detection and so on,combined with the characteristics of the patch elements and the working environment of the actual mounting,the method suitable for the image preprocessing of the patch elements is selected.(3)In this paper,the shape and pin characteristics of commonly used patch elements are analyzed,and the correction principle of patch elements is analyzed.Combined with theimage processing and other visual library of OpenCV,a sheet element(Chip),airfoil pin element(SOP),is proposed.Based on the improved correction algorithm of machine vision inspection,three kinds of spherical pin element(BGA)are used to detect the center position offset and angle offset of the component.The correction algorithm of Chip element is based on the method of maximum internal rectangle,finding the maximum inner rectangle of the image profile of the component,and calculating the deflection angle and the center position of the maximum inner rectangle.The correction algorithm of BGA and SOP elements is based on the method of center distance,the center of the pin profile is calculated,and the center of the pin is fitted by the improved least square method,and the center of the pin is fitted by the improved least square method,and the center of the pin is fitted in a straight line by the improved least square method,which is based on the center distance method.According to the fitting line,the deflection angle and the offset of the center position of the element are obtained.Theoretical analysis and experimental test show that the improved component correction algorithm based on machine vision can meet the precision requirement of high-speed mounting machine,and the time-consuming test is compared with the existing algorithm.It is concluded that the deviation correction algorithm proposed in this paper improves the speed and has a certain practical application value.