Research On Automatic Recognition Technology Of Circuit Board Data Based On Machine Vision

In view of the deepening of intelligent modern electronic industry,intelligent detection of solder joints plays an important role in improving the reliability and stability of the circuit.In this paper,the automatic recognition technology of solder joints based on machine vision is studied in this paper,the research in this paper is to use the machine vision system to collect and recognize the image of the solder joints of the circuit board and convert it into the compatible data of the test system to achieve the full automation of the course from the image gathering of the circuit board to the test data generation of the automatic test system.PCB solder joint identify and coordinate data extraction.This paper is in view of the home and board research on the related technology of circuit board solder joint image processing,based on the self-designed visual system and image processing software platform,combined with the design index requirements and algorithm characteristics,the following tasks should be completed::(1)using SIFT The feature extraction algorithm registers the solder joint image,and uses the smoothing algorithm to fuse the image in the image overlapping area;(2)The image color channel model selection,image noise reduction method,image segmentation,mathematical morphology,etc.Technology,using mathematical morphology to lock the target area of the solder joints on the circuit board,and accurately segment the background area and the target area and mark the solder joint area,accomplish the shortage the target area and accomplish the statistics on the number of solder joints;(3)after the development of the lifting edge detection algorithm,the more authentic and important edges of the soldering joint are acquired,and the edges are acquired through the point after the soldering joint edge scan point,and abnormal edge points are removed by fitting a circle to reduce the accuracy of false edges to coordinates.The Zernike sub-pixel torque is reused to locate the edge of the soldering joint at the level of the sub-pixel,and the high-precision solder joint center coordinates are obtained by fitting;(4)Actual lute coupling coordinates the system calibration method produces efficient soldering elements and February J extraction is complete Circuit plate parameters.Experiments show that the machine vision implemented in this design has an automatic recognition rate of more than 95% for circular solder joints with a diameter of 1-4 mm.The extracted solder joint coordinates are compared with the real coordinates exported using Altium Designer software,which proves that the method is extracted The maximum error of the solder joint coordinates is less than 0.1mm,and Excel coordinate data can be exported,which better provides a research basis for the coordinate position of the automatic test system.