| With the rapid development of microelectronics technology,the volume of electronic components is getting smaller and smaller,and the integration of printed circuit board is getting higher and higher.In the production process of printed circuit board,surface mounting technology gradually replaces the traditional Jack installation technology.The first process of surface mounting technology is printing solder paste.The quality of solder paste directly affects the quality of printed circuit board.Through on-line inspection of printing solder paste,on the one hand,the printed solder paste defective circuit boards can be repaired in time to reduce the production cost of printed circuit boards;on the other hand,the reasons for the defective printed solder paste can be analyzed and fed back to the solder paste printing process to optimize the process parameters related to the quality of solder paste printing,thereby improving the quality of printing solder paste.The thickness of solder paste printing is one of the important indicators to judge the quality of solder paste printing.At present,the detection of solder paste thickness mainly uses the three-dimensional detection technology in machine vision,while the three-dimensional detection mainly uses the phase measurement profilometry.Phase measurement profilometry mainly includes sinusoidal grating projection,phase unwrapping,phase-height calibration and three-dimensional reconstruction,among which phase unwrapping is the most critical link,which directly affects the speed and accuracy of three-dimensional online detection.Firstly,the principle of phase measurement profilometry is introduced,and then the basic principles and anti-noise effects of common phase unwrapping algorithms such as row-column point-by-point method,branch-cut method,quality graph guidance method and least square method are analyzed respectively.Aiming at the problem of error propagation in the phase unwrapping process caused by the through-hole area of printed circuit board,the four phase unwrapping algorithms mentioned above are presented in the through-hole area.The anti-noise ability and the time consumed are compared experimentally.According to the experimental results,the scheme of identifying and locating the through-hole region using standard machine vision algorithm package halcon,using least squares method for through-hole region and using row-column point-by-point method for phase unwrapping of other regions is determined.By comparing three printed circuit boards with different numbers and sizes of through holes,the phase unwrapping experiments are carried out by using the row-column point-by-point method with fast phase unwrapping speed,the least squares method with strong anti-noise ability and the phase unwrapping scheme determined in this paper.The experimental results show that the phase unwrapping scheme determined in this paper has the advantages of strong anti-noise ability and fast speed in the phase unwrapping process.On the basis of introducing the mathematical model of transforming the unwrapped phase into height and the process of three-dimensional reconstruction of the point cloud of height data,the phase unwrapping of a calibration block containing through holes and solder paste is carried out by using the phase unwrapping scheme combining the least squares method and the row-column point-by-point method.The phase unwrapping of the calibration block obtained by the four-step phase shifting method is carried out,and the height of solder paste is calculated by the unwrapped phase Three-dimensional reconstruction of point clouds.From the phase unwrapping diagram,it can be seen that the error of phase unwrapping in through-hole area is limited to a small range,which can effectively solve the error propagation problem caused by through-hole area in the phase unwrapping process,so as to obtain higher phase unwrapping accuracy.From the solder paste height data obtained after three-dimensional reconstruction,it can be seen that the measured solder paste height error meets the detection requirements.Finally,the causes of measurement errors are analyzed. |
PDF Full Text Request