| Automated Optical Inspection(AOI)is a critical part of the manufacturing,assembly and inspection process of Printed Circuite Boards(PCBs),mainly for position,shape and defect detection.With the miniaturization of electronic devices,traditional two-dimensional AOI cannot meet the increasing quality inspection requirements.3D AOI can effectively obtain 3D contour and depth information through structured light imaging,which improves the dimensionality of information acquisition and makes 3D AOI system have higher accuracy and stability.High-precision 3D AOI technology has a broader application prospect,so it is necessary to study the imaging algorithm.It has gradually become the mainstream choice for PCB manufacturing production lines,with a very broad application prospects.The structured light imaging in 3D AOI system can be divided into two parts:image coding and decoding algorithm and 3D information reconstruction algorithm.This paper focuses on the key technical issues such as high dynamic range imaging,inter-reflection suppression and imaging efficiency improvement,and the main work is as follows:(1)Design and build a 3D AOI imaging system.According to the quality inspection requirements of the PCB,the imaging system hardware was researched and selected,a high-precision structured light imaging system based on a telecentric lens camera and a tilt-shift projector was built,and the host computer software was developed for it for system testing and analysis.(2)A decoding algorithm based on multi-frequency phase fusion is proposed for the high dynamic imaging problem in PCB imaging environment.In this paper,the prior information of the multi-frequency fringe pattern is used to fuse and optimize the phase of the multi-frequency fringe pattern,which solves the problem of excessive noise in the traditional three-frequency heterodyne method.Experiments show that the algorithm significantly reduces the phase series jump problem and improves the signal-to-noise ratio of the phase map.(3)In order to improve the real-time performance and stability of imaging,a fast phase-height mapping algorithm combining cross-ratio invariance and local phase plane fitting is proposed.In this paper,the height mapping is performed based on the invariance of the cross ratio,which reduces the number of fringe images by half compared with the traditional triangulation imaging.At the same time,the phase distribution function is combined to achieve fast pixel matching,which significantly improves the imaging efficiency.Tests show that the algorithm has high accuracy and stability,and is easy to calibrate.(4)Aiming at the common problem of inter-reflection in structured light imaging,this paper studies the inter-reflection suppression based on the improved frequency shift method.Based on the traditional frequency shift method,this paper introduces four-step phase shift and window function to encode and decode images to improve imaging performance and efficiency.Based on the principle of multi-view geometry,epipolar constraints are used to correct the misidentified direct illumination.The experimental results show that the algorithm can effectively identify and suppress the indirect illumination caused by inter-reflection,and solve the problem of incorrect identification caused by excessive indirect illumination or multiple indirect illumination.The 3D AOI imaging system designed and constructed in this paper can meet the requirements of high-speed and high-precision three-dimensional imaging,and the imaging efficiency is doubled compared with the traditional imaging system.The experiment shows that the algorithm can image with an accuracy higher than 10 um within the depth of field of 3mm,which provides a reliable technical support for PCB quality inspection. |
