Vision-based electronic assembly part assembly precision detection method

Abstract

The invention relates to a vision-based electronic assembly part assembly precision detection method, and the method comprises the following steps: S1, collecting a scene image, and building an image coordinate system; S2, setting a search scanning range for each feature line and each feature hole, and respectively acquiring point set position coordinates of four feature lines on the chip and circle center feature point set position coordinates of a plurality of feature holes on the substrate by adopting an edge detection algorithm; S3, calculating the center distance between the chip and the substrate under an image coordinate system according to the point set position coordinates of the feature lines and the circle center feature points; and S4, comparing the center distance obtained in the step S3 with a standard center distance to obtain an assembly deviation value of the electronic assembly part. According to the method, a search scanning range is added for each feature, and an edge detection algorithm is adopted, so that the complexity of the algorithm is reduced.

Description

technical field [0001] The invention relates to the field of precision electronics manufacturing, in particular to a vision-based detection method for assembly precision of electronic assemblies. Background technique [0002] In the field of precision electronics manufacturing, a typical electronic assembly is usually involved. The electronic assembly is bonded to the substrate by a single chip through UV Epoxy glue. Machine vision is generally required to detect the position accuracy of the two after bonding. method, machine vision for non-contact measurement of electronic assemblies. The four characteristic lines of the chip to be processed are located on the surface of the chip, and the calculation results after the characteristic line image processing determine the geometric center position of the chip. The characteristic holes to be processed are distributed on the substrate surface, and the calculation results after the characteristic hole image processing determine T...

AI Technical Summary

Problems solved by technology

[0003] When the electronic assembly is placed horizontally for inspection, the surface facing the upward camera is not a flat shape, but is warped upward. The warped substrate structure makes the feature lines and feature holes to be processed that belong to two different component surfaces Not on the same plane, but there is a large height difference in the depth of field of the industrial camera. If a single industrial camera is used to process the feature lines and feature holes on the surface of two different components, only one of the feature lines and feature holes will be within the focal length, and the other A resulting feature image is not clear, which affects the detection accuracy; if the feature line and the feature hole are not in the same plane, it will also cause inconsistent illumination and uneven brightness of the assembly, which will also affect the detection accuracy

[0004] At the same time, the image algorithm for finding characteristic holes and characteristic lines in the prior art uses a template matching positioning algorithm based on gray value, the algorithm time is complex, and a large number of calculations lead to a decrease in the real-time performance of the detection process

Images