Applicative Framework for End-to-End AOI Implementation
In general , the key elements and the acquisition type described so far are selected based on use case , in order to build an architecture which guarantees a good balance between accuracy and speed .
The accuracy of the output can be improved also by studying the most appropriate illumination system . It is fairly common for the surface of the components analyzed to be enlightened by several lighting sources which are carefully chosen according to the refraction material characteristics and the type of defects to spot . By selecting the correct light type and making it diffuse homogeneously it is possible to amplify defects thus reducing the processing effort .
Built-in solutions are available in the market today , but despite optimized integration of all the components , they come at a remarkably high cost . Single architecture is calibrated and delivered to handle a specific check ; however it is difficult to scale the same system to different use cases due to its rigidity . This is why companies often opt for traditional methods such as manual inspection over AOI .
The scientific community is increasing their focus on Automatic Optical Inspection studies in the last 5 years ( Figure 2-1 ). Most of the published articles since 2016 include manufacturing-related use cases emphasizing the image analysis methods over the architectures themselves .
Figure 2-1 : Number of articles published on Scopus platform . Research strategy with keywords : “ Automatic Optical Inspection ”.
The following chapters illustrate a complete framework for an AOI system including hardware and software architectural paradigms for a from-scratch-implementation of a complete Automatic Optical Inspection system , and it will prove its scalability illustrating two different use cases .
IIC Journal of Innovation 5