The main application of the AOI algorithm - errors

The main application of the AOI algorithm - errors

The application of algorithms is a key part of the application of AOI(Automatic Optical Inspection Instrument) algorithms in the field of inspection. Shenzhou Vision AOI has over 20 algorithms, each of which has its specific application. Therefore, on the basis of being familiar with and understanding various AOI algorithms, applying AOI algorithms to each detection item is the prerequisite for AOI engineers to create detection programs.

The error component is mainly used for the inspection of the component itself to check whether there is any material error in the component. This test item is a routine test item for AOI inspection. There are four detection algorithms for errors, which are TOC algorithm, OCV algorithm, Match algorithm and OCR algorithm respectively. The detection algorithm for each error item has a different focus on the detection items.

The error detection of TOC algorithms is mainly used for the error detection of non-character components, which are mainly capacitors. This type of detection method detects faulty components by extracting the intrinsic color of the component and determining whether the intrinsic color of the component has changed. Among them, the body color parameters of the components have no default parameters. They are color extraction parameters given based on the actual body color.

The error detection of the OCV algorithm type is mainly used for the error detection of clear characters, and the components of this type are mainly resistors. This type of detection method determines whether a component has a fault by obtaining the degree of fit between the contour of the character to be tested and that of the standard character. The default range of the determination parameters for this type of detection is (0, 12). If the standard character is "123", the character to be tested is "351", the fitted return value is 28.3, and the determination range is (0, 12), then this component has a "wrong component".

The Match type detection algorithm is mainly used for error detection of fuzzy characters. The components of this type mainly include diodes, transistors, etc. This type of detection algorithm mainly determines whether the component has a "wrong part" by obtaining the similarity degree between the character area to be tested and the standard character area. The determination range of this type of error is defaulted to (0,32).

OCR type detection algorithms are mainly used for the detection of components in important parts, such as BGA, QFP, BGA, etc. This type of algorithm mainly detects and judges whether errors occur by identifying the character to be tested and determining whether the character to be tested is consistent with the standard character. If the standard character is "123" and the actual character is "122", then the OCR algorithm determines that this type of component has a "wrong component".