Detect Codes Neural Network

Recognizes different code types such as barcodes, QR codes and data matrix codes in the image provided using a trained neural network. The recognized codes are sorted from high to low according to their model confidence. If a code is recognized, the area is enlarged according to the ratio of the code region. Depending on the code type detected, the appropriate reading algorithm is then applied.

Optionally, a Boolean signal can be selected for the Enable Execution input so that the algorithm is only active if the value of the selected signal is TRUE.

Configuration options

• Num Codes: Maximum number of codes to be found.
• Model: Model of the neural network for code recognition. Different numbers of models are available. These include specialized models for 1D codes, 2D codes or all code types.
• Code Type: Fixes the code type(s) to be recognized.
• Code Region Ratio: (Optional) Enlarges or reduces the code region found in proportion to its current size.
• Dot Code Module Width: (Optional) Average dot diameter of the code in the image in pixels. The value must be at least 3, preferably 5 to 8. Only required for dot codes.

Output values

• Count Results: Is incremented when a new result is provided.
• hr: HRESULT error code that outputs the result of the processing.
• Last Event: Timestamp of the last result.
• Codes Detected: Number of codes found.
• Image: Initial image of the algorithm.
• Code Type 00..n: Type of the code found
• Decoded Data 00..n: Decoded data of the code found.
• Top Left Corner 00..n: Top left corner of the code found.
• Bottom right corner 00..n: Bottom right corner of the code found.
• New Result: (Optional) TRUE if a new result is provided.
• Image Preview: (Optional) Image in which the objects found are marked.