OCR Code Reading on Variable Packaging
Reliably read text on packaging that changes in size, color, and orientation
Related Products
Graphical programming environment for deep learning-based industrial image analysis
Powered by In-Sight ViDi Deep Learning-Based Vision Software
Container and package filling lines often switch container types, colors, shapes, and sizes depending on product and demand. All printed text on every container must be confirmed, particularly date/lot information, product line ID, and other information relating to product traceability.
Once the distance and orientation of packages are determined, conventional optical character recognition (OCR) with a fixed-mount vision system can accurately read characters on containers at high speeds. But any change of scale, angle, color, or background pattern requires significant adjustment or even reprogramming, making it difficult to produce and track smaller runs of various products.
Regulatory compliance increasingly demands full product traceability, with penalties for traceability failures, so accurate determination of product text without slowing production is essential.
The AI-based OCR tool accurately reads challenging OCR codes on packages on fast-moving production lines. The OCR tool is easily set up and deployed thanks to a pretrained font library. It takes only a small image set of the text on a new product or package type to train the OCR tool to find and read the necessary text, even at various angles, against a complex or reflective background.
Whenever a new product or packaging is produced, it is easy to train the OCR tool to identify and read the characters wherever they appear on the package, simply by labeling the characters in an image set of the new product.
