Variable-string reading
Capture changing letters, numbers and permitted symbols from every inspected product.
Industrial OCR inspection
Read variable printed text at production speed
Use optical character recognition to capture changing date, batch and lot strings, then compare the result with the information authorised for the active job.
Optical character recognition
Purpose-built vision inspection
Turn physical print into data the line can verify
OCR determines which characters are present when the printed string is not fixed in advance. It is ideal for reading genuinely variable information before applying date logic, database comparison, logging or product-to-code matching.
Data comparison
Compare the OCR result with a recipe, PLC value, database record or production-order scan.
Format rules
Check string length, position, separators and the allowed character set.
Confidence threshold
Reject a result when the system cannot identify the printed content with sufficient confidence.
System outcomes
Use the printed code as actionable production data
Once characters are read, the line can validate logic, match records, store results and respond differently to missing, incorrect or low-confidence codes.
Read unknown content
Recognise genuinely variable strings instead of comparing only with a taught image.
Apply business rules
Validate dates, prefixes, lengths and permitted combinations after recognition.
Exchange results
Send captured data and pass/fail states to line controls or authorised systems.
Application engineering
Design around the smallest critical character
OCR reliability depends on image resolution, focus, contrast and the range of approved character appearances. We calculate the field of view around the smallest feature that must be distinguished and stabilise the product so the code stays inspectable.
- Character height and stroke width
- Font consistency and permitted symbols
- Code contrast and background variation
- Line speed, exposure and product motion
Related inspection systems
Build a complete code-control point
Combine the inspection with complementary print, data or reject checks where they reduce the risk of a non-conforming pack reaching the customer.
Frequently asked questions
Industrial OCR inspection questions
Every inspection begins with real samples, representative print and the operating conditions found on your production line.
What is the difference between OCR and OCV?
OCR identifies which characters are present when the content may be unknown. OCV compares visible characters with a known expected string or taught reference. The best method depends on how variable the message is.
Can OCR read dot-matrix inkjet fonts?
Yes, provided the camera resolution, print quality and font variation allow reliable recognition. Training or font-specific tools may be used, and real degraded samples are important for setting limits.
Can recognised text be saved?
Yes. Depending on the required architecture, the captured string, pass/fail result, timestamp and image can be transmitted or stored with appropriate data-retention rules.
Oxford Vision Systems
Prove OCR performance with the full code range
Supply every legitimate font and format, plus marginal and failed samples. We will assess recognition confidence and false-reject risk.
Continue planning
What should a date-code inspection system check?
Define the expected code content, location and acceptable print condition for each product. Clarify whether the task is reading variable text, comparing against an expected value or checking presence. The system also needs a controlled way to receive the correct product data and respond to a failed inspection.
- Expected data and location
- Reading, comparison or presence task
- Product change and failure response
Related guidance: System guide · Retrofit inspection