A high precision OCR engine for reading cards and structured documents from an image. No matter what quality is the image, our Cards OCR can handle it.
Optical character recognition
Optical character recognition (OCR) components enable software to extract text data from an image (usually photograph) and use this data for further processing. OCR is one of the key technologies for automation - its implementation eliminates paperwork, reduces process errors and enhances user experience of customers and/or employees.
Cards OCR by Cognexa
Cards OCR by Cognexa is a high precision OCR engine suited for reading various types of cards and documents in use cases such as office automation, identity verification, onboarding process facilitation and more.
Examples of documents ideal for extraction using Cards OCR
Credit cards, Loyalty cards, etc
Usecases and requirements
Cards OCR requires a document with a standardized structure containing latin characters. It can have one or several layouts that define the position of text fields. For legal reasons, Cards OCR can not be used for text extraction from identity documents issued by government bodies.
The core of Cards OCR consists of OCXR software toolkit developed by Cognexa. The same technology is also used in our other products - tools for reading utility consumption meters
, car odometers
, signs and industrial product signatures.
We recommend using our OCXR based tools especially in cases where the accuracy of conventional tools such as Tesseract is not sufficient. That generally means if any of the following applies:
- The documents to be extracted don’t fall in the standard white-paper-black-text category
- When accuracy over 99% per field is required
- When extraction in diverse and tough conditions is required (OCXR tools are able to recognize characters with stunning accuracy even in blurred or dark images and photos taken from a distance)