A Medilink member is currently seeking a software solution that can be used to analyse the pages of individual digital documents and classify them according to their elements.
The categorization should then be able to assign different printing configurations to different categories of document pages without human intervention.
Interesting solutions should be able to:
1) Correctly identify images and extract any text from documents using Optical Character Recognition (OCR), or alternative technologies, e.g. page pixel analytics;
2) Derive significant features from the document, both text-based (e.g. word frequency or semantic analysis) and image-based (e.g. colour histogram statistics or co-occurrence matrices);
3) Be fed a sample group of already categorized document elements, derive their significant features and estimate patterns for each type of elements;
4) Be further fed uncategorized documents and correctly identify them, assigning them to different printing workflows accordingly

For further information about this opportunity, please contact SueW@MedilinkWM.co.uk