Optical character recognition or OCR is a method that ensures the processing of typewritten text from an electronic image. The so-called OCrKO is a widely used technology today. In fact, this technology is free. Here I dare to use the hackneyed phrase “what is worth nothing, is worth nothing”. What the OCRKO must meet today to be applicable in practice:
- Receive an electronic image from an email or file system
- Run OCR technology over documents
- Deliver the result to pass to the system, which creates metadata from the characters to the document
- Create searchable PDF and metadata in standard format (xml, csv, txt)
- It must be a server instance
- User-accessible, for possible correction, or more precisely, correction of OCR errors or erroneous information on the document
- It must not license the user.
So what is the success rate of OCR today?
.. OCR success rate from InfoMatic on the software platform MetaServer is +95%.
Proof instead of promises:
![](https://cdn.prod.website-files.com/621e047d28ca81795b58c57b/643f9736b6339330180da220_metaserver-ocr-test.png)
Quality OCR ensures document automation. The system correctly assigns the metadata that defines the document. Thus, they give the documents a “soul” and it is possible to further work with documents and search for them. Every core system of a company or organization, i.e. BPM /DMS/ERP/CRM needs documents that contain metadata. And you either give metadata to the core system manually or automatically.
![](https://cdn.prod.website-files.com/621e047d28ca81795b58c57b/643f97765ecc6b9ecd510393_ocr-lekarsky-predpis-recept-metaserver.png)
The right way to go digital is consulting with us to find solutions and not look for technology.
David Večera
david.vecera@infomatic.cz