Banking / Technical

ICR

Also known as: Intelligent Character Recognition, Reconocimiento Inteligente de Caracteres

Definition

ICR (Intelligent Character Recognition) is an advanced variant of OCR that uses AI models to recognize handwriting and non-standard typefaces. Useful for documents with handwritten annotations.

LATAM context

Pure digital statements almost never need ICR, but physical bank forms or legacy documents with handwritten notes are where ICR makes the difference between extracting the data or not.

Concrete example

A branch deposit voucher with a hand-written amount gets processed with ICR to pull the amount out as digital, numeric data.

How it shows up on your bank statement

Like OCR, ICR is part of the processing, not something that shows up in the document itself. It becomes relevant when a statement carries handwritten notes (a received-by signature, a margin note from an accountant) that printed-text recognition can't read.

How does finO$ handle this?

finO$ uses ICR in edge cases when it detects handwritten areas on scanned documents.

Related terms

OCR

Need to convert bank statements into structured data?

Start free Go to my dashboard

See the API reference

Just one file? Use the free converter to convert your bank statement to Excel.

Frequently asked questions about ICR

What's the practical difference between OCR and ICR?

OCR is trained on printed characters, which are uniform; ICR attempts handwriting, which varies person to person. On bank statements almost everything is printed, so OCR handles most of it and ICR only comes in for the occasional handwritten note.

Does ICR help read a hand-written bank statement?

Bank statements aren't issued by hand, so that case doesn't really come up. Where it does help is supporting documents attached to a file: receipts, vouchers or notes that accompany the statement.