When your Jæren Sparebank statement arrives as a scan or a photo of the page rather than a clean export, the numbers on it are just pixels until they are extracted. Statements that arrive as a scan or photograph present a different challenge from clean digital files. The text is an image of characters, not stored data. The engine applies optical character recognition to read every line, then verifies the running balance against the statement’s own printed closing figure. A single misread digit, a 3 mistaken for an 8, will cause the balance to break; the row is flagged for review rather than silently accepted. The layout detection works the same way it does for a native file, so the conversion path is identical once the text is lifted from the image.
If a transaction description on your Jæren Sparebank statement is long enough to wrap across two or three lines, it still needs to land in a single row once converted. Transaction descriptions that wrap across two or three lines inside the statement can break a naive extraction: the second line looks like a new row with no date or amount. The engine recognises wrapped text by the missing value column on the continuation line, then re-joins it to the parent row before writing the export. The result is one transaction with its full description intact, mapped cleanly into Excel, CSV or QBO, with no split rows to repair afterwards.