Photographs of receipts read into rows, loose scans merged into one document, and the details you did not mean to share taken out before it goes.
28 tools · 21 free with no account · 7 free to try with one
A photograph of a receipt is not a record. It is a picture of one, and it stays a picture until something reads the figures off it and puts them in a row with a date, a supplier and an amount. This is where most paperless projects stall: the paper is off the desk and in a folder of forty thousand images, and nobody can answer a question with it. The scanning was the easy half.
So the order is read it, file it, then check that what you filed is not carrying more than you meant. Reading is the part a person should not do by hand and the only part that uses a model. The merging, rotating, page numbering, redacting and metadata stripping all happen in your browser and send nothing, which matters more than usual here, because a pile of paper generally contains a bank statement, somebody’s ID and at least one thing you would not email.
1 · Read the paper
Photographs and scans turned into rows with figures in them.
That is what the scanned-invoice and receipt readers are for: a crooked, shadowed photograph of a faded thermal receipt, handled as an image rather than as text. They are slower and cost more per document than reading a file that is already digital, so use the plain extractor where you have a proper PDF and keep these for the paper.
What actually leaves my machine?
Only what the reading tools send, and each shows you exactly that before it goes: the page image or the text, with personal identifiers masked on your device first. Merging, splitting, rotating, redacting and stripping metadata send nothing at all, which is why they carry on working with the network off.
Is blurring the same as deleting?
Not always, and the difference has embarrassed people. Pixelated text has been recovered by researchers — the process is deterministic and there are not many possible characters — so use a solid block for anything that matters. A black box drawn over a PDF page in most editors is worse still, because the text is untouched underneath and can be copied straight out from under it. If a page has to be safe, remove the page or flatten it to an image, and check what the file is carrying besides: the author, the software and often an earlier filename.
