Compare scanned documents, even without a text layer
Two scanned deeds without a text layer cannot be compared with ordinary tools. CompareDesk opens up their content with text recognition that runs entirely on your own computer, and then compares the recognised wording.
In three steps
- Drop both scans into the fields. CompareDesk detects for itself that there is no text layer.
- Text recognition runs without further input; it needs no internet connection and transmits nothing.
- Compare. The result reports the differences and states on how many pages the text had to be recognised first.
What you get
The comparison of the recognised wording, together with an explicit note on its origin: recognised text can contain reading errors, and CompareDesk says so at every point where that applies.
Recognisable reading errors are treated as such and are not reported as substantive changes. Where the source was read with uncertainty, the passage is marked instead of asserting a change that does not exist.
Pages from which no text can be obtained even with recognition are left out of account and named.
What the comparison does not do
The quality of the result depends on the quality of the scan. At low resolution, with skewed feeding or a faint typeface, reading errors remain; review on your own responsibility is therefore particularly important with scans. CompareDesk does not conceal that limit.
Try it on your own scans
CompareDesk recognises and compares solely on your own computer. Neither the image nor the recognised text leaves the device.
Download for free To the user guide
Also possible: Compare PDF files Compare documents Deed comparison for notaries Comparison certificate and checksums Glossary of document comparison