Adobe OCR Text Recognition Distorting Signatures in the PDF

[thepianoman87]

As of about 3 months ago or so, when I open a document and perform OCR/text recognition, Adobe attempts to recognize the characters in a signature, thereby disorting and ruining the signature field, when all I really want to do it modify some of the content of the letter/document above (i.e. change the date at the top). I'm not sure why it does this only now, but it has rendered Adobe Acrobat essentially unusable for my line of work, and I cannot find any documentation online to support the issue I'm having.

Does anyone know if there is a setting or way to tell adobe not to auto-recognize certain bounding blocks? Or undo text recognition on certain bounding blocks? Or change the sensitivity so it only recognizes typed fonts?

Thanks.

[Bernd Alheit · Accepted solution]

There are no such settings.

[Luke Jennings]

I can suggest a work-around- go to Tools> Edit PDF and select the signature, go to Edit> cut (or copy). Run the OCR and when done go to Edit> Paste. If this doesn't work, you can edit the signature and open it in Photoshop and save it to a new file, then place it back into your PDF after OCR. Other options might be to move the signature to a new layer and hide or lock it.

[thepianoman87]

Luke Jennings I like the idea of moving the signature to a new layer- how is that accomplished?

[Luke Jennings]

The easiest way to make a PDF layer is in the application that created it, InDesign for example. To add a layer to an existing PDF, you can go the layers icon (View> Show/Hide> Navigations Panes> Layers, click on the little options box and select Import as Layer and navigate to the signature image you saved from Photoshop and place it into the PDF. Note, not all PDF viewers respect layers.

I think the cut/paste option might be your best bet. I don't do much OCR and don't have time to do any testing at the moment.

There is a trick to edit the object (signature) and replace it with the original image (that you previously saved with the exact same name) but it's been a while since I used it and don't recall the exact steps, perhaps it will come to me, or someone else might chime in.

Do you have PitStop? (an Acrobat plug-in)

[Susan5D11]

Open your document, go to Tools> Scan & OCR> Open> Recognize Text> In This File> Recognize Text.

[TrumanBurbank]

I'm having this problem in a document that was signed in hard copy then scanned into a pdf. When I OCR scan the document in order to add some required information below the signature, all typed text in the document scans correctly and appropriate text blocks are created for all text. However, the cursive signature in partially obscured by a rectangular white space that is not selectable after the OCR scan, i.e., it doesn't appear to be a text block, just rectangular damage to the signatire pixels. The rectangle is not a text block nor is it related to any of the nearby text blocks.  

[Abambo]

Don't do an OCR, add just your information as required. I've never seen the phenomena as you describe it, but the PDF OCR option was never thought to modify any text in the document, but to enable scanned PDF data to be searchable.

So, I'm wondering why you get that rectangle above the signature. At most, I would expect gibberish below the signature.