Converting Scanned Documents to Searchable Documents
Background
You probably know that our court's standard filing preference number two
indicates a strong preference for most documents filed in ECF to be converted
rather than scanned. The reason our judges issued this preference has to do with
text searchability. Documents scanned to PDF for filing are not searchable–that
is, you cannot search using Acrobat's Find tool (Ctrl+F or Edit>Find), nor can you
cut and paste text. However, Adobe Acrobat features an Optical Character
Recognition (OCR) tool which can, in many cases, convert a scanned PDF
document to a searchable PDF document.
When is OCR Necessary?
The first step to understanding and using OCR is to recognize when a document
is not already searchable. With a scanned document open and the Select tool
selected, if you click anywhere in the text, an unsearchable document will not
allow you to select any text OR it will turn blue, as shown here:
How to Use Adobe’s OCR
Here are the steps to use Adobe’s OCR Text Recognition Tool.
STEP ACTION
1 With the scanned document open, click Document on the
Acrobat menu bar.
Converting Scanned Documents to Searchable Documents
Page 1 of 3
STEP ACTION
2 Select OCR Text Recognition.
3 Select Recognize text using OCR.
4 The Recognize Text dialog box opens. Click OK.
Warning: As the OCR process is underway, a black status
bar will appear in the lower right corner of your Acrobat
window. Do not click in the document until the status bar is
gone. Interrupting the OCR process may lead to errors.
Watch for Conversion Errors
As helpful as OCR can be, it is not a perfect tool: Some documents will not
convert accurately. For example, a PDF document heavily laden with graphics or
a PDF document that has been copied, faxed, or scanned a number of times
may become unreadable to OCR. Here is an excerpt from one such document:
This is how OCR interpreted the same text:
Always be sure to proofread all text if you have used the OCR tool.
Handwriting May Not Convert
While handwritten PDF documents will not usually translate well through OCR,
typed text surrounding the handwriting will. Here is a scanned PDF document
with a handwritten date:
Even though pages on a site may have been updated, you may be viewing old information if your browser’s
cache (pronounced “cash,” a type of electronic memory) is not being cleared as frequently as it should be.
Even though pages on a sire may have Peen updated, you ma)4:l,e viewing old information if your browser’s
cache (pronounced “caSh: 11 type of electronic memory) lsmHletng cleared as ~. frequently as it stould be.
Converting Scanned Documents to Searchable Documents
Page 2 of 3
This is the same text as interpreted by OCR:
In this example, notice how OCR was able to recognize the text even though it
could not accurately recognize the handwritten characters.
ORDERS that the time in which SunTrust Bank may answer or otherwise respond to the
Complaint be extended through and including January 13, 2011.
ORDERS that the time in which SunTrust Bank may answer or otherwise respond to the
Complaint be extended through and including January 13, 2011.
Thi~oflJ~ $J},i)
Converting Scanned Documents to Searchable Documents
Page 3 of 3
