How are the PDF's generated? Are they scanned, or natively generated somewhere else? I've found that different PDF generators can write PDFs that fool with Laserfiche's ability to extract text.

For example, if I build a report in SSRS and add a field before I add the field label, SSRS renders the field "before" the label in the structure of the PDF. Since PDF was originally designed as a typesetting format, a writer can place elements in any order. The positioning is defined for each element. The result is that when I use Laserfiche's native text extraction for PDF, my field label actually shows up in the text after the field data itself. This is just one example of some of the weirdness that can happen when extracting text from PDFs.

If the PDF is an image, I don't have a ready explanation for the behavior that you are seeing. However, if it's generated, you might try changing your settings for generating text from PDFs and see if it makes a difference. How are you bringing the documents in? Are the users importing them directly into the client, or is it Import Agent?

Here are the settings for the desktop client, and they are similar in Import Agent:

As far as the certificate goes, we've found that the engraving around the edges of certificates tends to mess with things like orientation detection and OCR. You either need to clean them up manually, or run them through Quick Fields so that you can rotate and Zone OCR based just on the center portion.