Perhaps it is an image rather than text. What OCR options did you use?
You state that your PDF is an "OCR scanned" PDF.
I'll take that to mean the process active between paper to scanner and output PDF involved OCR of the image output from the scanner.
For such the OCR would be Searchabe Image or Searchable Image (Exact).
The PDF will have both the image and the OCR output.
So, yes, there is an image on each page with no alternate text.
If the document is in fact OCR scanned and contains readable text (as CT Dave mentions) and you are still getting this error, you could convert each image to an Artifact. This will enable your document to pass the built-in accessibility test in Acrobat X, but it will NOT necessarily make the document accessible or 508 compliant. Among other things, you need to ensure that the OCR is accurate, that the text is tagged and in the correct order, and that it in fact "reads" when the document is accessed using Assistive Technology (i.e., a screen reader).