A scanned PDF stores page images obtained from a scanner or photographs. Without an added text layer, words inside those images cannot be searched or copied as ordinary text.
Distinguishing a scan from a text PDF
Try selecting and copying a line and searching for a visible word. If you can only select an image or copying produces no readable text, the page may need recognition. Check several pages: a mixed PDF can contain both ordinary text and scans. Assess an existing text layer by the quality of the copied text as well.
Preparing for recognition
Use the original scan rather than a screenshot when possible. Check page orientation, small-print readability, cropped edges and glare. Simply enlarging an image cannot undo blur or heavy compression. Select the document language for OCR.
Choosing an expected result
Select an available output that matches your task: text for editing or a table format for rows and columns. The converter's direction list shows the supported choices. Recognition does not promise an exact copy of the layout; pictures, captions, columns and tables may need separate review.
Checking important information
After downloading, compare the number and order of relevant pages, headings, dates, amounts and identifiers with the original. If a table becomes plain text or its cells are grouped incorrectly, check the table conversion and the source quality. No detected text does not prove that the document is blank: the image may be unreadable.