Recognize text in scanned PDFs (OCR)
A scanned PDF contains only page images, so you cannot search, select or copy its text. Recognize Text (OCR) reads the words on the page and adds them as a searchable text layer.
Requires Pro edition or higher
OCR runs fully on your computer with the bundled Tesseract engine. No internet connection or upload is needed. Speed depends on your CPU, memory and the number of pages.
Choose an OCR command
Open the scanned PDF, then select Convert › OCR. The menu offers four commands:
- Quick OCR recognizes the whole document and shows the text in a panel, ready to copy or save.
- Recognize Current Page recognizes the page you are viewing.
- Recognize All Pages recognizes every page of the document. This takes longer.
- Searchable Layer writes the recognized text into the PDF. It becomes available after recognition.
Recognize text on the page
- Select Convert › OCR › Recognize Current Page, or Recognize All Pages.
- The Recognize Current Page panel opens on the right and recognition starts. Progress is shown in the panel and in the background task list.
- When OCR is done, a message shows how many words were recognized. Each word is outlined on the page.
- To correct a single word, point to it on the page and edit the text in the box that appears.
The OCR section of the panel holds the recognition settings. They apply to the next run:
| Setting | What it does |
|---|---|
| OCR language | English + Bahasa Indonesia (default), English or Bahasa Indonesia. Choose the single language when the document uses only one. |
| Select OCR region | Drag a rectangle around text to recognize only that area of the current page. |
| Review OCR words | Opens the review panel described below. |
| Straighten text before OCR | Corrects a slightly tilted scan before reading it. On by default. |
| Text orientation | Auto detect, As displayed, or a fixed rotation of 90°, 180° or 270°. |
| Reduce scan shadows | Evens out dark edges and shadows from photographed pages. On by default. |
The same settings are available from Options in the status bar.
Review uncertain words
OCR can misread small text, stamps and numbers. Review the results before you make them part of the PDF.
- In the OCR section, click Review OCR words. The Review OCR words panel opens.
- Under Confidence below, choose 50, 70 or 90, or All words. Words under that confidence are listed.
- Under Review focus, choose Uncertain words, Numbers and identifiers or Possible names.
- The current word is highlighted on the page. Its page and confidence are shown above Corrected text.
- Fix the text if needed, then click Save & next or press
Enter. Use Previous and Skip to move without changes. - When you are done, click Apply searchable text.
Make the PDF searchable
Click Apply searchable text in the review panel, Searchable in the recognition panel, or Convert › OCR › Searchable Layer. The same command is on the Accessibility tab as Searchable Layer.
The recognized words are added as an invisible text layer behind the scan. The page looks the same, but you can now search, select and copy the text. Save the document with File › Save or Save As to keep the layer.
A searchable layer also lets you export the scan to Excel or text formats, and lets screen readers and Read Aloud use the text. See Convert PDFs and Accessibility and local AI tools.
Quick OCR
Quick OCR opens the OCR PDF panel for the active document. It is useful when you only need the text.
- Under Language, choose English, Indonesian or Both.
- Click Run OCR. Pages are processed one at a time.
- Read the text below the buttons, grouped by page.
- Click Copy, Save .txt, or Save text PDF.
Straighten a crooked scan
Select Convert › Deskew Scan. Farino PDF measures the tilt of each page and straightens it. The result opens in a new tab with "-straight" added to the file name.
Pages tilted more than 6 degrees are left unchanged, to avoid over-correcting artwork. The straightened pages are rebuilt as images, so any existing text layer is not carried over. Run OCR on the straightened copy.
Turn a scanned table into Excel
Select Form › Scan Table to Excel with the PDF open. Farino PDF reads the columns of each page and writes real spreadsheet cells to an .xlsx file. Pages without text are recognized with OCR automatically, in English and Indonesian. When the workbook is ready, the Windows Save As dialog opens.
Check the totals and numbers in Excel. Faint lines, merged cells and handwriting can shift values into the wrong column.
Tips for better results
- Scan at a resolution where the text is sharp. Low-resolution or blurred scans produce more errors.
- Choose a single OCR language when you can. It is faster and more accurate.
- Use Select OCR region for a single paragraph, a stamp or a table cell.
- Large documents take time. Recommended memory is 8 GB for large documents and OCR.