How to Apply OCR to Only Part of a Scanned PDF
Question details
The user wants to apply Optical Character Recognition (OCR) only to specific text areas of a scanned PDF to avoid altering or distorting charts, diagrams, and other graphics.
- Product
- PDF Editor
- Device & OS
- not provided
- Scenario
- Performing text recognition on a scanned document that contains a mix of standard text and complex graphical elements.
- Observed behavior
- Standard OCR processes the entire page and may attempt to convert images and charts into text or editable objects, changing their original visual appearance.
Identify the exact pages or visual sections containing the charts and diagrams you want to preserve, as you will need to adjust your OCR output settings or isolate these areas before running the recognition process.
Use Searchable Image Mode for the Entire Page
This is the safest method to ensure charts and diagrams remain completely unaltered while still making the text selectable and searchable.
Most PDF editors, including Adobe Acrobat, apply OCR to the entire page by default. To stop the software from changing the visual appearance of your graphics, you should change the OCR output style.
Open your scanned document in your PDF editor and navigate to the OCR or Text Recognition tool.
In the OCR settings menu, look for the output style or PDF output option. Change it from 'Editable Text and Images' or 'ClearScan' to 'Searchable Image' or 'Searchable Image (Exact)'.
Click 'Recognize' or 'OK'. The software will place an invisible layer of text behind the scanned image, keeping your original diagrams looking exactly as they did.
Isolate Text Areas Using Crop and Combine
Use this method if you strictly need editable text output but want to protect specific graphic regions on the same page.
Accurately OCR Scanned Documents with WPS PDF
WPS PDF features an advanced OCR engine that lets you extract text from specific areas or convert entire scanned pages into searchable documents without ruining your original layout and charts.
- 1. Open Scanned PDF: Launch WPS Office and open your scanned document in the WPS PDF editor.
- 2. Access the OCR Tool: Navigate to the 'Tools' tab on the top ribbon and click on the 'OCR' button.
- 3. Choose Area Extraction: If you only need text from a specific paragraph, select the 'Screen Grab OCR' or 'Extract Text' option and drag a box strictly around the text, avoiding the diagrams.
- 4. Export or Save: Copy the extracted text to your clipboard or choose 'Save as Searchable PDF' to add an invisible text layer to the entire page while keeping graphics intact.
Frequently Asked Questions
Why do my charts and diagrams look distorted after running OCR?
When OCR is set to 'Editable Text' or 'ClearScan', the software attempts to convert not just text, but all visual elements into editable font characters and vector paths. Complex images and charts often confuse the engine, resulting in visual distortion.
Can I manually draw a box to apply OCR only to that area?
Many standard PDF editors process the whole page by default. However, using tools like the 'Screen Grab OCR' feature in WPS PDF allows you to draw a selection box and extract text exclusively from that highlighted area.
What is the difference between 'Searchable Image' and 'Editable Text' in OCR?
Searchable Image leaves the original scanned picture exactly as it is and adds an invisible layer of text over it for searching and copying. Editable Text completely replaces the scanned image with newly generated font characters, which can alter the layout.
