If you’ve ever tried searching through a scanned PDF or attempting to edit text inside of it, you know that it’s difficult — if not outright impossible — to search for a word inside a scanned PDF.
This is because a scanned PDF doesn’t store the actual words as text. Instead, they are stored as images (or pages), which means assistive technologies like screen readers cannot read them either. Worse yet, users can’t select text inside the document, nor resize or reflow text on the page. And while you can change the background color of a PDF document, you can’t edit the text in it, nor change its text color.
The authors of the document also have limited ability to manipulate the content of the PDF file for accessibility; the PDF itself is static, unchangeable, and not searchable by default.
But what if we could convert those images into selectable, editable, searchable text? That’s where OCR (optical character recognition) comes in handy.
Optical character recognition converts visual characters (like letters, numbers, and symbols) into stored machine-readable text so you can then do things like copy, highlight, and rewrite. In fact, converting non-selectable PDF files into selectable and searchable PDFs has high accuracy with optical character recognition, and turns those static images into fully functioning text.
In short, OCR unlocks many capabilities within a scanned PDF. You can now perform search-and-replace functions, and extract words from a PDF file to use in another application. You can even make an image or scanned document searchable by adding OCR.
How online OCR tools work
All online OCR tools follow a similar upload-process-download workflow. First, you’ll need to choose which PDF files to upload. Then, you’ll need to modify the settings. Once processing completes, you’ll be able to download your resulting searchable PDF.
Free online OCR tools typically require no software installation, and often don’t require registration. The work is done on cloud servers, rather than your own computer.
For example, PDF24 allows you to select the files you want to process, modify the OCR settings, and then download your new searchable PDF. Adobe Acrobat uses a simple drag-and-drop upload workflow; Acrobat will scan text in the uploaded PDF, and after a few moments, the PDF will appear as downloadable content or shareable via link. iLovePDF lets you upload a file, then recognizes the text automatically. Sejda’s free service is quick and easy, too. Smallpdf is accessible docs in seconds without downloads, signups, or complicated settings.
Many online tools offer several different options for how you’d like the processed PDF to be saved, which may affect the output quality and format. For example, selecting the correct language increases the accuracy of detection. PDF24 provides some great options, including whether you’d like to save the PDF as a PDF/A, remove artefacts and noise, deskew pages, set meta information, or join all of the output files together into one single file.
Accessibility and OCR
OCR satisfies the requirements under WCAG technique PDF7, which requires performing OCR on a scanned PDF in order to provide actual text for reading. Performing OCR fulfills the success criteria 1.4.5 and 1.4.9 (Images of Text). If you’re using Adobe Acrobat Pro, you can add accessibility tags manually. Acrobat Pro may automatically add tags when the file is run through OCR.