Turnitin Rejected My Scanned PDF: Why Image-Based Files Fail Submission

A scanned PDF (or any image-based file) fails Turnitin submission because Turnitin requires extractable, selectable text — and a scanned page is an image, not text. Turnitin's own support page states the file must contain at least 20 words of extractable prose text, and that scanned documents, image-based PDFs, forms, or portfolios are automatically rejected. The fix is documented too: convert the image into text (OCR) before submitting, or paste the text directly into the submission box. Changing the file format without converting the image does not help.

Turnitin's support page "Why am I unable to submit my paper to Turnitin?" lists the technical rules for files that should generate a Similarity Report:

"Supported file types: Your file must be in a supported format (e.g., .docx, .pdf, .txt, .rtf)."

"Size & length limits: The file must be under 100 MB and fewer than 800 pages."

"Minimum word count: The document must contain at least 20 words of extractable prose text."

"No images or scanned text: The text in your document must be selectable and copyable. Scanned documents, image-based PDFs, forms, or portfolios will be automatically rejected."

The key word is extractable. The text needs to be selectable with your cursor and copyable — the pipeline reads the text layer of the file, not the pixels.

The most common error for this failure is documented in Turnitin's help center:

"The error message 'Your submission must contain 20 words or more' indicates that the file you are attempting to submit does not meet the minimum requirement of 20 words of extractable text."

"This commonly occurs when the submitted document does not contain enough extractable text. For example, the file may consist primarily of images rather than selectable text, or it may contain fewer than 20 words."

The official self-check is a copy-and-paste test:

"Try highlighting the text in your document and copying and pasting it into another application. If you cannot select or copy the text, the document may contain images instead of extractable text."

If you cannot select and copy the text, the file is an image from the pipeline's point of view — regardless of how good the scan looks to a human.

Turnitin's submission error page names the issue directly:

"If you scanned the document you are trying to submit, the document is likely an image. In order to submit a document that has been scanned, you must use a program to convert the image into text."

The same page explains the underlying check:

"The file you are trying to submit does not contain at least 20 words or 100 characters of text."

So the fix is not "export the same scan as a different format." The fix is converting the image into a text layer — OCR — or pasting the text directly into the submission box.

Turnitin's file requirements page groups related rejections together:

"Turnitin will not accept PDF image files, forms, or portfolios, files that do not contain highlightable text (e.g. a scanned file - usually an image), documents containing multiple files or files created with software other than Adobe Acrobat."

The same guidance recommends how PDFs should be created:

"Turnitin recommends using Adobe Acrobat or Microsoft Word 365 to create PDF files. PDFs generated by other applications may not always be processed successfully."

Related failures that look different but come from the same root:

• First page contains an image and no text — the file can fail to process entirely.
• Files with dense vector images can take too long or fail to process.
• Password-protected files are rejected.

All of these are "the pipeline cannot read text the way it needs to."

Turnitin's own guidance gives the workable path, in order:

1. Check whether the file has extractable text: select and copy part of the text into a plain-text editor. If nothing copies, the file is an image.
2. Convert the image into text with an OCR program, then proofread the result — OCR can introduce character errors.
3. Re-submit the converted, text-based file (or paste the text directly into the assignment's text input).

The boundary that matters: converting the format (e.g. PNG to PDF) does nothing if the text is still pixels. The text must exist as selectable characters before the pipeline can compare it.

A scanned PDF fails Turnitin submission for a documented reason: the pipeline requires at least 20 words of extractable, selectable text, and a scanned page is an image. Turnitin's own pages state the minimum, name the automatic rejection of image-based PDFs, and give the official self-check: if you cannot select and copy the text, it is not text to the pipeline. The fix is converting the image into text (OCR) and proofreading it, or pasting the text into the submission box — not re-exporting the same scan in another format. Understand the rule, check your file the official way, and the error stops being a mystery.