Improving Acrobat OCR accuracy for lines in the document (e.g. blank lines for signatures or table grid lines)

Question

I am using Acrobat Pro. I often use the OCR function on documents that I get from clients as photos. Often, these documents include lines, such as underlined text or lined spaces for a signature or grid lines as part of tables. Whether Acrobat does its OCR with the “Scan & OCR” tool or the “Edit PDF” tool, it consistently treats these lines as part of background imagery. It does an okay job recognizing underlining of text (but not better than okay) and pretty much never recognizes blank lines as pseudo text (in the case of spaces for signatures) or formatting (in the case of boxing in and around tables).

Often, the images I get from clients including yellowing or low-res smearing in the white background of the document. OCRing turns them into editable images. I would like to be able to select and delete these image background, but because Acrobat’s OCRing includes the blank lines and table lines described above as included in the background image, deleting the background fuzz also causes these “real” and necessary lines to disappear. I’m left keeping the background fuzz in place, resulting in larger file sizes and dirtier documents.

Is there a way to improve this with settings at my end? Is this something that needs fixing at Adobe’s end? Thanks.

Anand Sri Bhattacharya · Accepted Answer

Hello @jonathan_2510,I hope you are doing well, and thanks for sharing the details. We're sorry for the trouble you had.To make sure I cover what you need:Which OS platform and version of Acrobat Pro are you using?Are you running OCR via:A) All Tools > Scan & OCR > Recognize Text, orB) All Tools > Edit PDF (which triggers OCR implicitly), orC) Both (same result in each)?Is your primary goal to:A) Preserve lines as vector content (so they survive background cleanup), orB) Improve table recognition/editability, orC) Enable background cleanup without losing signature/table lines?Assuming you are on the latest build of Acrobat: 26.001.21563. Planned update, May 18, 2026. Acrobat’s OCR is optimized for text characters. When you run OCR, Acrobat “analyzes the image and replaces text bitmaps with searchable text”, but non-text elements like lines or borders are left as part of the image layer. In other words, blank lines or table rules are not converted into separate vector lines or underlined spaces; they remain embedded in the scanned image for visual fidelity.Please note that, unfortunately, there isn’t a user-accessible setting in Acrobat to improve this. All the OCR settings (language, scan resolution, output style) focus on text and do not include options to interpret or preserve drawn lines. So, under current functionality, you can’t force Acrobat to identify those lines as text or shapes; it will always consider them part of the background image. To learn more, please check this Adobe article: Fix text issues in scanned PDFs.Suggestions:While you can’t change how OCR classifies lines, you can optimize input quality to get slightly cleaner results:Pre-process the scan to reduce noise: Before OCR, use Tools > Scan & OCR > Enhance (or enable “Background Removal” in custom scan settings) to clean up uneven backgrounds. This filter lightens speckles and discolouration, potentially preserving true lines while dropping only the unwanted “fuzz”. Check this article to learn more: Scanned PDF settings.If you control the scan/photo, scan in black & white at high DPI (Adobe recommends ~600 dpi for monochrome) so lines are as crisp as possible. Clear, dark lines on a white background are less likely to be mistaken for noise. OCR output style: If not already, use the default “Editable text and images” mode (Acrobat’s standard when you click Edit PDF). This mode at least attempts to retain formatting; lines under detected text might carry over as actual underlines.Right now, there’s no Acrobat setting to have OCR treat drawn lines as text or separate objects. The workaround is to optimise scan quality to help keep those lines visible and manually remove unwanted background noise. You can use the Acrobat Wish form to file a feature request with the product team: https://adobe.ly/4u9bQJlI hope this helps, and let us know if you need any assistance.Regards,Anand Sri.

Sign up

To post, reply, or follow discussions, please sign in with your Adobe ID.

Sign in to Adobe Community

To post, reply, or follow discussions, please sign in with your Adobe ID.

Scanning file for viruses.

This file cannot be downloaded