OCR Online: Extract Text from Images and Scanned PDFs

By FileConvertLab

Published:

OCR reader converting scanned document image to editable digital text
Diagram showing scanned image with locked text being converted to editable text via OCR

OCR (Optical Character Recognition) converts images containing text — scanned documents, photos, screenshots — into actual editable text you can copy, search, and reformat. If you have a scanned PDF you can't select text from, or a photo of a receipt you need to transcribe, OCR solves it in seconds.

Use our online OCR tool to extract text from any image or scanned PDF — no account or software required. The guide below explains how OCR works, what affects accuracy, and which format to choose for output.

How OCR Extracts Text from Images

When you scan a document or take a photo of text, you get an image — a grid of pixels with no inherent text data. The computer sees shapes, not characters.

OCR software analyzes the pixel patterns and matches them against character templates. It identifies letter shapes, groups them into words, and reassembles the text in reading order. Modern OCR uses trained neural networks that handle varied fonts, sizes, and layouts.

The output is plain text (or a searchable PDF) where the original page layout is approximated. Tables, columns, and indentation are partially preserved depending on the tool and the document complexity.

What You Can Extract Text From

Scanned documents — contracts, invoices, forms, books, legal papers

Photos of text — business cards, receipts, whiteboards, signs, menus

Screenshots — error messages, software dialogs, website content, presentations

Scanned PDFs — PDFs where text is stored as images, not as selectable text

Old documents — historical records, typed letters, faxes

To check if a PDF is scanned: try selecting text with your cursor. If nothing highlights, the PDF is image-based and needs OCR. If text selects normally, it's already searchable.

How to Extract Text from an Image Online

Open our image to text tool

  1. Upload your image (JPG, PNG, TIFF, BMP) or PDF
  2. Select the document language for better accuracy
  3. Click Convert
  4. Copy the extracted text or download as Word / plain text Works in any browser including Safari on iPhone. No installation required.

How to Get the Best OCR Accuracy

OCR accuracy depends heavily on image quality. These factors have the most impact:

Resolution

300 DPI is the minimum for reliable OCR. Most scanner defaults are 200 DPI — increase to 300 before scanning documents you plan to OCR. Phone photos are usually sufficient if the text is sharp and fills most of the frame.

Low-resolution images (below 150 DPI or very small text) produce errors on similar characters: l/I/1, 0/O, rn/m, cl/d.

Contrast and lighting

Dark text on a white background gives the highest accuracy. Problems arise from:

  • Shadows across the page (common in phone photos of thick books)
  • Colored backgrounds with text in a similar shade
  • Faded or yellowed paper
  • Glossy surfaces causing light reflections

Skew and rotation

OCR performs best on straight, horizontal text. Pages scanned at an angle — even a few degrees — reduce accuracy noticeably. Most scanning apps include auto-deskew; enable it before processing.

Language selection

Always select the correct language before running OCR. The engine uses language-specific character sets and dictionaries. Running English OCR on a French document causes errors on accented characters (é, à, ç). For mixed-language documents, select the primary language.

Output Format: Text, Word, or Searchable PDF?

FormatBest forPreserves layout?
Plain text (.txt)Copy-paste, data extraction, translationNo — text only
Word (.docx)Editing, reformatting, sharingPartial — headings, paragraphs
Searchable PDFArchiving, keeping original appearanceYes — original image + invisible text layer
Searchable PDF is the best choice for document archiving: it preserves the
original visual appearance while adding a hidden text layer for search and copy. Choose Word
when you need to edit the content or reformat it significantly.

Extracting Text from Images in Other Tools

Google Docs (built-in OCR)

Upload an image or PDF to Google Drive, right-click it, and choose Open with → Google Docs. Google automatically runs OCR and opens the result as an editable document. This is free and works well for English text.

Windows Snipping Tool

Windows 11's Snipping Tool includes a "Text Actions" feature that extracts text from screenshots directly in the clipboard. Take a snip, click the Text Actions button, and copy the recognized text.

iPhone / iOS Live Text

On iPhone (iOS 15+), point the camera at text and tap the Live Text icon to select and copy it. This works for photos in the Photos app too — open any image, tap the Live Text button, and select the text you want.

Summary

OCR turns images of text — scans, photos, screenshots — into editable text you can search and copy. Accuracy depends mostly on image quality: 300 DPI, dark text on light background, straight pages, and the correct language selected. For one-off jobs use the online OCR tool — no install required, works on iPhone and desktop alike.

Related Tools

Image to Word (OCR) — extract text and save as editable Word document

Scanned PDF to searchable PDF — add a text layer without changing the appearance

OCR essentials guide — how OCR works and when to use it

Scanned PDF to Word — step-by-step OCR for scanned documents

Frequently Asked Questions

What is an OCR reader?

OCR (Optical Character Recognition) is software that reads text from images and converts it into editable digital text. Instead of seeing a photo of a page, OCR produces actual characters you can copy, search, edit, and reformat. It works on scanned documents, photos of signs, screenshots, and any image that contains text.

How accurate is online OCR?

For clear, well-lit scans of printed text, modern OCR is 99%+ accurate. Accuracy drops with handwriting (expect 85–95%), low resolution (below 150 DPI), unusual fonts, or images with text at an angle. For best results, use scans at 300 DPI or higher and straighten pages before processing.

Can I extract text from a scanned PDF?

Yes. A scanned PDF is essentially a collection of images — the text isn't actually stored as text data, which is why you can't select or copy it. Our OCR tool processes each page as an image and converts it to real text. The output can be a searchable PDF or a plain text/Word document.

What languages does the OCR support?

Our OCR engine supports 100+ languages including Latin-script languages (English, French, German, Spanish, Portuguese, Italian), Arabic, Japanese, Chinese (Simplified and Traditional), Korean, Cyrillic (Russian, Ukrainian), Hindi, and many more. Select the language before running OCR for best results.

Can OCR read handwritten text?

Basic OCR handles printed text only. Handwriting recognition (ICR — Intelligent Character Recognition) is a separate, harder problem. Our tool handles most neat block handwriting at reduced accuracy, but cursive or messy handwriting may not produce usable results. For handwritten text, AI-based tools specifically designed for handwriting recognition work better.

What image formats work with the OCR reader?

JPG, PNG, TIFF, BMP, and PDF. For best results: use 300 DPI or higher scans, ensure text is dark on a light background, and keep the image straight (not rotated or skewed). Photos taken with a phone work fine if the text is sharp and the lighting is even.

How do I translate text from an image?

First use OCR to extract the text from the image, then copy the extracted text into a translation tool. Our OCR reader produces plain text output that you can paste directly into Google Translate, DeepL, or any other translator. This is faster than trying to find an image translation tool.

Can I extract text from a screenshot?

Yes. Screenshots are standard PNG or JPG images — the OCR tool processes them the same way as scanned documents. This is useful for extracting text from error messages, software dialogs, presentations, or any screen content you can't otherwise select.

OCR Online: Extract Text from Images and Scanned PDFs