A while ago, I was working on a project, and the way error was being displayed, I could not copy the text directly to paste it in ChatGPT. I had to type the whole error message manually, which was a bit frustrating. I thought, what if I could just take a screenshot and extract the text from the image directly? That’s when I decided to enhance GNOME Screenshot with OCR. I mean windows snipping tool has this feature, why not GNOME Screenshot?
That’s one the perks of using Linux, I can just make something for my needs and share with others, someone who needs can also use, someone who doesn’t need can just ignore it. So, I started working on this project, and I am happy to share that I have successfully enhanced GNOME Screenshot with OCR. In this article, I will share how I did it and how you can also use it.
The code is available here with MIT license: github.com/funinkina
π Prerequisites
I had strict requirements for this project, I wanted to use only open-source tools and libraries, with minimal dependencies, so other’s don’t have to install a lot of things to use this. Also it needed to simple and single file, so it can be executed easily. Here are the tools and libraries I used for this project:
- Python 3.13
- Tesseract OCR (Only thing you need to install)
- GTK 4.0
- Adwaita 1.0
- XDP
π¨ Design Overview
The design of the project is simple, I have used Python to create a script that takes a screenshot using XDP (xdg-desktop-portal), then it extracts the text from the image using Tesseract OCR, and then it displays the extracted text in a dialog box using Adwaita and GTK. The script is designed to be executed using a keyboard shortcut, so you can take a screenshot and extract the text from the image with just a single key press.
The script follows a modular and event-driven design with the following components:
- Graphical Interface β Uses GTK 4 for the text display and user interactions.
- Screenshot Handling β Utilizes Xdp.Portal to capture screenshots interactively.
- OCR Processing β Uses pytesseract (Tesseract OCR) to extract text from the image.
- File Management β Saves extracted text to a file if requested.
- Clipboard Integration β Copies the extracted text to the clipboard.
ποΈ Architecture Breakdown
The script is object-oriented, with the following key classes:
π₯οΈ 1. TextDialog (UI for Extracted Text)
Displays the recognized text in a Gtk.TextView inside a scrollable window.
Provides “Save to File” and “Copy to Clipboard” buttons for user actions.
Uses Adw.ToastOverlay for improved UI experience.
π 2. GnomeOCRApp (Main Application)
Handles the application lifecycle and integrates:
Screenshot capture via Xdp.Portal.take_screenshot().
OCR text extraction via pytesseract.image_to_string().
GUI window management using GTK 4.
βοΈ 3. argparse (CLI Arguments)
Supports optional flags:
--enablesaving: Keeps the screenshot after extraction.--nocloseonaction:Prevents the app from quitting after copying or saving.--lang <languages>: Specifies OCR language(s).--save-location <path>: Defines a default save directory.
β¨ Key Features & Design Considerations
β Minimal UI Footprint β The main window is invisible, and only the extracted text is displayed.
β Flexible Text Handling β Users can edit the extracted text before saving/copying.
β Language Support β The OCR language can be customized via –lang.
β Clipboard & File System Integration β Text can be saved or copied seamlessly.
β Automatic Cleanup β The script deletes temporary files unless explicitly saved.
π Here’s a demo screenshot of the script in action:
π οΈ Installation & Usage
You can find the installation instructions and usage guide in the README file of the project repository: Gnome-OCR-Screenshot
Please star the repository if you find the project useful and feel free to contribute to the project by creating issues or pull requests.