Screenshot Search Engine

A fully offline, Python-based desktop application that automatically indexes your screenshots and allows you to search through them using natural language or keywords via OCR.

(Preview image placeholder)

Features

• Fully Offline: No API keys, no cloud dependencies. Your data stays on your machine.
• Background OCR: Extracts text from images using PyTesseract without freezing the UI.
• Smart Search: Combines exact keyword matching with fuzzy matching to tolerate typos.
• Auto-Indexing: Uses a file watcher to automatically index new screenshots added to watched folders.
• Duplicate Detection: Uses Perceptual Hashing (pHash) to prevent indexing exact duplicates.
• Modern UI: Built with CustomTkinter for a sleek, dark-mode ready interface.

Architecture

This app uses a clean architecture pattern:

• GUI: CustomTkinter handles the modern look and feel.
• Database: SQLite3 stores image paths, metadata, hashes, and OCR text.
• OCR Engine: PyTesseract handles text extraction. OpenCV is used for image preprocessing (grayscale, Otsu's thresholding) to improve accuracy on UI screenshots.
• Background Workers: watchdog monitors folders, while a threaded queue processes images sequentially.
• Search Engine: thefuzz ranks results based on filename exact matches, OCR text exact matches, and fuzzy token ratio scores.

Prerequisites

Before running the application, you must install Tesseract-OCR on your system.

Installing Tesseract-OCR (Windows)

1. Download the installer from UB Mannheim.
2. Run the installer.
3. Ensure Tesseract is installed in a standard location like C:\Program Files\Tesseract-OCR\tesseract.exe. The app will attempt to auto-detect it.

Installation (Source)

1. Clone this repository or download the source code.
2. Install Python 3.9+ and pip.
3. Install the dependencies:

   pip install -r requirements.txt

4. Run the application:

   python app/main.py

Packaging as an Executable

You can build a standalone .exe using PyInstaller.

1. Install PyInstaller:

   pip install pyinstaller

2. Run the following build command from the root directory:

   pyinstaller --noconfirm --onedir --windowed --add-data "app/database/schema.sql;app/database/" --add-data "assets;assets/" --name "ScreenshotSearchEngine" app/main.py

3. The built application will be located in the dist/ScreenshotSearchEngine folder.

Note: Even as an executable, the user's system must have Tesseract-OCR installed. To create a completely standalone app, you would need to bundle the Tesseract binaries within the assets/ folder and update app/ocr/engine.py to point to the bundled executable relative path.

Usage

1. Open the App.
2. Go to Settings and click Add Folder. Select your screenshots folder.
3. The app will immediately start indexing in the background. You can see logs in app.log.
4. Go to the Search tab and type keywords.
5. Click a thumbnail to see a large preview and copy the extracted text.

Future Improvements

• Semantic search using local sentence-transformers.
• Image similarity search (finding visually similar screenshots).
• Multi-language OCR configuration from the UI.