A Python tool that automatically aligns and centers specific words in images using OCR.

Perfect for creating that "busy" effect in documentary-style videos.

• 🔍 OCR-powered word detection - Automatically finds target words in images
• 🎯 Automatic centering - Centers the detected word in the output frame
• 📐 Consistent sizing - Maintains uniform word size across all images
• ⚡ Parallel processing - Processes multiple images simultaneously for speed
• 🔤 Partial matching - Optionally match partial words (e.g., "warp" matches "warpdotdev")
• 📁 Batch processing - Process entire folders of images at once

1. Python 3.8 or higher
2. Tesseract OCR installed on your system
   • Windows: Download installer from GitHub releases
   • Mac: brew install tesseract
   • Linux: sudo apt-get install tesseract-ocr

# Clone the repository
git clone https://github.com/yourusername/quick-cuts.git
cd quick-cuts

# Install dependencies with uv
uv pip install -r requirements.txt

# Clone the repository
git clone https://github.com/yourusername/quick-cuts.git
cd quick-cuts

# Install dependencies
pip install -r requirements.txt

Process images with a specific word:

python quick_cuts.py images/*.png -w "breaking"

Process a folder with partial matching:

python quick_cuts.py images/ -w "warp" --partial

Custom output size and word height:

python quick_cuts.py images/*.jpg -w "news" -s 1920x1080 --word-height 150

Specify output directory:

python quick_cuts.py images/ -w "alert" -o output_folder

Use multiple workers for faster processing:

python quick_cuts.py images/ -w "update" --workers 8

1. OCR Detection: The tool uses Tesseract OCR to detect text in each image
2. Word Location: Finds the target word and its bounding box
3. Alignment Calculation: Calculates the transformation needed to center the word
4. Image Generation: Creates a new image with the word perfectly centered
5. Batch Processing: Processes all images in parallel for efficiency

The tool creates aligned images with consistent properties:

• All images have the same resolution
• Target word is centered in every frame
• Word size is consistent across all images
• Original aspect ratios are preserved where possible

Output files are saved as aligned_[original_filename] in the output directory.

1. Image Quality: Higher resolution images produce better OCR results
2. Text Clarity: Ensure text is clear and readable in source images
3. Contrast: Good contrast between text and background improves detection
4. Preprocessing: The tool automatically applies image preprocessing, but very poor quality images may still fail

• Ensure Tesseract is properly installed and in your PATH
• Try adjusting the confidence threshold in the code
• Check that the text is clear and readable in your images

• The tool handles Unicode characters in filenames automatically
• If issues persist, try renaming files to use only ASCII characters

• Reduce the number of workers if you experience memory issues
• Process images in smaller batches for very large datasets

This tool is perfect for creating dynamic news-style cuts where specific words need to be emphasized:

1. Export frames from your video clips
2. Run quick_cuts to align all frames on your target word
3. Import the aligned images into Premiere/After Effects
4. Create smooth transitions between different clips with the word always centered

MIT License - feel free to use this tool for any purpose!

Contributions are welcome! Please feel free to submit a Pull Request.