Photo Mosaic Generator 🎨

Transform any image into a stunning photo mosaic by replacing each pixel block with another image! This project uses multiprocessing for fast generation and creates beautiful artistic effects.

What is Photo Mosaic?

Photo mosaic (also called photomosaic) is an artistic technique where a large image is composed of many smaller images. Each small image acts as a "tile" that represents a pixel or block of pixels in the original image, creating a unique visual effect when viewed from a distance.

Features ✨

• High Performance: Utilizes multiprocessing for fast mosaic generation
• Customizable Tile Size: Adjust the size of individual mosaic tiles
• Automatic Image Preprocessing: Smart upscaling and resizing
• Real-time Progress Monitoring: Track processing progress and CPU usage
• Flexible Tile Sources: Use any collection of square images as tiles

Default Example: Cat Mosaic 🐱

By default, this project transforms your photos into cat mosaics - replacing every pixel block with adorable cat photos! The result is a beautiful artwork where your original image is recreated entirely using cat pictures.

Quick Start 🚀

Prerequisites

pip install numpy pillow psutil

Basic Usage

1. Place your input image in the project folder (name it input2.jpg or modify the path in code)

2. Navigate to project directory

   cd photo-mosaic-generator

3. Run the mosaic generator

   # For Korean interface
   python KOR_main.py
   
   # For English interface  
   python EN_main.py

4. Check your results! The output will be saved as mosaic_multiprocessing1.jpg

Adjusting Cat Size

If the cats appear too small or too large in your mosaic, modify the min_size parameter:

generate_mosaic_real_multiprocessing(
    # ... other parameters ...
    min_size=5000  # Increase for larger cats, decrease for smaller cats
)

Using Custom Images 🖼️

Want to create mosaics with your own images instead of cats? Follow these steps:

Step 1: Prepare Your Tile Images

1. Create a new folder for your tile images
2. Add your images to this folder (square images work best!)
3. Use diverse images for better mosaic quality
4. Ensure consistent naming (e.g., dog_001.png, dog_002.png, etc.)

Step 2: Generate Image Database

1. Modify the preprocessing script parameters:

   compute_dataset_mean_colors(
       data_dir='YourImageFolder',    # Your new folder name
       output_file='custom_avg_rgb.json',
       prefix='dog_'                  # Your image prefix
   )

2. Run preprocessing:

   python imgtojson.py

Step 3: Update Main Script

Modify the main script parameters:

generate_mosaic_real_multiprocessing(
    img_path='input.jpg',
    avg_rgb_path='custom_avg_rgb.json',      # Your JSON file
    tile_dir_path='YourImageFolder',         # Your image folder
    output_path='output.jpg',
    base_size=64,                            # Match your tile image size
    chunk_size=512,
    min_size=5000
)

Step 4: Generate Your Custom Mosaic

python EN_main.py  # or KOR_main.py

Parameters Explained 🔧

Parameter	Description	Default
img_path	Input image file path	'input.jpg'
avg_rgb_path	JSON file with tile RGB data	'image_avg_rgb.json'
tile_dir_path	Directory containing tile images	'Data'
output_path	Output mosaic file path	'output.jpg'
base_size	Size of each mosaic tile in pixels	64
chunk_size	Chunk size for multiprocessing	512
min_size	Minimum image size (auto-upscaled if smaller)	3000

File Structure 📁

photo-mosaic-generator/
├── EN_main.py              # English version main script
├── KOR_main.py             # Korean version main script  
├── imgtojson.py            # Image preprocessing script
├── Data/                   # Default cat images folder
│   ├── cat_0.png
│   ├── cat_1.png
│   └── ...
├── image_avg_rgb.json      # Preprocessed RGB data
├── input.jpg              # Your input image
└── README.md

Performance Tips 💡

• Image Quality: Use high-quality, diverse tile images for better results
• Processing Speed: The script automatically uses all available CPU cores
• Memory Usage: Large images may require significant RAM
• Tile Diversity: More diverse tiles create more detailed mosaics

Troubleshooting 🔧

Issue: Mosaic looks blurry or low quality

• Solution: Increase min_size parameter or use higher resolution input image

Issue: Process runs out of memory

• Solution: Reduce chunk_size or use smaller input image

Issue: Colors don't match well

• Solution: Use more diverse tile images or ensure good color coverage

License 📄

This project is open source. Feel free to modify and distribute!

Contributing 🤝

Contributions are welcome! Please feel free to submit pull requests or open issues.

Credits 🙏

• Cat Dataset: Cat Dataset from Kaggle - https://www.kaggle.com/datasets/borhanitrash/cat-dataset

---

Created with ❤️ for photo mosaic enthusiasts!