Mosaic Image Project 🎨🖼️🧩

Project Overview

This project generates a mosaic image by recreating a target image using smaller images from a database. The smaller images (tiles) are selected based on their color similarity to the corresponding regions in the target image, creating a visually mosaic effect.

Features

• Create mosaic images from any target image.
• Use a custom image database to generate the mosaic.
• Adjust tile size and other parameters to fine-tune the mosaic creation process.

Installation

1. Clone the repository:

   git clone https://github.com/White-On/Mosaic_Image
   cd Mosaic_Image

2. Install the required dependencies:

   • Make sure you have Python 3.x installed.
   • Install necessary Python libraries by running:

     pip install -r requirements.txt

Usage

1. Prepare your image database:

   • Collect the small images (tiles) that you want to use in your mosaic.
   • Store them in a folder (e.g., image_database/).
   • The images will be resized to the tile size specified in the mosaic generator. For best results, use images of similar aspect ratios or crop them to the desired size before creating the mosaic.

2. Run the mosaic generator:

   • Use the following command to create the mosaic:

     python mosaic_image.py --target-image path/to/target_image.jpg --tiles-path path/to/image_database/ --output path/to/output_mosaic.jpg 

3. Parameters:

   • --target-image: Path to the image you want to recreate as a mosaic.
   • --tiles-path: Path to the folder containing your small images.
   • --output: Path where the generated mosaic image will be saved.
   • --tile-size: Size of the tiles in the mosaic (default is 50 pixels).
   • --partition-size: The target image is partitioned into tiles of this size (default is 50 pixels).
   • --similarity-method: Method used to compare color similarity between tiles, you can choose between average, histogram, and color (default is color).
   • --histo-method: Method used to compare histogram (if similarity method is histogram), you can choose between correlation, chi-square, intersection, and bhattacharyya (default is intersection). See the OpenCV documentation for more information on histogram comparison methods: OpenCV Doc
   • --max-random-range: Maximum number of random tiles to be selected for each partition (default is 5).
   • --select-frequency: Frequency of selecting the most similar tiles. You can choose between best, uniform or normal (best is default and normal is usefull for images with very uniform area and when the image database is big enouth)
   • --unsafe: Disable safety checks ( won't allow to run for some parameter combo).

Example

python mosaic_image.py --target-image images/sunset.jpg --tiles-path images/tiles/ --output output/mosaic_sunset.jpg --tile-size 50 50 --partition-size 25 25 --similarity-method color --max-random-range 10 --select-frequency normal

This will create a mosaic of the sunset.jpg image using tiles from the images/tiles/ directory and save the mosaic to output/mosaic_sunset.jpg.

Contributing

Feel free to contribute to this project! You can fork the repository and create a pull request for any improvements, new features, or bug fixes.