https://github.com/soumik12345/colorization-using-optimization

Python and C++ implementations of a user-guided image/video colorization technique as proposed by the paper Colorization Using Optimization
https://github.com/soumik12345/colorization-using-optimization

colorization computer-vision cpp14 eigen opencv python

Last synced: 22 days ago
JSON representation

Python and C++ implementations of a user-guided image/video colorization technique as proposed by the paper Colorization Using Optimization

• Host: GitHub
• URL: https://github.com/soumik12345/colorization-using-optimization
• Owner: soumik12345
• License: mit
• Created: 2021-08-15T20:08:23.000Z (over 3 years ago)
• Default Branch: master
• Last Pushed: 2021-08-18T12:11:56.000Z (about 3 years ago)
• Last Synced: 2024-10-04T19:48:18.676Z (about 1 month ago)
• Topics: colorization, computer-vision, cpp14, eigen, opencv, python
• Language: Python
• Homepage:
• Size: 9.87 MB
• Stars: 21
• Watchers: 2
• Forks: 2
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Colorization using Optimization



Python and C++ implementations of a user-guided image/video colorization technique as proposed by the paper 

[Colorization Using Optimization](https://dl.acm.org/doi/10.1145/1015706.1015780). The algorithm is based on a simple premise; neighboring pixels in space-time that have similar intensities should have similar colors. This premise is formalized using a quadratic cost function that obtains an optimization problem that can be solved efficiently using standard techniques. **While using this alogorithm, an artist only needs to annotate the image with a few color scribbles or visual clues, and the indicated colors are automatically propagated in both space and time to produce a fully colorized image or sequence.** The annotation can be done using any drawing tool such as [JSPaint](https://jspaint.app/) or [Gimp](https://www.gimp.org/).

## Instructions

### Instructions for running python version

1. Create a virtualenv using:

    - `virtualenv venv --python=python3`

    - `source venv/bin/activate`

    - `pip install -r requirements.txt`

2. Colorize images using the CLI:

    ```

    python colorize.py

    

    Options:

        --original_image TEXT    Original Image Path

        --visual_clue TEXT       Visual Clue Image Path

        --result_path TEXT       Colorized Image Path (without file extensions)

        -i, --use_itercative     Use Iterative Mode

        --epochs INTEGER         Number of epochs for Iterative Mode

        --log_intervals INTEGER  Log Interval

        --help                   Show this message and exit.

    ```

3. Alternatively, you can run on Google Colab using 

### Instructions to build C++ version

1. Install dependencies using `sh install.sh`

2. Create a build directory `mkdir build && cd build`

3. Generate makefiles and compile using `cmake .. && make`

4. Run the executable using `./colorization [input-image] [visual-clues] [result] [gamma] [threshold]`

5. Alternatively, you can download the executable from [here](https://github.com/soumik12345/colorization-using-optimization/releases/download/0.1/colorization) and run it (installation of dependencies is still needed).

## Results

    

        Original Image

        Visual Clues

        Colorized Image