The project showcased a python implementation of Bilateral Guided Upsampling introduced by J. Chen et al. (2016), and extended its application to Tone Mapping and Gradient Enhancement Operators. More explainations can be found from the report.pdf

Running the code

The full pipeline is demonstrated in main.ipynb. A short demo is in demo.ipynb which uses the main.py as a module.

Experiment Results

High-resolution output from the experiments are documented in ./report

Main Dependencies

Basics

opencv
scikit-image
scipy
matplotlib

For solving least square problem

sudo apt-get install python-scipy libsuitesparse-dev
conda install -c conda-forge scikit-sparse

For Tone Mapping operators and TMQI

#pip install git+https://github.com/dvolgyes/TMQI
import imageio
imageio.plugins.freeimage.download()
from TMQI import TMQI

If some libraries are still missing, feel free to check out the environment/environment_graphics.yml