Skip to content
Code for paper "Near-Infrared Fusion for Photorealistic Image Dehazing"
Python Jupyter Notebook
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.


Code for the paper "Near-Infrared Fusion for Photorealistic Image Dehazing".

Presented at the Electornic Imaging Conference 2018, Burlingame CA, on January 31 2018.


  • Frederke Dümbgen *
  • Majed El Helou *
  • Natalija Gucevska
  • Sabine Süsstrunk

*the first two authors have equal contribution.

Last updated: 18.07.2019


Requires installation of OpenCV 3+. We can recommend this guide for installing OpenCV 3+ for Python on Ubuntu.

Other libraries required (NumPy, Matplotlib) are listed in requirements.txt. Install by running:

pip install -r requirements.txt

How to use

The pipeline for dehazing using photorealistic fusion can be found in pipeline.ipynb. Note that you need to run beforehand, creating the multiresolution decomposition of the image. Since this is a costly operation, it has to be done only once per image, all detail and average images are stored for later use.


The dataset used in this publication can be found here:


Please report any bugs via this repository, and create pull requests if you wish to contribute.


Please cite the paper as

      title = {Near-Infrared Fusion for Photorealistic Image Dehazing},
      author = {Dümbgen, Frederike and El Helou, Majed and Gucevska, Natalija and Süsstrunk, Sabine},
      journal = {IS&T Electronic Imaging: Proceedings},
      pages = {321-1-321-5(5)},
      year = {2018},
      note = {First two authors have equal contribution},
      doi = {10.2352/ISSN.2470-1173.2018.16.COLOR-321}
You can’t perform that action at this time.