Skip to content

Synthesize Hazy/Foggy Image using Monodepth and Atmospheric Scattering Model

License

Notifications You must be signed in to change notification settings

Chi-SquareX/haze-synthesize

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

40 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Synthesize Hazy/Foggy Image using Monodepth and Atmospheric Scattering Model

This project is to synthesize hazy/foggy image using Monodepth and atmospheric scattering model.

Medium article: Synthesize Hazy/Foggy Image using Monodepth and Atmospheric Scattering Model

Dependencies

  • OpenCV
  • Matplotlib
  • Numpy

How to Use

1. Generate Depth Map (using Monodepth2)

  • An easy-to-use implementation of Monodepth2 can be reached via this link

2. Locate Image Files (optional)

  • Put the original (clean) image in the folder "docs"
  • Put the generated depth map in the folder "docs"
  • Say "fname" is the file name of the clean image, the file name of the generated depth map should be "fname_depth"

3. Generate Hazy Image

  • Run notebook: synthesize-haze.ipynb

Results

  • 1st row: original images
  • 2nd row: synthesized hazy images with sparse haze
  • 3rd row: synthesized hazy images with dense haze

Citation

@article{tran2022novel,
  title={A Novel Encoder-Decoder Network with Guided Transmission Map for Single Image Dehazing},
  author={Tran, Le-Anh and Moon, Seokyong and Park, Dong-Chul},
  journal={arXiv preprint arXiv:2202.04757},
  year={2022}
}
@inproceedings{tran2022novel,
  title={A Novel Encoder-Decoder Network with Guided Transmission Map for Single Image Dehazing},
  author={Tran, Le-Anh and Moon, Seokyong and Park, Dong-Chul},
  booktitle={International Conference on Industry Science and Computer Sciences Innovation (iSCSi’22)},
  year={2022}
}

Have fun!

LA Tran

Oct. 2021

About

Synthesize Hazy/Foggy Image using Monodepth and Atmospheric Scattering Model

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 100.0%