utkarsh-deshmukh / single-image-dehazing-python

python implementation of the paper: "Efficient Image Dehazing with Boundary Constraint and Contextual Regularization"

Hi there 👋, I am Utkarsh. A computer vision enthusiast.

On my Github page, I try my best to keep all my projects as an end2end runnable modules. What that means is I want to make these modules as hassle-free and easily runnable as possible.

About me:

🔭 I’m currently working as a Computer Vision Developer at Rivian Automotives.

📫 How to reach me:

Languages and Tools:

single-image-dehazing-python's Issues

How to remove popup window while running code

There is always a popup window when we run code for every single image how to resolve it?

Not working with pip install

Hello brother shared code example is not working with pip install

The following error

File "/home/mrtan/ritvik/image_dehazer.py", line 15, in
HazeCorrectedImg, HazeMap = image_dehazer.remove_haze(HazeImg) # Remove Haze
AttributeError: partially initialized module 'image_dehazer' has no attribute 'remove_haze' (most likely due to a circular import)

Does not work with Python2.7

  Downloading image_dehazer-0.0.4.tar.gz (5.0 kB)

DEPRECATION: Python 2.7 reached the end of its life on January 1st, 2020. Please upgrade your Python as Python 2.7 is no longer maintained. pip 21.0 will drop support for Python 2.7 in January 2021. More details about Python 2 support in pip can be found at https://pip.pypa.io/en/latest/development/release-process/#python-2-support pip 21.0 will remove support for this functionality.
DEPRECATION: The -b/--build/--build-dir/--build-directory option is deprecated and has no effect anymore. pip 21.1 will remove support for this functionality. A possible replacement is use the TMPDIR/TEMP/TMP environment variable, possibly combined with --no-clean. You can find discussion regarding this at https://github.com/pypa/pip/issues/8333.
ERROR: Could not find a version that satisfies the requirement numpy==1.19.0 (from image-dehazer) (from versions: 1.3.0, 1.4.1, 1.5.0, 1.5.1, 1.6.0, 1.6.1, 1.6.2, 1.7.0, 1.7.1, 1.7.2, 1.8.0, 1.8.1, 1.8.2, 1.9.0, 1.9.1, 1.9.2, 1.9.3, 1.10.0.post2, 1.10.1, 1.10.2, 1.10.4, 1.11.0b3, 1.11.0rc1, 1.11.0rc2, 1.11.0, 1.11.1rc1, 1.11.1, 1.11.2rc1, 1.11.2, 1.11.3, 1.12.0b1, 1.12.0rc1, 1.12.0rc2, 1.12.0, 1.12.1rc1, 1.12.1, 1.13.0rc1, 1.13.0rc2, 1.13.0, 1.13.1, 1.13.3, 1.14.0rc1, 1.14.0, 1.14.1, 1.14.2, 1.14.3, 1.14.4, 1.14.5, 1.14.6, 1.15.0rc1, 1.15.0rc2, 1.15.0, 1.15.1, 1.15.2, 1.15.3, 1.15.4, 1.16.0rc1, 1.16.0rc2, 1.16.0, 1.16.1, 1.16.2, 1.16.3, 1.16.4, 1.16.5, 1.16.6)
ERROR: No matching distribution found for numpy==1.19.0 (from image-dehazer)

Real time

Can This algorithm works with a real time video?

difference in calling dehazing using pip package

In the readme it show the implementation of the dehazing as follows

$ import image_dehazer										# Load the library

$ HazeImg = cv2.imread('image_path', 0)						# read input image
$ HazeCorrectedImg = image_dehazer.remove_haze(HazeImg)	

That means we are loading the image as grayscale and simply passing that image in remove_haze function.

But in example.py we are doing the following

Single-Image-Dehazing-Python/src/example.py

Lines 14 to 40 in 393a514

	HazeImg = cv2.imread('../Images/foggy_bench.jpg')
	# Resize image
	'''
	Channels = cv2.split(HazeImg)
	rows, cols = Channels[0].shape
	HazeImg = cv2.resize(HazeImg, (int(0.4 * cols), int(0.4 * rows)))
	'''
	# Estimate Airlight
	windowSze = 15
	AirlightMethod = 'fast'
	A = Airlight(HazeImg, AirlightMethod, windowSze)
	# Calculate Boundary Constraints
	windowSze = 3
	C0 = 20 # Default value = 20 (as recommended in the paper)
	C1 = 300 # Default value = 300 (as recommended in the paper)
	Transmission = BoundCon(HazeImg, A, C0, C1, windowSze) # Computing the Transmission using equation (7) in the paper
	# Refine estimate of transmission
	regularize_lambda = 1 # Default value = 1 (as recommended in the paper) --> Regularization parameter, the more this value, the closer to the original patch wise transmission
	sigma = 0.5
	Transmission = CalTransmission(HazeImg, Transmission, regularize_lambda, sigma) # Using contextual information
	# Perform DeHazing
	HazeCorrectedImg = removeHaze(HazeImg, Transmission, A, 0.85)

That is we are using the color image and calculating Transmission Airlight and then passing those values to removeHaze(HazeImg, Transmission, A, 0.85)

Is the pip documentation wrong/outdated? or image_dehazer.remove_haze(HazeImg) function calculates them internally? If so why we are using grayscale image?

How to reduce run time per image?

This is awesome work, do you know how to reduce runtime for image? Right now it takes about 3-4 seconds for 1280p

Your website does not work

I have tried to use your test site:
https://utkarsh-deshmukh-streamlit-image-dehaze-run-haze-removal-vo1yua.streamlit.app/

but it does not work, Can you fix it please?

Unexpected keyword argument

remove_haze() got an unexpected keyword argument showHazeTrasmissionMap