noahzn / fohis

Towards Simulating Foggy and Hazy Images and Evaluating their Authenticity

Hi there 👋

🔭 I’m a Research Engineer working on UAV-based computer vision.

fohis's Issues

I can't train new pictures

Hi, thank you for your contribution. Now I have a question to ask: I want to add simulation to the new image, but I found that there is an IMgd.jpg in the sample image, how to generate the IMgd.jpg? Hope to answer my doubts, thank you very much!

compute_authenticity.py issues

I used the code you provided to calculate the image authenticity index for my graph, but the result is sometimes one and sometimes not in the range 1-5. I want to ask which hyperparameter I need to adjust to calculate the correct data. Thank you for your work

global name 'pnoise3' is not defined in toolkit.py

About the data

Hello,thank you for your work.
And i want to ask you that how can i get the depth images? Is that can generate from rgb images?

How to obtain the depth images

As the code requires the RGB Image and the Depth image to simulate fog effects. how to obtain Depth images

Image Depth

Hey I'm having issue with the image depth as I'm using an RGB image and generated a depth with code I found and ran into this error:

Traceback (most recent call last): File "fog.py", line 28, in <module> const.CAMERA_ALTITUDE) File "C:\Users\hashi\FYELabs\DND\FoHIS\FoHIS\tool_kit.py", line 30, in elevation_and_distance_estimation File "C:\Users\hashi\FYELabs\DND\FoHIS\FoHIS\tool_kit.py", line 75, in get_image_info KeyError: 'dpi'

This is the image I am using:

If you can provide your method for depth that would be great

Do we need to get a depth image first?