A Telecom Paris project about fusion of images with different time of exposition.
https://perso.telecom-paristech.fr/gousseau/IM01/exposure_fusion.pdf
pip install opencv-python
(Pour Lucas : export QT_QPA_PLATFORM=xcb)
pip install scikit-image
pip install matplotlib
All python files are at the root. "Raw" images are situated in the "Images/" folder and grouped together. The "Images/Taken-Lucas/" folder is the folder with different exposures of a photo taken with Lucas' phone.
The "results" images are situated in the "Results/" folder.
Interesting images are located in the "Generated/" folder. They help to understand every step we went through this project. They are generated by the "generation.py" script which executes everything we did in this project step by step.
When an image/variable/... has "correct" in it, it refers to the second fusion method (the one with the pyramids).
Method 1 is implemented in "fusion.py".
Method 2 is implemented in "fusion_correct.py".
The anti-ghosting technique is also implemented in "fusion_correct.py".
To follow everything we have coded, we would advice to start by reading the "generation.py" function, and go to the respective file when a function is called.