/ExposureFusion

Exposure Fusion Technique

Primary LanguagePythonMIT LicenseMIT

Exposure Fusion

Exposure Fusion Technique This code implements the Exposure Fusion, a Low-Dynamic-Range technique. It blends multi-exposure sequence of photo into a high-quality image, and is guided by measurement as Contrast, Saturation and Well-exposedness.

You can find the Research paper where the algorithm comes from here [1]

This code used python 2.7 and the packages Numpy and SciPy.

Here some multi-exposure sequence of photo used:

Here the result with the Naive implementation:

Here the result with the Exposure Fusion algorithm with the Laplacian Pyramid:

Usage

To test the code with your own images, put them in a new folder under the folder image_set and edit the text file list_images.txt by putting the names of your images. Then run main.py with the desired arguments. You can see all the possible arguments by running: python main.py -h

Authors

References

[1]: Exposure Fusion: A Simple and Practical Alternative to High Dynamic Range Photography Mertens, T.; Kautz, J.; Van Reeth, F.

[2]: P. Burt and T. Adelson. The Laplacian Pyramid as a Com- pact Image Code. IEEE Transactions on Communication, COM-31:532–540, 1983.