Automatic Photo Album
Project Book Presentation Poster Code
Since the digital
camera became affordable and widely used, the amount and variety
of photos we take is
greater in comparison to the times we used the film camera.
Therefore, we find
it difficult to decide which photos out of hundreds we took are
the best to represent
the whole variety and are enough aesthetic photos.
Automatic photo
album is an algorithm that receives multiple photos and selects
a number of photos
out of them as specified by the user.
The need for this
algorithm arose due to the transition to digital photography,
which practically
removed the limitation on the amount of photos one can store.
The goal is to
select the photos as if the user would have selected them himself.
We achieve that by
using several basic photography rules, combined with
the photos'
Co-saliency values. Co-saliency is an algorithm we developed to
find the photo that
best represents all the other photos in a certain group of photos.