Last evening, Associate Professor of Information Sciences and Technology James Wang was interviewed on BBC Radio regarding his research project Acquine. Acquine is an online tool that rates photographs based on mathematical cues and provides the user with a rating out of 100. Professor Wang spoke about how his program goes about rating a photograph:

“We know there are certain rules, like say the rule of thirds or certain pictures with low depths of field give people a better impression. Or pictures with higher saturation or contrast give people a better impression. You know that all of this has been encoded into our features.”

I gave the program a test using a couple of well known photographs from recognized masters of photography to see how the system’s ratings compared to general opinion.

The first photo I used is by landscape master Ansel Adams, who is generally regarded as one of the greatest photographers of all time. Acquine seemed to agree with the consensus, giving the picture below a rating of 89.1 out of 100.

I thought perhaps that was a little low, so I ran the program again using one of the most famous photographs of all time. National Geographic rated this as the number one photograph ever to grace the cover of their magazine (cool tidbit: It was taken by Penn State alumni, Steve McCurry). “Afghan Girl” is almost instantly recognizable as a photographic masterpiece, so I ran it through Acquine to see what it thought. Acquine rated “Afghan Girl” at 87.4 out of 100.

So clearly Acquine is on the right track when it comes to judging photographs, though as is to be expected, the computer program can’t and won’t always agree with general consensus about what makes a photograph appealing to the human eye.

If you want to see how your own photo’s stand up, you can go to http://acquine.alipr.com/ to give it a shot.

[source]
[photo]
[photo]