Computer program detects 'editing' in images

A professor-student duo from the Dartmouth University has devised a way to tell a photoshopped image from an untouched one. Hany Farid, a computer science professor and Eric Kee, a doctoral student, according to the study published in the college's journal, studied 468 sets of photographs that were unedited, as well as those, which were worked upon. A computer program and mathematical description of augmentations, later, they were able to point out the exact differences between the unedited and the edited images. 

Spot the difference (BEFORE and AFTER) (Image credit:

Spot the difference (BEFORE and AFTER) (Image credit: Dartmouth College)



Interestingly, along with the ability to detect if any given image has been photoshopped or not, Farid and Kee's study also allows to gauge the extent of it. Simply put, by a simple method of rating, they can assert the levels of editing done to an image to give it its final look. According to the report, "The authors propose a rating system that takes into account common practices such as cropping and color adjustment while providing assessment of other kinds of modifications that dramatically change a person’s appearance. They consider geometric alterations such as slimming legs, adjusting facial symmetry, and correcting posture, as well as photometric manipulations that might include removing wrinkles, “bags” under the eyes, and skin blemishes."


It not unusual that we read about young girls embracing crash diets, just to achieve a figure similar to what the model in a particular ad has, which for all we know may be photoshopped.  This technique could work as an eye-opener to those who think all that meets the eye is true. 


To read the entire report, follow the link above. 

Published Date: Dec 05, 2011 04:19 pm | Updated Date: Dec 05, 2011 04:19 pm