Picture this - “You’ve had a tough day at work (or school). You return home only to find the door locked and no one home. You use your keys to get in, head straight to your computer and log in to your email. You’ve got new mail; it’s a forwarded message from your closest friend. He chose to share with you a picture from the trip you took earlier in the week, only that all your friends have had their faces replaced by the faces of your favourite celebrities. You smile in amusement.”
With the increasing popularity of digital cameras and steady improvement of photo editing software, photo editing and photo fakery has emerged as a popular pastime over the years. Fake photos have become so common on the internet of late that one encounters at least one each day, somewhere or the other on the internet. We may have all seen images of giant grasshoppers climbing the empire state building, but we naturally assume that these images are fake and assume that they were created simply for our amusement. (Fiete, 2005).
It’s only when fake photos are created with motives other than entertainment that our discretion is called upon to decide whether we can believe the image or not. Hence being cognizant of how fake images are made will help us from being deceived by them.
What is a fake image?
A fake image can be defined as an image of an object or scene that wasn't captured as the image would imply.
How are fake images created?
Fake images can be created using -
- Dark room and photo developing techniques
- Using a computer and a image processing software The discussion in this article is limited to image processing software alone.
Anatomy of a digital image
Take a look at this image –
A digital image is made up of many dots (called picture elements or pixels) each of varying hue/brightness. This image is stored in memory as a grid of numbers each representing the brightness of the respective pixel (the numbers vary from 0 for black to 255 for white). An 8 bit grayscale image hence has 2^8 = 256 gray values (0 to 255). A 24 bit colour image is formed by combining together different gray values from a green image, a blue image and a red image There are hence 224 (16777216) colour options for every pixel.
While it is possible for an artist to “paint” out an image from scratch by assigning the gray values to each pixel in an 8 bit image, the number of colour options in 24 bit images restricts this possibility.
Hence fake images are now better created using computer graphics software designed to generate 3D objects with realistic illumination.
The most common way of creating fake images is by altering an existing image caught on camera.
How do I decide if an image is fake or real?
Take a look at these 2 images:
 |
 |
One of them is a real photograph but the other is a fake and has been created by “The Credibles” artwork team. Let us take you through a set of steps and tell you how to spot fake images from real ones.
Altering context or altering content
A fake image is a photograph with either its context or contents not the same as the original. A context altered image is one which is a photograph in itself but claims to be what it is not. A content altered image is a suitably edited photograph that wasn’t actually captured.
Examples :
The Surgeon's Photo, 1934, reportedly showing the Loch Ness Monster, is a context altered image. This photo was revealed to be merely a toy submarine with a model of a serpent head attached.
Well, in our case the fake images of the hand and counters are made by editing the pixels of a photograph using software and are hence they are content altered images.
Adding elements to an image
 |
Talented image editors can merge two images by adding elements of one into the other. When elements are added, fakery can be spotted by looking out for unexplained shadows, which defy the laws of physics.
Examples: The photograph on your left is that of Vietnam veteran John Kerry addressing an audience at an anti-war rally. A picture of American actress Jane Fonda addressing a crowd during the civil rights movement has been used to create the deceptive image. Though the photo has been faked skillfully, the shadows near Fonda’s arm are questionable. |
Painting out elements from an image
Painting out elements and people is a very common trick and has been popular from the 20th century, for propaganda in political regimes. When people fell out of favour with the ruling governments it was popular to paint them out of historic documents. 1920 : Vladimir Lenin with Leon Trotsky at Bolshoi Theater Moscow
Seven years later, power struggles forced the revolutionary to quit the Soviet Communist Party–and so a photo retoucher meticulously painted him out of the picture. 1927 : Vladimir Lenin at Bolshoi Theater. But where is Trotsky?
Sometimes when elements are painted out of an image, noise levels may be inconsistent or pixel contrast might be minimal. When enlarged we can see that image no. 3 is real while image 1 and 2 are fake. :
;) FAKE |
;) FAKE |
;) REAL |
Click on the images for larger versions.
Other things to consider while spotting a fake image
- Use Common Sense: If the image looks credible, and it depicts a subject that can be photographed, made easily, such an image would, in all probability, be real as there is no reason to fake it. If the image looks unbelievable, it could be fake.
- Lay Focus on subtle details : Focus on the subtle details in composure and texture. Most computer generated scenes/images generally have a cartoon feel to them and don’t fool anyone unless created very skillfully.
- Visually inspect to find physical inconsistencies: The physical traits of the image that can be assessed include the illumination conditions, edge sharpness, resolution, tone, relative scale, and noise characteristics.
References
- Fiete, Robert D. Photo Fakery. OE Magazine
the SPIE magazine of Phototonics Technologies and Applications - Jan '05
Retrieved : 26/03/2007 - “The Loch Ness Saga,” by Dr. Maurice Burton, New Scientist, June 24, 1982, p. 872; July 1, 1982, pp. 41-42; July 8, 1982, pp. 112-113.
- US News
Retrieved : 26/03/2007
Click to translate
External Links
Print Page