Thursday, April 26, 2018
Amit Dubey(Cyber Expert)

MORPHED IMAGES AND VIDEOS

How often do we find pictures going viral these days, insinuating people in different situations, while never even pausing to wonder if the said image or video was real in the first place? We have our cyber expert Amit Dubey on how to detect the same.

Friends, these days, we are surrounded by images and videos and large multimedia content has become an integral part of our daily lives. While there are some good, positive and inspirational images, we also frequently encounter content which creates negativity and hatred.

Last month, a news caught my attention, where a 29-year-old Delhi man allegedly downloaded photographs of girls and morphed them only to blackmail his unsuspecting targets. He told the police that he procured the cut-and-paste application for his mobile phone from Google Play Store using which he could paste the faces/pictures of anyone and superimpose them, on other’s pictures. When raided by police, lots of objectionable morphed pictures of several girls were found in his mobile phone.

This was a shocking case for many of us as the criminal was not an engineer and neither was he well educated. Still, he could manage to do all this with easily available free mobile applications.

Today, when almost everybody is on social networking sites like Facebook or WhatsApp, it’s definitely advisable to take care of your privacy on such platforms or else one can get hold of your profile photograph and misuse it with some bad intent.

Actually every day, such morphed contents is posted on different social media platforms.  Even though, you yourself are not creating these morphed contents, but some times by sharing or liking such images/videos, you unknowingly endorse the objective and become part of such criminal activities. Mostly such morphed contents are created just for fun but yes, some time there are organized criminal intentions too.

As they say, a picture is worth a thousand words. In today’s world, we hardly get time to verify authenticity of a picture, almost 350 million pictures are uploaded every day only on Facebook.

Some time, we see a negative post and create an opinion about that person but we hardly verify if the picture is real or not. Anybody can easily play with our emotions by creating a morphed image targeting an innocent person. Pictures and Videos are very powerful tools now a days. A video can create hatred among people to that extent that it can even lead to riots.

So, here, we should not only protect ourselves by keeping our personal images private and secure but also be cautious not to share any image or video which has been created with an evil intention.

Here are few guidelines which you should follow while creating a DP Image.

  1. Use a low resolution image for your Profile Picture.
  2. Always delete the Profile Picture history.
  3. Click on the icon just beneath the bottom right corner of an album’s cover image that has a small arrow next to it. A privacy options menu will be revealed; this allows you to decide who can see the album.
  4. Choose “Friends” from the privacy options menu to stop anyone who is not on your friend list from seeing the album or downloading the photos in it.
  5. Choose “Custom” from the privacy options to be even more specific about who can see your albums. If you want all of your friends but a few to have access to an album, choose “Friends” from the “Make this visible to” menu and then enter the name of a friend or friend list in the “Hide this from field” — you can enter more than one friend in the field. Click “Save Changes” to complete the process.
  6. If you want to lock your profile pic, click on your profile picture & go to visibility link & click on only me, your profile pic will be locked..

It’s not easy to distinguish the morphed image from an original one with naked eye only, so there you can take help of some online tools which are capable of distinguishing if the images are morphed or not.

Fotoforensics (http://fotoforensics.com/) is a very popular and easy online tool, which you can use to validate the authenticity of an image.

Actually, you need to extract the Image ELA(Error Level Analysis),  JPEG% , Meta Data, etc. to check if an image is  morphed or photo-shopped or not.

If you have a doubt on any image and want to check its authenticity, you can upload the same to this website (http://fotoforensics.com/), after uploading the image, you will get two image outputs, one is original image and another one is analyzed image.

Check the ELA, Iif the image is edited or morphed, then it shows some color in image analysis. If the image is not changed by using any software, then it shows normal white color on the image.

 

Fig 1: Example of morphed or tampered image

Fig 2.0 Untampered Image

You can also analyze the image morphing ratio using the JPEG%. It shows the quality of image when it was last saved. If the quality has decreased, then it must have been modified using software.

Meta data information is also useful to check the authenticity of an image. Through meta data headers, you get to know many details about that photograph such as when was it created, when modified and which camera was used to take that photograph, etc.

You can also check from headers if the image is taken from Facebook, or not. If you upload images which are downloaded from Facebook, then the header will show a message like “Photo has been taken from Facebook” and also give detailed information about that photograph like which software is used to edit this photograph.

For Video Forensics, the easiest way is to use the VLC player and see the data frame by frame and visually analyze the frame. You need to open a video with VLC and press the letter “e” on the keyboard, this will allow you to move the video image –by-image. Every time you press the letter “e” a text “next frame” will appear and if on pressing “e”, image doesn’t change, then this is definitely a duplicate image.

If it’s morphed then you can detect such sudden transition in frames, when you analyze it frame by frame.

December-2016