They say that a picture is a thousand words, but in our time there are in the folds of digital pictures that we take with our modern cameras, there is much more than that; As researcher Jeron Andrews says according to a BBC report.
On the third of October 2020, the White House published two pictures of then-US President Donald Trump, signing papers and looking at briefings, one day earlier, Trump had announced that he had contracted the new Corona virus, which made the step to publish these two pictures as an attempt To show that he is in good health, his wife Ivanka posted one of the two pictures through her Twitter account, with a comment saying: “Nothing can stop him from working for the American people, unabated!” But those with close eyes noticed something unusual in these two Trump photos.
The two pictures, which were taken in two different rooms at Walter Reed National Military Medical Center, showed Trump in one of them wearing a jacket while in the other he was satisfied with a white shirt, apparently suggesting that he was exercising his presidential duties throughout the day, despite his illness, especially in light of the simultaneous publication of them. With the statements that he is in good health, and that he performs his work normally, but the timing column shown on the sides of the two pictures indicated something different, as it showed that the time difference between each did not exceed 10 minutes.
Although there were, of course, several explanations for this extreme temporal convergence, such as that the time available to the photographer was perhaps only those 10 minutes, and that this coincided with Trump’s desire during the same period to move from room to room, the White House must be He was not happy with those who noticed the timing in the two pictures, as this led the media and commentators to question whether their filming was arranged to convey a political message, and even to question whether Trump had actually acted “relentlessly” during his time in the hospital, or not.
But it is not only the timing block in digital photos that brings unwanted consequences. In this regard, we can mention what happened with John McAfee, the developer of the anti-virus program that bears his name. In 2012, this man was a fugitive from the authorities of Belize, located in Central America. At the time, photographers working for Vice magazine tracked him down and posted a picture of him on the Internet, under the headline “We are with John McAfee now, you idiots.” And without these photographers realizing or intending either, the data about the location that the photo was taken on, which appears on it was revealed. Automatically, it was taken in Guatemala, and so McAfee was quickly found, before he was arrested.
The above are just two examples, which show how the data contained in digital images can reveal information that exceeds what the people who captured it and those who appeared in it want to unveil. Does this happen with your own photos that you post on social media? Could these photos reveal to the world more information than you could previously realize?
When you take a photo, your smartphone or digital camera stores what is known as “metadata” in their photo file. Automatically and intrusively as well, that data reserves a place for itself in every image that it takes, and if we want to define “metadata”, we will say that it is data about the data in itself, as it provides information to identify things related to the image itself, such as when and where it was taken, and the type of camera. That was used in it.
Although it is possible to erase this type of data, using tools that can be used for free, such as the program “Exvetol”, many do not even know its existence, let alone its potential uses, so they do not care to do anything about it, before publishing their pictures on the Internet. Although some social media platforms remove geolocation information from the photos you post, many other platforms and websites do not.
While the lack of awareness of this matter proved useful to investigators working with the police, as it helped them to identify – for example – the whereabouts of criminals, who overlooked what this “metadata” could reveal, this negligence also represented a problem threatening the privacy of citizens. Others who respect the law, if the authorities can track their activities through the photos they publish, on social media platforms, and unfortunately, smart criminals can use the same tricks that the police can resort to to find out the location where a picture was taken, which may lead to Exposing you to the risk of becoming a victim of crimes such as robbery and stalking.
But the details hidden in digital images are not limited to “metadata”. There is also a unique identifying element, linking every picture you take to the camera that was used for it. This element is something you may never even suspect, and even professional photographers themselves may not be aware of it or remember it.
To understand the nature of that element, we must first understand how this or that image is taken. To begin with, it can be said that the main component of all digital cameras, including those embedded in smartphones, is a part known as the “imaging sensor” or “imaging sensor”, which is a complex “electronic network” of semiconductors, consisting of millions of Light-sensitive silicon circuits that are able to absorb photons (light). Due to what is known as the “photoelectric phenomenon,” the absorption of photons causes electrons to be expelled from the outside.
Then, the charge of the electrons emitted from each circuit of this “imaging sensor” is measured digitally, and converted into a digital value as well. This results in the crystallization of numerical values, each of which determines the amount of light that was detected emitted from each circuit, thus forming the image, which we may describe as “drawing with light.”
Due to defects in the manufacturing processes of “these imaging sensors”, the dimensions of their circuits differ slightly from each other, and the sensitivity of some of these circuits to light is greater or less than it should be, regardless of the nature of what is being photographed.
Therefore, even if you use two cameras of the same model and manufacturer, to take two pictures of a surface of equal light intensity on the back, that is, each point on this surface has the same degree of brightness, there will be differences between the two images, depending on the nature of each camera.
The difference in the sensitivity of the “imaging sensor” circuits between each camera and the other, makes each of them – unintentionally – a kind of subtle watermark, which forms what can be described as a “fingerprint” for each “imaging sensor”. This “fingerprint” is printed on every photo you take. And because no imaging sensor is ever identical, there are also no alike fingerprints.
Among those who use digital images as evidence in criminal cases, this “imaging sensor” fingerprint is known as “image response asymmetry”. Jessica Friedrich, a researcher at Binghamton University in New York State, says that it is very difficult to delete this “fingerprint”, even if one tries to do so, it is embedded in the “imaging sensor”. Friedrich notes that this “fingerprint” is also present intentionally, not unintentionally, as is the case with “metadata” of images.
The positive side of having this fingerprint is that it helps researchers like Jessica Friedrich to distinguish false and fabricated images, which is of great importance.
Because images constitute – in principle – a detailed reference to the physical world; They can be used in terms of inferences, conclusions, and proof of certain things, given that they depict what is already there. But in the current climate of misinformation, exacerbated by the large number of available photo-editing programs, it is becoming increasingly important to know the origin and nature of digital images, and whether they are real or fake.
Friedrich obtained a patent for the use of the “asymmetric image response” fingerprint technology. This technology has been approved officially for use as forensic evidence, in cases before US courts. This means that the investigators will be able to determine the areas in which these pictures have been manipulated, and link them to specific cameras, or know the details of the treatment processes that were exposed to them.
This researcher believes that there is a possibility to take advantage of this technique to reveal the truth about the fabricated images, which are fabricated using a method known as “deep falsification”, a vision that is reinforced by the results of the initial research conducted in this regard. The method of “deep faking” is the use of artificial intelligence programs to fabricate images and video clips, and it was notoriously infamous in 2018, after it was used at the time to install pornographic clips. The distinctive feature of “deep falsification” revolves around the fact that its use produces images with realistic features, which makes it a tangible threat to the information environment, given that our inability to distinguish between what is real and what is false, will make there logical reasons for skepticism about all media materials that We are exposed to it.
Thus, in an age in which the objective criteria by which we can distinguish truth from falsehood are disappearing, the ability to spot false images is clearly a positive. But the “asymmetric image response” fingerprint technology has negative uses as well, says Hany Farid, professor of electrical engineering and computer science at the University of California, Berkeley, who is also the founder of the forensic field in which digital images are relied upon.
Although Farid has benefited from this method of linking specific images to specific cameras, in cases of child sexual exploitation, which is a positive thing of course, he urges the need to be careful, to ensure that this technology is not misused, as it happens with any other recognition technology.
This fear is even more important for people working in fields such as photojournalism and human rights, as well as for those who report on irregularities or corruption scandals, as the personal safety of these people may depend on their remaining anonymous. Farid says that there is a possibility to target these people, by linking photos to their cameras, mobile phones, or pictures that were previously published on the Internet.
Our consideration of these privacy issues may lead to another, similar recognition technology, which is being supplied this time with many color printers. These printers add covert yellow tracking points to the papers they print with. These points, invisible to the naked eye, reveal the serial number of the printer, and the time and day on which the paper was printed. And in 2017, the US Federal Bureau of Investigation (FBI) may have used this hidden dot technology to identify who leaked a document belonging to the National Security Agency in the country, containing details about supposed Russian interference, in the 2016 presidential election. in the United States.
Regardless of your personal opinion, regarding the issue of revealing secrets of this kind, there is no doubt that these espionage and surveillance technologies, can threaten us all. The European Commission has expressed its concerns in this regard, saying that these methods may erode the individual’s right to “privacy and private life”. If we consider that the “digital photo fingerprint” is similar to the hidden points that exist on printed papers, then this may lead us to ask whether this fingerprint, in turn, constitutes a violation of each of uss freedom to protect his personal data or not.
In spite of our chronic tendency to reveal ourselves through what we publish on the Internet, at the same time we strictly reserve the right to protect our privacy during our use of the Internet. Although we should be able – in principle – to limit the amount of personal information that we share over the Internet with others, this may seem out of reach, in fact. In light of what we now know about image-tracking technologies, the issue of our ability to control what we reveal to others of information about ourselves may have become an illusion on our part that we are only in control of this matter.
However, it is very difficult to avoid recording the “metadata” that is usually found in digital photos, which means that you have to delete it, after it has been recorded. The only information that you can avoid registering in advance is that of determining the geographical location in which the photo was taken. Deleting the “asymmetric image response” data is much more difficult. Although it is technically possible, Farid says, to reduce the ability to monitor and identify this data, by making the image of lower quality; But how feasible would this be? Of course, this depends on many factors, including the type of device that was used to capture the image, in addition to the algorithm that was used to test the similarity between “photo fingerprints” and each other, which means that there is no single solution that can be used to erase those fingerprints. Whatever it is, between one camera and another.
So, how concerned should we be about these “photo fingerprints”, and what might be caused by their recognition? When I asked the researcher, Jessica Friedrich, about the implications of the various uses of these “fingerprints,” she answered frankly, saying: “A carpenter can do wonders with his hammer, but this hammer may also be a tool for killing.” Of course, we do not suggest here that the data hidden in your digital images could be deadly, but that researcher points out with this analogy, that a technology such as this, may harm you, if you fall into the wrong hands.