Why are people camera shy

Photograph 100,000 people, within a few hours, and then compare these photos with hundreds of thousands of other photos. Impossible? A person who moves through a large building, from room to room, from floor to floor, can be observed by cameras, without any gaps, without a person controlling the video surveillance? Science fiction? Camera systems that learn independently, recognize unusual situations and then automatically trigger an alarm? Utopian dream of all security authorities? No, reality.

More than ten years ago, more than 100,000 spectators were scrutinized by the police at the American Super Bowl in Tampa (Florida). When they streamed into the stadium in 2001, people were filmed - without their knowledge - by permanently installed video cameras at all entrances. While the game was still going on, the security authorities compared these photos with images from local and state criminal cards. 19 petty criminals were identified and arrested as they left the stadium.

Would something like that also be possible in Germany? And how fast is the technical development going? Answers to important questions.

Do cameras offer more security?

Deterrence and prevention, education, strengthening the feeling of security. These are the three arguments with which the constant expansion of video surveillance is usually justified. Arguments which, however, are not always based on facts and which are all but undisputed.

  • Enlightenment: Search successes from the past show that cameras help in the search for criminals. How many would have been arrested even without cameras cannot be verified, however.
  • Sense of security: Video surveillance only increases the public's sense of security for a short time - if at all. Studies show that. "This effect fizzles out quickly," says the Hamburg social researcher Nils Zurawski, who has been dealing with the topic for years. From his research he concludes: "Security at Marienplatz, for example, does not come from cameras." A camera could even reinforce existing uncertainty. "People say: 'There, a camera! I always knew that something was wrong here.'"
  • Prevention: "The deterrent effects of video surveillance are at least doubtful, but in any case strongly dependent on the context," says Tobias Matzner, referring to studies on the subject. He is responsible for the "Ethics of intelligent video surveillance" project at the University of Tübingen. In fact, the study results are inconclusive. Studies from Great Britain in particular show a moderate but significant decrease in the crime rate. However, mainly in parking garages. Comparable studies from the USA could not find a comparable effect. Matzner draws the conclusion from this: Video surveillance cannot be a universal solution. In addition, one must assume that crime is not prevented by cameras, but only suppressed. However, it is precisely this phenomenon that is difficult to scientifically investigate due to the available data.

Can cameras recognize us and identify us as a person?

Since the attempts by Facebook to introduce automatic face recognition, which failed at least in Europe, this type of pattern recognition has become a matter of public awareness. Software that compares images with one another exists. So it is quite conceivable that security authorities couple their photo files with a video camera - and thus automatically recognize whether one of the stored persons is moving through the field of view of the cameras. However, face recognition processes are still very error-prone, especially when used in conjunction with video cameras. There are often incorrect assignments. The pilot test in America, which started at the Super Bowl 2001, was discontinued due to a lack of accuracy. "Identifying the individual is not yet possible efficiently for a large number of people at the moment," says futurologist Britta Oertel.

The Federal Criminal Police Office has also had bad experiences. In 2007, it installed a camera system at Mainz main station that was supposed to automatically recognize faces. But in many cases it was simply not bright enough to identify individuals. The success rate in good lighting conditions was at least 60 percent. "There are systems that can recognize stationary objects," says Nils Zurawski, who as a social researcher deals with information and communication technologies. But that doesn't work well enough for faces, says Zurawski. Nevertheless, scientists in Tokyo are currently developing glasses that can block facial recognition software. Better safe than sorry.