The history of video surveillance: the path from the TV and the Third Reich to the clouds and neural networks
An iconoscope in the exposition of the Czech National Museum of Technology.
Watching something (or someone) is boring, sometimes dangerous, but often necessary. People constantly come up with and improve technical means that can make life easier and reduce the risk for the observer, as well as improve the quality of observation. The history of remote monitoring systems is counted from the inception of electronic television, or, more precisely, from the invention of the iconoscope - an electronic tube transmitting an image.
However, observation as a phenomenon arose much earlier than all technical means. Primitive hunters peering out from under the vaults of the cave already knew the importance of constant monitoring. In the 19th century, police visited prisons, where they examined prisoners, remembering their faces and appearance. With the advent of photography, and then television, it became much easier to transmit important information, but the issue of the safety of data storage has not yet been resolved. Video surveillance should not only record important events, but also, if possible, save them. Forever.
The path to the iconoscope
Attempts to create equipment for receiving and transmitting a television signal have been made around the world by a large number of researchers. As early as 1907, an application was filed for the first patent for electronic television technology, “Method for the electrical transmission of images.” And in 1911, the Russian physicist Boris Rozing managed in his laboratory to receive images of the simplest figures designed by his kinescope. But a real breakthrough in the clarity of electronic television images (and the last nail in the lid of the coffin of mechanical television ) was the appearance of a transmitting television tube (iconoscope), invented in 1931 by our former compatriot, Russian-American engineer Vladimir Zvorykin.
In 1919, Zvorykin emigrated to the United States, where he began to engage in technology for transmitting images at a distance. By 1923, while working for Westinghouse Electric, he managed to create a television device based on an original transmission tube with a mosaic photocathode. However, the invention did not make any impression on the owners of the company, and Zvorykin soon had to go on a “free voyage”. In just a few years, Zvorykin single-handedly created many important devices for the future of television (photocells, a sound recording system, and several others).
In 1928, Zvorykin met another emigrant from Russia, David Sarnov, vice president of Radio Corporation of America (RCA). Heading the RCA electronics laboratory, the inventor quickly completed the development of his own high-vacuum television receiving tube - a kinescope. Following this, having improved the technology of accumulating the charge of point photocells, Zvorykin created an iconoscope. In 1932, with the help of an iconoscope from a 2.5 kW transmitter installed on the Empire State Building, the first experimental electronic television transmissions began.
Shot from the film by Leonid Parfenov “Zvorykin-Muromets”
The iconoscope (from Greek: εἰκών “image” and σκοπεῖν “watch, see”) consists of a vacuum glass bulb, inside which a photosensitive target is mounted, on which the image is projected by the lens, and an electron beam gun located to the side or bottom of the lens. The image in the iconoscope falls onto a plate with a mosaic of photocells isolated from each other. In those days, this mosaic was created from mica with a photosensitive layer of cesium. But Zvorykin improved the method: a thin silver film was burned in mica so that it coagulated into many tiny drops. The iconoscope plate (6 by 10 cm) uses 1,200,000 such drops. Each drop is a photocell; when the target is illuminated by the photoelectric effect, silver droplets acquire a positive charge proportional to the illumination.
Vladimir Kozmich was not only engaged in exclusively peaceful television broadcasting systems, but also wanted to build an “air torpedo with an electronic eye,” or, in modern terminology, a guided missile. To demonstrate the concept, in 1937 he put a large iconoscope on a plane and let it fly around the Statue of Liberty. The military on the ground were able to fully see the attraction on the TV screen, which was pretty impressed. RCA got an order for a remote-controlled weapon: planning an air bomb and an unmanned kamikaze plane. The project was ultimately considered a failure - the television signal was easily jammed by the enemy, but along with the air force and the US Navy received several television intelligence systems.
From V-2 to VHS
The Olympia-Kanone electronic television camera during a live broadcast from the Berlin stadium at the Summer Olympics in 1936.
As a result of the hard work of scientists from different countries, television systems developed rapidly, and real video surveillance appeared in the modern sense. In 1941, German electrical engineer and television pioneer Walter Bruch installed a CCTV system (Closed Circuit Television - a closed loop television system) at the Peenemuende training ground, where he tested the famous "retaliation weapon" - the V-2 missiles.
Launch of the V-2 in the summer of 1943
At the research center of the Third Reich, missiles at the start often exploded, and two installed cameras made it possible to monitor the launch from a safe distance of 2.5 km. Video surveillance data about problems could greatly help rocket scientists, but there was not enough opportunity to record observations. They knew how to transmit images, but not to record. The operator had to sit in front of the monitor in order not to miss something.
The first Ampex VR 1000B VCR created under the guidance of engineer and innovator Alexander Ponyatov
In 1951, the first VTR (VideoTape Recorder) devices appeared that recorded images on magnetic tape. They were about the size of a desk, and were "like a cast-iron bridge." The first VCR, created in 1956, was able to fully record sound and images onto magnetic tape using magnetic video heads, but it cost $ 50,000. For comparison: a Chevy Corvette car cost about $ 3,000. However, this did not prevent the rapid growth of the device’s popularity - after half a year the device began to be used in all leading US television studios.
The visit of the Thai royal family in London in July 1960
Since the end of the 50s, cameras began to be installed on roads, in crowded places, at important facilities. In 1960, the London police installed two cameras on Trafalgar Square - specifically to watch the crowd gather to watch the official visit of the Thai royal family. After the visit, the cameras were removed - this kind of equipment in those days cost millions of dollars.
Monitoring system at the Munich Police Headquarters, 1973
Towards the end of the decade, remotely controlled pivoting mechanisms for cameras were invented, which made it possible to place one camera where several were previously needed. Nevertheless, the video surveillance posts were equipped with dozens of monitors - each camera needed its own. Operators had to constantly run their eyes across the entire array of monitors, their attention was scattered.
The first device providing the possibility of telephone video communication was presented on April 20, 1964.
In 1969, a patent was issued for a home security system (video intercom) that allows you to see who is behind the door on the TV screen and remotely unlock the lock.
When video surveillance took over the world
A new era of video surveillance began with the invention in the early 70s of domestic video recorders. Video recording became available to private individuals, and cameras began to appear everywhere: in homes, shops, banks, educational institutions. Now the witness did not have to affirm in court that he would recognize the thief in the defendant - a VHS cassette with the incident was attached to the case.
With the advent of multiplexers that allow you to display images from multiple cameras on one monitor and record it on one tape, video surveillance has become truly convenient. But there is no limit to perfection: in the 80s, cameras based on a CCD array (CCD, charge-coupled device), a general designation of a class of semiconductor devices that use the technology of controlled charge transfer to semiconductor volume). The resolution of the matrices of the first CCD cameras left much to be desired, but they were smaller and significantly more sensitive than old cameras.
Another qualitative leap in video surveillance occurred in the 90s, when fully digital systems appeared. DVRs equipped with hard drives learned to record the image in a circle when the “tail” of the recording erases the beginning, and also enable recording when motion is detected. Video cameras appear in ATMs, the first baby monitor was launched on the market, and personal computers received a completely new peripheral device - a web camera. The cameras are equipped with sensors and record motion or sound detection.
The new millennium is a new round of the development of video surveillance, which has become networked. IP cameras transmit the image through the network, both local and via the Internet, and the DVR can be located anywhere, or you can do without it at all by adapting the computer to store the recordings. In parallel with this, in the 2000s, video analytics systems that were able to recognize objects and events in the frame were recognized, due to which the observation and analysis of video was simplified and accelerated many times.
Present and future
An IP camera that stores an archive in the cloud The
boom of "clouds" of the early 2010s has borne fruit; cloud technologies have come to many industries, including video surveillance. A camera that supports cloud technology is very different from a classic IP camera: ease of deployment and configuration with absolutely no troubles, a minimum of maintenance, the availability of recordings anytime anywhere, integration with almost any web services. Where a ramshackle IP system with video recorders and video analytics servers used to be fenced, now it’s enough to hang modern cloud cameras, putting aside all the headache for the share of the service provider.
But the main innovation brought by the cloud is not that. The way records are stored has changed dramatically. Local DVRs are vulnerable and potentially unreliable - their hard drives fail or simply become full, and in order to steal records, an attacker just needs to have physical access to the device. In the cloud storage, data is encrypted, backed up, and guaranteed to be available (unless, of course, your Internet channel works) - while they are stored theoretically forever. A cloud provider is responsible for the quality of the service with its own money, which means that you can expect that no accidents, such as the administrator’s activities, overheating or filling up the hard drive, or even unauthorized persons entering the premises, will lead to the loss or compromise of records.
Nvidia Engineerweaned cats to spoil the lawn using a video surveillance system, machine vision and deep training
It may seem that video surveillance systems have reached the "ceiling" of development. However, if we continue the chronological series of innovations in this industry, it becomes clear that by 2020 we should expect a new breakthrough that will again radically change the ideology of remote observation. Most likely, a new round of development will be associated with the improvement of systems for recognizing the behavior of objects. Already, neural networks can quickly analyze certain objects during translation. In a few years, they will be able to detect many events - for example, illegal actions - and automatically adjust the entire system to changing conditions in real time. Surveillance cameras with neural networks and clouds will become truly smart: they will not only record events, but also decide what to do next: