Network video: development vectors.

Network video: development vectors.

Notes from the AXIS Communications press conference dedicated to new directions in the development of IP video surveillance

In the modern world, nothing stands still; it seems that the rhythm of life itself makes us constantly go somewhere, develop, improve. As before, one generation is replaced by another, new technologies replace the old ones, offering us more and more advanced and thoughtful solutions. At the moment, we are witnessing, and to some extent, participants in the advent of the era of IP video surveillance.

A little history. It all began in the 50s of the XX century, when the first analog cameras appeared. In 1980, they began to conduct analog video recording, which was replaced by digital in the 90s. A historical moment can be considered 1996, when the Swedish company AXIS Communications released the world's first network video surveillance camera.

Experts identify three main areas of IP video surveillance development. One of them was and remains improving the quality of the transmitted image. This is achieved by using megapixel matrices in cameras, which allow for higher image resolution and, as a result, a larger viewing area and higher detail. However, using megapixel technologies also has its drawbacks. These are primarily low light sensitivity, low frame rate (FPS) and the higher cost of such cameras.

Improving image quality inevitably entails an increase in the volume of video streams, and, therefore, requires more memory for storing information and greater network bandwidth. Today, the M-JPEG and MPEG-4 standards are widely used, the capabilities of which are already insufficient for processing large volumes of data. Therefore, another direction in the development of IP video surveillance is becoming obvious — the need to develop a new, more advanced compression format. Several companies, including AXIS Communications, are actively working on creating a new compression standard that would allow transmitting good-quality images with the lowest traffic costs.

But the main emphasis in the development of network video surveillance is on the use of intelligent technologies. The expression «intelligent video» implies, first of all, a certain logical analysis of the information received from the camera and a prompt response in accordance with the specified parameters. Of course, such functions are already performed by various software applications, but the uniqueness of the new direction in the development of IP video surveillance is that the intelligent video function is built directly into the camera itself.

The first new generation network cameras were presented by AXIS Communications. Using software algorithms, they are able to track the contours and trajectory of an object, and sometimes even identify cases of petty theft. The motion detection function can be combined with the remote control function, which will allow recording only if there is suspicious activity. At the same time, cameras with a built-in logical analysis function allow not only to record the fact of theft or other illegal activity, but also help to prevent it.

According to Simon Harris, video data analysis is extremely important. “CCTV operators receive too much data and do not have time to monitor everything that is happening,” says Harris. “Research has shown that after 22 minutes of surveillance, the operator begins to miss up to 95% of the events on the screen.” Intelligent video will help eliminate the human factor from the surveillance process and increase the effectiveness of security systems.

A «smart» camera allows you to set motion detection parameters, which significantly expands its functionality. In this case, the camera serves as a kind of filter and focuses the operator's attention on precisely defined moments. For example, when observing a uniform flow of people, the camera will clearly react to a person walking against the flow or at a different speed. Analysis of the audio signal allows you to adjust the camera to the sound of breaking glass, a gunshot, as well as noise manifestations of panic or aggression.

The «virtual line» technology has found wide application. For example, a camera scans the area along some object in real time. An installation is set that, for example, one cannot approach the object closer than half a meter — a so-called virtual line is drawn. When a person or any other object crosses this line, the camera immediately transmits a signal to the operator. This technology is already being implemented in the Moscow metro and is aimed at warning drivers about people and objects falling onto the rails.

Network cameras not only have motion and sound detection functions, but also allow you to notify the operator if the camera itself is tampered with. This is especially important in areas that are potentially subject to vandalism, such as schools, prisons, public transport, and in difficult operating conditions where weather conditions, vibrations or high levels of pollution can adversely affect the camera's operation. The active tamper alert function records incidents such as the camera accidentally changing its direction, being blocked or out of focus, and actively responds if the camera is attacked, spray painted or deliberately covered.

Using license plate recognition technology allows you to accurately track the movement of cars on the roads, parking lots and other sites. When you request a specific number, the «smart» camera will instantly transmit information about the wanted car. Unfortunately, there are often cases when people hit by a car are dumped in hospitals. In Moscow, several hospitals were equipped with network surveillance cameras, and now you can track which car brought the victim and identify the owner by the number.

In addition, intelligent technologies offer ample opportunities for merchandising: the camera analyzes the behavior of the visitor and his routes, signals traffic jams and warns that a certain product is running out. Due to the use of a heat map in the cameras, a people counting function is implemented, which allows you to determine the number of people entering or leaving, as well as areas of crowded people.

There are many such examples. Because the penetration of IP video surveillance into all spheres of human activity is an obvious fact. The development of intelligent technologies in CCTV allows to reduce the influence of the human factor and increase the efficiency of the systems themselves. Well, how accurate are the experts' forecasts about the timing of the complete displacement of analog solutions, time will tell. There is not long to wait.