NII MVS SFU: Video system for early detection of drones

The main articles are:

• Video analytics (terms, applications, technologies)
• Unmanned aerial vehicle (drone, UAV)

2023: Development of an early drone detection video system

Scientists from Taganrog have developed a technology for detecting, identifying and escorting group or single unmanned aerial vehicles (UAVs) on video obtained from cameras and thermal imagers, and have created a mobile complex based on it, which is quickly installed in an urban environment and works effectively in various weather conditions. This was announced on January 16, 2023 by the author of the project - head of the laboratory of the Research Institute of Multiprocessor Computing Systems named after Academician A.V. Kalyaev of the Southern Federal University Anatoly Kalyaev.

Illustration: dzen.ru

The rapid development of unmanned aerial vehicle technologies has led to the fact that UAVs are increasingly used for drawings attacks critical targets and unauthorized tracking of specially protected areas. Therefore, the problem of early detection and identification of UAVs and their groups has become very urgent. In January 2023, expensive radar systems are used for this, but they are not able to effectively identify small targets, and in addition, they are ineffective in an urban environment, explained Anatoly Kalyaev.

The detection system is based on the use of television and thermal imaging cameras, the images obtained from them are processed using a number of methods and algorithms.

Small drones, up to 40 centimeters, the radar sees poorly, so you have to work with a video signal, and image processing is a very difficult business, associated with time-consuming calculations. To solve this problem through a mobile complex, it was necessary to create parallel and multithreaded methods and algorithms that allow improving the quality of video, highlighting contrasting objects, filtering out false ones and identifying the remaining ones. In simple words, the system works like this: first, images are fixed and improved, then objects are searched, identified, coordinates and directions of movement are determined, then it is decided what to do with it. For example, civilian UAVs can be grounded, specified scientist.

The technology created by Anatoly Kalyaev is already used in the creation of some prototypes of promising Russian systems for automatic video detection and identification of air targets for various purposes.

Tests of prototypes have shown that the use of technology makes it possible to process large amounts of incoming data in real time (uncompressed video up to gigabit per second) and analyze information about the goals currently visible, noted Anatoly Kalyaev.

Anatoly Kalyaev stressed that the scope of its development is quite wide, but first of all, it is the protection of the perimeter of critical infrastructure facilities, for example, nuclear power plants and industrial enterprises.

In January 2023, Anatoly Kalyaev, together with his colleague Maxim Khisamutdinov, continues to work on the project. Scientists are engaged in the creation of a universal hardware and software distributed video processing complex. It will allow you to design on its basis many different UAV surveillance systems, consisting of a large number of cameras and many computers that automatically distribute video processing among themselves depending on tasks: search, identification, coordinate calculation and others.

Шаблон:Quote 'author = scientist emphasized.