Young scientists from KBSU help the Ministry of Emergency Situations

KBP traditionally introduces readers to the participants in the exhibition of innovative projects of young scientists, which took place at KBSU on Russian Science Day. The first of them is a project to create a system for determining the area of fires by analyzing video surveillance data.
The project was recognized as the winner in the “Best Innovative Project” category, and its name is “System for determining the area of a fire from a video image based on convolutional neural networks in real time.” The author is a master’s student at KBSU, Institute of Artificial Intelligence and Digital Technologies Astemir Kalmykov.

“The project is the development of a system that uses computer vision and machine learning algorithms to analyze the video stream from a camera located above a potential fire,” explains the author. – The processed data is displayed on the screen indicating the area of the fire, its area and scale. By doing so, we hope to improve the efficiency and safety of firefighting operations. The system is intended for the Ministry of Emergency Situations and relevant ministries.”
The essence of the project is to create a system for recognizing and classifying the scale of fires based on video analysis. In practice, this is a software and hardware complex that gives an accurate determination of the area and scale of fires. In fact, it is a combination of specialized software using computer vision and machine learning algorithms with CCTV cameras or drones.

The process goes something like this: an observer takes a position in the fire area and launches the drone. It transmits information to the laptop screen. The data obtained helps the operator decide whether to call in additional forces or make do with the existing ones. At the moment, information is displayed on the screen in the form of data on the fire area and location. In the future, there is an idea to add here the characteristics of fires into which they are divided by the Ministry of Emergency Situations.

In principle, any drone will do here. But a high-resolution camera is needed. The height at which the camera is located and the characteristics of the camera itself are also important: knowing them, you can calculate the area of the fire.

The author also had to work on “training” the software:

“To begin with, I took one of the neural network models for pattern recognition. Then I found videos and images of fire and fires, and trained a neural network to recognize images of fire in images. I tested how the neural network itself copes with this. Then he mathematically read and converted the pictures into text information and mathematical formulas. And displayed the information on the screen.”
The idea of creating the system was suggested to the author by news reports:

“Today, as you know, there are many reports of fires. Recently, for example, a warehouse of an online retail network burned down. In summer, forests often burn. And in general, the problem of fires is relevant everywhere and always: fires can spread quickly and lead to serious consequences, including damage to the environment, loss of property and a threat to human life. So I became interested in whether it was possible to minimize the risks of fire spread using modern methods of artificial intelligence and neural networks. Well, I formulated a specific topic when I came across the possibilities of neural networks on the Internet and decided to develop this idea.”
Let us remind you that the traditional exhibition of innovative projects of young scientists from the North Caucasus took place at KBSU for the eighteenth time. During the event, over 130 innovative projects in agriculture, medicine, construction, industry, the social sphere, and production processes were presented. A number of projects concerned the creation of new generation materials, robotics and innovative information technologies. Their authors were young scientists from Kabardino-Balkaria, Stavropol Territory, Ingushetia, South Ossetia and Rostov-on-Don. Interesting and promising developments were presented not only by undergraduate and graduate students, but also by schoolchildren. More than 60 projects belong to young scientists from KBSU.
A panel of judges determined the winners in the categories “Best Innovative Project”, “Best Innovative Product”, “Best Innovative Idea”, “Best Project Presentation”, “Best Business Idea”, “Best Social Project”.

Askhat Mechiev. Photo by the author

Source: Electronic newspaper Kabardino-Balkarian Pravda dated February 15, 2024