SOFTWARE AND HARDWARE COMPLEX OF ACCESS CONTROL TO CRITICAL OBJECTS USING COMPUTER VISION TECHNOLOGIES

Abstract

The article proposes a software and hardware complex for controlling access to critical objects using Computer Vision technologies. The hardware component is implemented at the level of project solutions. The software component has a practical implementation.

The technological basis of the software component is the Computer Vision methods, models and algorithms adapted to a specific software product. The software component was implemented using the Python programming language and relevant libraries with the integration of information flows from several surveillance cameras. The software and hardware complex of access control to critical objects is based on the use of technologies of distributed computer systems and networks.

An example of the implementation of the proposed solutions provides the functions of controlling the access of cars to parking spaces of critical infrastructure objects (shopping and entertainment centers; sports / concert venues; institutions of higher education; residential complexes, etc.); blocking of unregistered users; detection of violators. Identification is carried out by license plates and, if necessary, by other indicators.

The feature of the proposed development is as follows. The system provides control of the internal and external perimeter of the critical object; object tracking is built on the use of YOLOv8 convolutional neural network, which provides object identification in the Object Tracking process based on a certain number of frames; the training of the network was carried out according to its own Data set array. The development has a practical implementation.