The article presents the results of the development and research of a software module for monitoring and managing the access of vehicles to the territory. Based on the analysis of literary sources, an algorithm for recognizing state license plates was developed, its software implementation was completed, its own data set was formed, and a neural network model was trained. The functionality of the program lies in the application of one-stage detection methods to solving the problem of recognizing license plates of vehicles. Thanks to this, the software product can be used for educational and research purposes, and can also be used both standalone and connected to building security systems as an embedded module. In the course of a comparative analysis of the developed software module with similar software products, it was concluded that the use of the developed model is appropriate. It can be used to recognize license plates in real time, showing good accuracy and completeness, as well as excellent frame processing speed. As a result, a software module for vehicle access control and management based on license plate recognition was implemented and studied, which has the following advantages: the ability to work on devices with low computing power; high speed image processing; possibility of integration with other software products; low cost.