Description

The research laboratory complex is designed to create digital twins of road network nodes in order to carry out simulation modeling of the road situation and work out measures to optimize the parameters of the site. The software contains tools for modeling traffic flow with subsequent capture of a synthesized image of the road situation and its further processing using convolutional neural network technologies. The resulting images can be used to solve problems of classification, accounting and logging of traffic flow parameters using neural networks.
The data generation module supports connecting a synthesized video stream from a virtual camera. The video stream is obtained by rasterizing images of the road situation, synthesized in real time using the complex’s three-dimensional engine. The software package also contains a mathematical model of road flow. Flow parameters can be adjusted in real time, changes reflected in a visual demonstration followed by image capture and objective flow metrics.