Implementation of Adaptive Metrics for Vehicle Counting Algorithm via ViDeS Video Monitoring System

In recent decades, with the development of computing technology, decreasing size and lowering the cost of the component base of devices, it is becoming cheaper and easier to develop systems for capturing, processing and analyzing information about the elements of transport infrastructure. With the growth of cities, increasing population and motorization rates, traffic congestion is becoming more and more of an issue, resulting in lost time, increased pollution and increased likelihood of traffic accidents. The paper presents the application of adaptive metrics for vehicle counting based on the virtual detector system - «ViDES». The algorithm of the system operation, the equipment used and the set of video data for which the study was conducted are described. The object of the study is the influence of metrics implemented in the algorithm of vehicle detection on the accuracy of their counting. The obtained data can be used for more flexible tuning of the developed system, which should lead to an increase in the accuracy of recognizing the characteristics of traffic flows.