The granular computing implementation for road traffic video-detector sampling rate finding

The method discussed in this contribution allows to estimate the necessary data granularity for an on-line traffic controlling, using the information recorded by digital video-camera. Due to define the data sampling rate modelling and analysis methods were applied. They are used for extracting prediction rules of the traffic descriptors. The discussed scheme combines granular computing algorithms with assumptions of a cellular automata traffic model. It enables direct determination of temporal characteristics for the recognised and extracted traffic states. The traffic parameters prediction algorithm was introduced that allow determining the sampling time intervals of the video detection system.