Detection and classification of vehicles using selected methods of image processing

Methods of vehicles detection and classification using image processing are becoming increasingly popular, especially due to their non-invasiveness in the road surface and relatively lower installation and maintenance costs. These methods are commonly used in traffic flow monitoring systems and detection of vehicles with specific parameters. Importantly, the use of video analytics methods is still characterized by sensitivity to external disturbances such as variable weather conditions. The work discusses selected data processing mechanisms that have been applied within the functioning vehicle recognition subsystem. As part of the analysis, the effectiveness of the applied solutions and sensitivity to the occurring weather conditions were assessed.