VISION-BASED TRAFFIC PARAMETER ESTIMATION SYSTEM FOR AIR QUALITY MONITORING

摘要

The present invention relates to a vision-based traffic parameter estimation system for air quality monitoring. The present invention includes a road color estimation unit estimating a road color by using color changing information of a road color from an inputted image; a traffic information extraction unit extracting a vehicle speed by using a color difference between the road color estimated through the road color estimation unit and a color of a sampling pixel of a current image, and extracting a vehicle road occupancy rate by using a time covered by vehicles per total measurement time; and a vehicle type classification unit detecting a vehicle object by using a color difference of a current frame of an image and a background model, and classifying vehicles through the matching of vehicle templates.