Road Intersection Monitoring from Video with Large Perspective Deformation
Takashi Furuya, Camillo Jose Taylor · 2014
The authors propose a method to extract traffic information from videos of road intersections, taken from a low-resolution camera fixed 2 to 7m above ground. First, feature points are detected and tracked with the Kanade-Lucas-Tomasi tracker to obtain trajectories. These trajectories are then segmented into groups corresponding to vehicles so that the vehicle’s path, velocity, time of entry and exit can be determined.