The invention discloses a road-section-
pedestrian-crossing control method based on a
machine vision, and belongs to the field of intelligent traffic safety management. The method includes the steps that firstly, related data of road-section pedestrians and motor vehicles is collected, and characteristic data such as the waiting quantity of the pedestrians, watering time of the pedestrians,
road traffic flow and the queuing length of the motor vehicles is further identified and calculated; secondly, normalization
processing is carried out on the basis that the longest
waiting time of the pedestrians, the maximal waiting number of people, the maximal vehicle flow and the longest queuing length of the motor vehicles are set; finally, the going-through priority range is divided by determiningdifferent threshold values. When
traffic conflict is generated between the road-section pedestrians and the motor vehicles, characteristic threshold values can be compared through a control center, going-through priorities of the pedestrians and the motor vehicles are reasonably distributed accordingly, and corresponding
signal lamp controlling is carried out. The method is simple in design, roadtraffic conditions are obtained and analyzed in real time and comprehensively considered through the
machine vision, people-vehicle conflict can be effectively reduced, and the occurrence rate of rear-end accidents of motor vehicles is decreased while the
pedestrian safety is guaranteed; the method has the extremely-practical
engineering application value in the aspect of traffic safety management.