Traffic flow statistical method based on road video and license plate detection

Abstract

The invention belongs to the technical field of intelligent traffic systems and computer vision, and discloses a traffic flow statistical method based on road video and license plate detection. The method comprises the steps of initializing, calibrating a license plate detection area, labeling the type of the license plate detection area, and drawing a vehicle tracking cut-off line; acquiring a current frame of the camera; detecting a new license plate image, and adding the new license plate image into a tracking list; performing matching tracking on the old license plate image, and updating tracking vehicle list information; and making a judgment when the vehicle is driven out of the tracking cut-off line and updating vehicle statistical information. Compared with a road traffic flow statistical method based on vehicle motion characteristics and vehicle detection and tracking, the road traffic flow statistical method based on license plate detection and tracking is higher in anti-interference capability, wider in universality, higher in detection accuracy and capable of meeting the real-time requirement.

Description

technical field [0001] The invention belongs to the technical fields of intelligent traffic systems and computer vision, and in particular relates to a traffic flow statistics method based on road video and license plate detection. Background technique [0002] At present, the existing technologies commonly used in the industry are as follows: With the rapid development of society and economy, the number of vehicles has increased sharply, and intelligent transportation systems have emerged as the times require. In order to observe road traffic conditions in a more comprehensive and timely manner, most cities in my country have deployed real-time road monitoring systems to deter illegal driving, observe road conditions, and guide traffic. Among them, the traffic flow detection system based on road video is an important link in the intelligent transportation system, which lays the foundation for other systems. [0003] There are two main types of current road vehicle flow sta...

AI Technical Summary

Problems solved by technology

[0004] To sum up, the problems existing in the existing technology are: there are many shapes and colors of existing vehicles, resulting in poor versatility of existing vehicle detection; the shadows, lights, and road shadows of vehicles have similar characteristics to vehicles. characteristics, causing great interference to vehicle tracking, making the accuracy of road traffic flow statistics based on vehicle detection and vehicle tracking low

Especially in low light, the accuracy of vehicle detection and vehicle tracking is significantly reduced, and the characteristics of vehicle movement are not obvious, resulting in inaccurate traffic statistics

Images