Statistical method of traffic flow

Abstract

The invention discloses a statistical method of traffic flow. The method includes the steps of defining a lane which in the shape of a closed polygon; detecting vehicles of the current frame, and forming an observation list for the vehicles of the current frame, locating in the lane; predicting the states of the vehicles in a prior vehicle list; calculating correlation between the vehicles in the observation list and the vehicles in the prior vehicle list, and according to the different correlations, updating the states of the vehicles in the prior vehicle list, correlated to the vehicles in the observation list, by the states of the vehicles in the observation list, adding the vehicles in the observation list to the prior vehicle list or deleting the corresponding vehicles in the prior vehicle list. The states of the vehicles in the prior vehicle list include location, speed and size of the vehicles in the prior vehicle list. According to the statistical method of vehicles, all vehicles locating on the lane are accurately tracked by comparing the vehicles in the observation list to the vehicles in the prior vehicle list, so that accurate traffic flow statistics is achieved.

Description

technical field [0001] The invention relates to the technical field of traffic control, in particular to a traffic flow statistics method. Background technique [0002] Traffic flow statistics is an important function and task in intelligent transportation, which is of great significance to management departments and drivers. Traffic flow statistics methods include methods based on magnetic induction coils, methods based on electromagnetic waves (radar) and methods based on video. [0003] The method based on the magnetic induction coil needs to embed the magnetic induction coil under the road surface, which will damage the road, and the cost is high and the maintenance is complicated. The method based on electromagnetic waves measures the traffic flow through electromagnetic waves by setting radar above the road, and the accuracy is poor in the case of vehicle congestion. [0004] Traditional video-based methods use background modeling technology to detect moving vehicles...

AI Technical Summary

Problems solved by technology

[0003] The method based on the magnetic induction coil needs to embed the magnetic induction coil under the road surface, which will damage the road, and the cost is high and the maintenance is complicated

The method based on electromagnetic waves measures the traffic flow through electromagnetic waves by setting radar over the road. In the case of vehicle congestion, the accuracy is poor.

[0004] Traditional video-based methods use background modeling technology to detect moving vehicles on the road, which are greatly affected by light, shadows, and bad weather. When vehicles are congested, they cannot accurately segment each vehicle, and the accuracy is poor.

Images