Density traffic flow counting method and system based on space-time counting characteristics

Abstract

The invention provides a density traffic flow counting method and system based on space-time counting characteristics. The density traffic flow counting method comprises the steps: obtaining a trafficvideo, and carrying out the vehicle detection in a selected salient region; constructing a line-crossing probability function to evaluate the probability that the detected vehicle may pass through the counting line, selecting the vehicle needing to be tracked in each frame of image according to the probability value, and extracting the spatial feature and the time feature of the vehicle; constructing a counting model based on a long-short-term recurrent neural network, and analyzing bidirectional space-time counting features in the video to obtain a classification and a corresponding countingresult; and estimating the flow based on the counting result, and calculating to obtain the flow traffic volume, density and speed. According to the density traffic flow counting method, the whole dense traffic flow can be described, and the traffic flow can be accurately determined.

Description

technical field [0001] The disclosure belongs to the technical field of road traffic vehicle flow detection, and relates to a density traffic flow counting method and system based on spatio-temporal counting features. Background technique [0002] The statements in this section merely provide background information related to the present disclosure and do not necessarily constitute prior art. [0003] Accurate and real-time estimation of traffic flow parameters (volume, density and speed) is a key technology in traffic management and intelligent transportation. Estimating these parameters usually requires vehicle counts, and accurate vehicle counts are still difficult in dense traffic scenes. The main methods can be divided into two categories, including sensor-based methods and vision-based methods. Sensor-based approaches often use different types of monitoring equipment, which are often difficult to install and may miss slow-moving vehicles or vehicles that are temporar...

AI Technical Summary

Problems solved by technology

[0004] As far as the inventors are aware, existing LOI counting methods typically require detection and tracking of each vehicle, which is a complex and difficult process

Extracting each vehicle is more complicated when vehicle objects are close to each other

Therefore, the performance of LOI counting in dense traffic scenes is relatively poor

Images