Method and system for detecting traffic jam in monitoring video based on AI deep learning

Abstract

The invention provides a method and device for detecting traffic jam in a monitoring video based on AI deep learning. The method comprises the steps of: (1), obtaining a moving vehicle in a video through a moving object detection method; (2), tracking the moving vehicle; (3), on the basis of the step (2), obtaining traffic characteristic parameters, and establishing a traffic congestion model based on deep learning; and (4), calculating a road jam index delta, and judging the jam level of a road section. The system comprises a traffic monitoring video obtaining system, a traffic monitoring video identification processing system and a road jam judgement system. According to the method and system for detecting traffic jam in the monitoring video based on AI deep learning in the invention, automatic perception, comprehensive perception and precise perception of all kinds of traffic jams can be realized; furthermore, the cost is low; the efficiency is high; a new solution is provided for highway management; and a solid foundation is established for a highway to increase the management efficiency.

Description

technical field [0001] The invention relates to the field of artificial intelligence, in particular to a method and system for detecting traffic congestion in surveillance video based on AI deep learning. Background technique [0002] In the traditional traffic flow research, it is believed that traffic flow has three basic characteristic parameters: flow rate, speed and density, and there is a close relationship among them. The accuracy is not high, and the pairwise relationship model between them is called the relationship model of the basic parameters of traffic flow. [0003] (1) Vehicle speed [0004] The vehicle speed V refers to the average speed of a vehicle running on a road. [0005] (2) Vehicle density [0006] Vehicle density D refers to the ratio of the number of vehicles on the road to the length of the road. [0007] (3) Congestion index [0008] The road congestion index δ refers to the degree of road congestion. The higher the vehicle speed and the lowe...

AI Technical Summary

Problems solved by technology

The existing methods have the following problems: damage to the road surface, and the construction is difficult, which is more likely to cause traffic congestion; there is a certain false detection rate; the cost is high and the management is difficult

[0010] With the continuous improvement of AI algorithms and the continuous improvement of chip performance, AI technology is becoming more and more mature, and major manufacturers have begun to study AI algorithms. However, the traffic congestion detection method in surveillance video based on AI deep learning is still relatively blank.

Images