Traffic jam area real-time detection method based on deep learning

Abstract

The invention discloses a traffic jam area real-time detection method based on deep learning. The traffic jam area real-time detection method comprises the following steps: 1, setting a camera preset position and calibrating a camera; 2, initializing a convolutional neural network model; 3, acquiring a current video frame and video frame time; 4, checking the working state of the camera; 5, performing vehicle target detection on the ROI by using a convolutional neural network model; 6, maintaining a stationary target tracking queue; 7, detecting a congestion area; and 8, congestion status reporting. The congestion detection area congestion judgment algorithm provided by the invention has relatively strong robustness to environmental changes, and realizes a real-time detection effect and relatively high congestion identification precision.

Description

technical field [0001] The invention relates to the technical field of deep learning, in particular to a method for real-time detection of traffic jam areas based on deep learning. Background technique [0002] In recent years, with the popularization of automobiles, the capacity of urban roads has been insufficient, and with improper design and too many road intersections, the problem of traffic congestion has become extremely prominent. Traffic congestion will not only affect people's travel efficiency, but also Even more can cause serious traffic accident, seriously endanger people's travel safety. Therefore, it is particularly important to accurately detect and report traffic congestion in real time. [0003] At this stage, there are mainly three solutions proposed by researchers for the traffic congestion problem. The first one is to predict congestion based on the statistical characteristics of road traffic flow, such as historical traffic volume, headway, average spe...

AI Technical Summary

Problems solved by technology

[0005] In order to overcome the shortcomings of low detection accuracy and slow detection speed in the prior art, the present invention provides a real-time detection method for traffic congestion areas based on deep learning, using deep convolutional neural network (CNN) features for vehicle target detection, and integrating video spatiotemporal Continuity for accurate and fast congestion area detection

Images