Traffic signal lamp control method based on deep convolution neural network

Abstract

The invention discloses a traffic signal lamp control method based on a deep convolution neural network. The system hardware comprises a dead-corner-free omni-directional vision sensor used for getting a panoramic video image of a whole intersection, a computer used for segmenting the video image, analyzing the state of traffic congestion in a key place using a deep convolution neural network, and sending a processing result and a control command to a signal lamp controller, and the signal lamp controller used for receiving the control command of the computer and controlling the signal lamps at the intersection. Through the method, violation of the traffic rules 'green for go and red for stop' under traffic congestion at intersections is solved effectively, the occurrence of large-area traffic congestion caused by running the green lamp is prevented effectively, and the difficulty and intensity of law enforcement by the traffic police are reduced greatly.

Description

technical field [0001] The invention relates to the application of computer technology, pattern recognition, artificial intelligence, applied mathematics and biological vision technology in the field of intelligent transportation, and in particular to a traffic signal light control method based on a deep convolutional neural network. Background technique [0002] At present, there are red, yellow and green indicator lights, "arrow" signal lights, multi-phase signal lights, etc. on the roads of cities in our country; with the sharp increase in the number of vehicles, there will often be congestion during the peak traffic period due to poor clearing and overflow of downstream queuing Question: According to the new traffic regulations, motor vehicles should stop outside the intersection and wait when there is a traffic jam at the intersection ahead. At present, there will be some corresponding warning signs above the signal lights at the intersection to inform the driver. It is...

AI Technical Summary

Problems solved by technology

[0010] To sum up, there are still several thorny problems in the road condition recognition using computer vision: 1) How to accurately locate and segment the overall image of the vehicle under test from the complex background and accurately frame each frame of the image 2) How to reduce the influence of conditions such as weather and light, and increase the adaptability and robustness of the system; 3) How to balance recognition accuracy and detection efficiency, while minimizing training and learning time

Images