Road condition monitoring system based on YOLOV5

Abstract

The invention provides a road condition monitoring system based on YOLOV5. The road condition monitoring system comprises an edge terminal, a message middleware module and a central platform, the system combines video analysis and edge computing technologies, can monitor various road targets at one time, including trucks, large buses, medium buses, SUVs, cars, minibuses and the like, and has the characteristics of high recognition accuracy, strong real-time performance, wide monitoring range, low cost and the like.

Description

technical field [0001] The invention relates to the field of intelligent traffic and road condition monitoring, in particular to a road condition monitoring system based on YOLOV5. Background technique [0002] With the progress of society and the improvement of people's living standards, the number of motor vehicles has increased rapidly, and the demand for traffic has also increased sharply. Traffic congestion occurs from time to time, especially in cities. The problem of traffic congestion has become one of the urgent problems in our major cities. . Intelligent transportation system is emerging to solve this problem and is currently in a stage of rapid development. Comprehensive, real-time and accurate collection of traffic information is the basic guarantee of intelligent transportation. At present, there are three main technologies for traditional traffic road information collection, one is a permanent embedded system based on a piezoelectric loop, and the other is a ...

AI Technical Summary

Problems solved by technology

[0003] Among the above-mentioned three traditional traffic road condition information collection technologies, the piezoelectric circuit-based permanently embedded system needs to permanently bury the sensor underground, and its construction and maintenance require excavation of the road surface, which is time-consuming and laborious and affects traffic, and the detection accuracy is limited by the road surface. Deformation is greatly affected; and another suspension-based system, although relatively based on piezoelectric circuit technology, does not need to damage the road surface during installation and maintenance and is not affected by road surface deformation, but its detection is greatly affected by the environment, such as in strong winds. The detection accuracy will be greatly reduced in weather conditions. When the wind exceeds level 6, the system is basically unusable. At the same time, when non-vehicle objects appear under the probe, it is easy to cause false detection. Finally, the third traffic flow detection system based on image processing technology, due to It has the advantages of simple installation, convenient maintenance, no damage to the road surface, wide detection coverage and many detection parameters, etc. Compared with other methods, it has obvious advantages, but this method is very dependent on the performance of the video analysis server, and it is necessary to realize large-scale real-time monitoring , a large number of expensive high-performance GPU servers need to be configured, which seriously affects its wide application and promotion

Images