Research On Traffic Flow Statistics Of Intelligent Transportation Based On YOLO Network

Intelligent transportation is the main trend of urban development at present.Traffic flow statistics,as an important part of intelligent transportation,can provide decision-making data for traffic management departments.In this way,road transportation efficiency can be improved and road patency can be increased.However,the low accuracy of traffic flow statistics is widespread,which restricts the development of intelligent transportation.In this paper,the study of traffic flow statistics,analysis of the key factors affecting the accuracy of traffic flow statistics,respectively from the aspects of vehicle identification,vehicle tracking and vehicle count to improve the accuracy of traffic flow statistics.In the aspect of vehicle identification,YOLO network is adopted to identify vehicles.In view of the shortcomings of YOLO network in vehicle identification,the network model is trained on the manually annotated vehicle identification data set,and the traffic video is used to verify the vehicle identification effect of the trained network model;In terms of vehicle tracking,mean shift algorithm is adopted to track moving vehicles in the detection area.Aiming at the target tracking loss problem of mean shift algorithm,the mean shift algorithm is improved by combining the regression characteristics of YOLO network,and the effectiveness of the improved vehicle tracking algorithm is verified;In the aspect of vehicle counting,a traffic flow statistics method based on YOLO network is proposed.The traffic flow statistics area is set in the road to identify and track the vehicles in the detection area.The counter function completes the tra:ffic flow statistics according to the reg:istration information of the vehicles entering the area and the vehicle leaving information provided.by the tracking algorithm;The experimental system of traffic flow statistics was built to verify the feasibility of the algorithm.The above research results show that the traffic flow statistics algorithm based on YOLO network has good real-time performance and high robustness.It effectively improves the accuracy of traffic flow statistics,which is of great significance to solve the problem of traffic jam and guide urban traffic rationally.