Research on traffic congestion detection methods based on deep learning

With the rapid development of China's economy and the advancement of urbanization, China's per capita automobile ownership rate is also increasing year by year. But with the increasing number of cars, the problem of traffic congestion in cities is becoming increasingly serious. The prediction and analysis of the road status can provide a basis for the urban road traffic planning and guidance. Therefore, it is necessary to study the traffic prediction and improve the accuracy of traffic state prediction. Deep learning can learn and extract features from massive data; suitable for various fields, such as image recognition, speech recognition, natural language processing, etc., with good scalability and flexibility, can improve performance by increasing the network level and the number of neurons, and can conduct autonomous learning and iterative optimization. With the increase of data volume and computing power, the performance can be continuously improved. Therefore, in this paper, deep learning is applied to the field of traffic state prediction, which is based on residual network and VGG16 network. The final experimental results show that the classification accuracy of the residual network can reach 92.11%, and the accuracy of the VGG16 network can reach 89.60%, so the residual network is more suitable for the field of traffic congestion detection.