Application Analysis of Vehicle Type Identification System in Nanping Expressway Toll Station
In order to meet the increasing requirements of the current expressway toll station on the model identification system,it is necessary to make full use of modern high-tech optimization of the corresponding system,explore the vehicle classification algorithm based on video image technology,improve the application effectiveness of the vehicle identification system,and provide necessary ideas for improving the efficiency of the expressway toll station.Based on this,the data set construction and vehicle classification network de-sign were analyzed,and the key points of vehicle classification label design and image preprocessing were explored.At the same time,the feature extraction network was explored,and it was proposed to optimize the EfficientNet V2 network,and the network structure reparameterization technology was designed to improve the image recognition accuracy.According to the actual research results,the rec-ognition accuracy of the vehicle type recognition system of the proposed algorithm has been significantly improved,reaching 98.4%.At the same time,the image recognition speed has also been improved to 26 ms with the support of network structure reparameterization technology.
expressway toll stationvehicle type identification systemimage processing
NETL
NSTL
万方数据
