Artificial intelligence-based vehicle brand identification method and device, equipment and medium

Abstract

The invention relates to the technical field of artificial intelligence, and provides an artificial intelligence-based vehicle brand identification method and device, equipment and a medium, which adopt a multi-scale parallel structure under an intersection-parallel ratio, can effectively alleviate the problem of sample imbalance of simple and difficult scenes, significantly improve the vehicle brand identification accuracy under the difficult scenes, and the cross-parallel ratio loss is increased; the position and the size of the detection frame can be fitted more accurately, the false recallphenomenon of non-vehicles can be effectively reduced, the overall recall rate and accuracy of vehicle brand identification are improved, meanwhile, the loss function can evaluate the loss of the model from multiple aspects, the training effect of the model is better, and thus, automatic identification of the vehicle brand is realized. In addition, the invention can also be applied to smart traffic, thereby promoting the construction of smart cities. The invention also relates to a blockchain technology, and an identification result can be stored in the blockchain node.

Description

technical field [0001] The present invention relates to the technical field of artificial intelligence, in particular to an artificial intelligence-based car brand recognition method, device, equipment and medium. Background technique [0002] With the continuous increase of vehicle violations, the rapid positioning and identification of vehicles in traffic checkpoints has become a very important task in urban traffic management. [0003] The traditional car brand recognition mainly uses the YOLOv3 algorithm. However, the YOLOv3 algorithm is only suitable for simple scenes such as sunny days, daytime, front of the car, and the back of the car. For difficult scenes such as haze, rainy days, night, and the side of the car, the accuracy of the YOLOv3 algorithm Both the recall rate and the recall rate are relatively low, and there is still a lot of room for improvement. Contents of the invention [0004] In view of the above, it is necessary to provide an artificial intellige...

AI Technical Summary

Problems solved by technology

[0003] The traditional car brand recognition mainly uses the YOLOv3 algorithm. However, the YOLOv3 algorithm is only suitable for simple scenes such as sunny days, daytime, front of the car, and the back of the car. For difficult scenes such as haze, rainy days, night, and the side of the car, the accuracy of the YOLOv3 algorithm Both the recall rate and the recall rate are relatively low, and there is still a lot of room for improvement

Images