Artificial intelligence-based traffic signal light recognition method, device, equipment and medium

Abstract

The present invention relates to the technical field of artificial intelligence and image detection, and provides an artificial intelligence-based traffic signal recognition method, device, equipment and medium, which can use the darknet53 network to extract the target feature information of the resized target image and input it to the Mixup algorithm And the traffic signal recognition model trained by the residual attention network, output the target feature map to accurately extract the detailed features, improve the recall rate and accuracy rate, obtain the target anchor box for recognition on each target feature map, and output the target anchor box coordinates and Target score, the target anchor box coordinates with the highest score are used as the predicted position coordinates and mapped to the image to be recognized to obtain the recognition result, and then the automatic recognition of traffic lights is realized based on artificial intelligence means, and the recognition accuracy is higher. The invention also relates to block chain technology, and the recognition results and traffic signal light recognition models can be stored in the block chain. The present invention can also be applied to smart traffic scenarios, thereby promoting the construction of smart cities.

Description

technical field [0001] The present invention relates to the technical field of artificial intelligence and image detection, in particular to an artificial intelligence-based traffic signal recognition method, device, equipment and medium. Background technique [0002] With the increasing number of illegal activities such as running red lights, the rapid positioning and identification of signal lights in traffic checkpoint images has become an extremely important and challenging task in urban traffic management. [0003] The traditional YOLOv3 detection algorithm has become one of the mainstream target detection algorithms in the industry due to its advantages of fast detection speed and good accuracy. There is still a lot of room for improvement in the effect, especially for detection and recognition in complex scenes such as night or haze, the effect is weak. Contents of the invention [0004] In view of the above, it is necessary to provide an artificial intelligence-ba...

AI Technical Summary

Problems solved by technology

[0003] The traditional YOLOv3 detection algorithm has become one of the mainstream target detection algorithms in the industry due to its advantages of fast detection speed and good accuracy. There is still a lot of room for improvement in the effect, especially for detection and recognition in complex scenes such as night or haze, the effect is weak

Images