Traffic sign recognition (TSR) in automotive driving ensures smooth driving and prevents accidents by collecting and recognizing the road traffic sign information, giving instructions or warnings, and making operations more manageable and safer. Most current traffic recognition methods are based on machine learning methods and perform well. However, Road traffic sign recognition in Automatic extreme environments needs to be further emphasized. In this paper, CNN is used to recognize road traffic signs under extreme weather conditions to help drivers deal with barriers in foggy or highlight situations. This technique provides an efficient way for drivers to identify the traffic signs in the foggy or highlight conditions, decreasing the frequency of traffic wrecks. Firstly, we adopt the methods to blur the picture, and then the processed image is used for training. A multi-layer convolution neural network learning algorithm is proposed. The results of the tests show that the average recognition rate of the algorithm for traffic signs in the natural environment is 97.1%, and the false recognition rate is less than 1%. The results of the experiments suggest that the algorithm we employed in our research works has an accurate recognition degree and can improve the recognition speed of the existing algorithms.
Road Traffic Signs Recognition in Automatic Driving under Extreme Environments
CAIBDA 2022 - 2nd International Conference on Artificial Intelligence, Big Data and Algorithms ; 2022 ; Nanjing, China
CAIBDA 2022 ; 4
2022-01-01
4 pages
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Automatic Recognition of Road Signs
SPIE | 2002
|Real-time detection and recognition of road traffic signs
Online Contents | 2012
|Evaluation of Road Traffic Signs
NTIS | 1981
|