In order to study the technical issues of traffic safety early warning on roads with poor sight distance, we first collect the sounds of vehicles running on the road, select representative vehicle sounds as recognition objects, pre-emphasize the vehicle sound signals, add windows into frames, and calculate the power spectrum. Input the Mel filter bank to obtain the MFCC of the vehicle sound signal and construct a feature vector with its characteristic values; then the BP neural network algorithm is improved to classify and identify the vehicle signal, thereby achieving the purpose of vehicle identification. Experiments show that the accuracy of the proposed voice recognition technology reaches more than 90%. This technology can be applied to road sections with poor sight distance to identify passing vehicles.
Research on Safety Early Warning Technology for Road Sections with Poor Sight Distance based on Acoustic Signals
Advances in Engineering res
International Symposium on Traffic Transportation and Civil Architecture ; 2024 ; Tianjin, China June 21, 2024 - June 23, 2024
Proceedings of the 2024 7th International Symposium on Traffic Transportation and Civil Architecture (ISTTCA 2024) ; Chapter : 39 ; 392-399
2024-09-24
8 pages
Article/Chapter (Book)
Electronic Resource
English
Guiding device and operation method for high-curvature road with poor sight distance
European Patent Office | 2025
|