With the rapid development of deep learning methods in the field of computer vision, the effective methods can be applied to the classification and positioning of traffic targets based on real traffic data. This paper studies the application of convolutional neural network in the real traffic information extraction field. The YOLOv3 target detection algorithm, the Faster RCNN target detection algorithm and the SSD target detection algorithm are used to detect vehicles in the images collected in same road condition, and generate the result graph of the target detection. The experimental results show that the YOLOv3 target detection algorithm, Faster RCNN target detection algorithm and SSD target detection algorithm can identify all traffic targets in the traffic road images. It reveals the three algorithms meet the detection needs, and help to achieve non-contact traffic flow collection and feature recognition. Meanwhile these algorithms can provide technical means for traffic supervision and safety improvement.
Research on Traffic Information Extraction Approach Based on Image Deep Learning
2021-10-22
3073894 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Traffic distribution prediction method based on feature extraction and deep learning
Europäisches Patentamt | 2021
|Traffic Rule Violation Detection System: Deep Learning Approach
Springer Verlag | 2022
|