The Improved Framework for Traffic Sign Recognition Using Guided Image Filtering
Loading...
Files
Size: 2.9 MB, File format: Adobe PDF
Date
Authors
Xing, J
Nguyen, M
Qi Yan, W
Supervisor
Item type
Degree name
Journal Title
Journal ISSN
Volume Title
Publisher
Springer Science and Business Media LLC
Abstract
In the lighting conditions such as hazing, raining, and weak lighting condition, the accuracy of traffic sign recognition is not very high due to missed detection or incorrect positioning. In this article, we propose a traffic sign recognition (TSR) algorithm based on Faster R-CNN and YOLOv5. The road signs were detected from the driver’s point of view and the view was assisted by satellite images. First, we conduct image preprocessing by using guided image filtering for the input image to remove noises. Second, the processed image is input into the proposed networks for model training and testing. Three datasets are employed to verify the effectiveness of the proposed method finally. The outcomes of the traffic sign recognition are promising.
Description
Keywords
Traffic sign recognition, Faster R-CNN, GTSDB dataset, FRIDA database
Source
SN Computer Science. 3, 461 (2022). https://doi.org/10.1007/s42979-022-01355-y
Publisher's version
Rights statement
Creative Commons Attribution 4.0 International
Permanent link
Endorsement
Review
Supplemented By
Referenced By
Creative Commons license
Except where otherwise noted, this item's license is described as Creative Commons Attribution 4.0 International