Abstract
Traffic sign detection is an important task in traffic sign recognition systems. Chinese traffic signs have their unique features compared with traffic signs of other countries. Convolutional neural networks (CNNs) have achieved a breakthrough in computer vision tasks and made great success in traffic sign classification. In this paper, we present a Chinese traffic sign detection algorithm based on a deep convolutional network. To achieve real-time Chinese traffic sign detection, we propose an end-to-end convolutional network inspired by YOLOv2. In view of the characteristics of traffic signs, we take the multiple 1 × 1 convolutional layers in intermediate layers of the network and decrease the convolutional layers in top layers to reduce the computational complexity. For effectively detecting small traffic signs, we divide the input images into dense grids to obtain finer feature maps. Moreover, we expand the Chinese traffic sign dataset (CTSD) and improve the marker information, which is available online. All experimental results evaluated according to our expanded CTSD and German Traffic Sign Detection Benchmark (GTSDB) indicate that the proposed method is the faster and more robust. The fastest detection speed achieved was 0.017 s per image.
Highlights
Traffic Sign Recognition Systems (TSRs) are important for highway maintenance, driver-assistance systems, and self-driving cars [1]
The research of traffic sign recognition is divided into sign detection and sign classification, which both have been discussed for a long time; in particular, the latter has achieved greater success on the German Traffic Sign Recognition Benchmark (GTSRB)
Concerning the test set for CCTSDB—excepting tiny traffic signs that are not clear enough to divide into subclasses—we found that the test set contained a total of 209 mandatory signs, 376 prohibitory signs, and 151 danger signs
Summary
Traffic Sign Recognition Systems (TSRs) are important for highway maintenance, driver-assistance systems, and self-driving cars [1]. For traffic sign detection, the CNN-based method [3] is not superior to real-time performance due to the complex computation. Types of convolutional networks for object detection have emerged and been proven to be efficient in processing unit (GPU). The series of methods based on regression exhibits real-time performance; for example, many new types offields. The series of 45 methods based on regression exhibits real-time performance; for example, the speed of can reach above frames per second. Figures and show some samples of the three super categories in Germany traffic sign dataset. Details shouldwhen be taken into account when detection methods aresign applied to Chinese traffic signs. Account when detection methods are applied to Chinese traffic signs
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
