An Efficient Small Traffic Sign Detection Method Based on YOLOv3
- PDF / 1,788,327 Bytes
- 13 Pages / 595.276 x 790.866 pts Page_size
- 103 Downloads / 276 Views
An Efficient Small Traffic Sign Detection Method Based on YOLOv3 Jixiang Wan 1,2
&
Wei Ding 1,2 & Hanlin Zhu 1,2 & Ming Xia 1,2 & Zunkai Huang 1 & Li Tian 1 & Yongxin Zhu 1 & Hui Wang 1
Received: 31 July 2019 / Revised: 2 November 2020 / Accepted: 3 November 2020 # Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract In recent years, target detection framework based on deep learning has made brilliant achievements. However, real-life traffic sign detection remains a great challenge for most of the state-of-the-art object detection methods. The existing deep learning models are inadequate to effectively extract the features of small traffic signs from large images in real-world conditions. In this paper, we address the small traffic sign detection challenge by proposing a novel small traffic sign detection method based on a highly efficient end-to-end deep network model. The proposed method features fast speed and high precision as it attaches three key insights to the established You Only Look Once (YOLOv3) architecture and other correlated algorithms. Besides, network pruning is appropriately introduced to minimize network redundancy and model size while keeping a competitive detection accuracy. Furtherly, four scale prediction branches are also adopted to significantly enrich the feature maps of multi-scales prediction. In our method, we adjust the loss function to balance the contribution of error source to the total loss. The effectiveness, and robustness of the network is further proved with experiments on Tsinghua-Tencent 100 K traffic sign dataset. The experimental results indicate that our proposed method has achieved better accuracy than that of the original YOLOv3 model. Compared with the schemes in relevant literatures our proposed method not only emerges performance superiors in detection recall and accuracy, but also achieves 1.9–2.7x improvement in detection speed. Keywords Computer vision . Convolutional neural networks . YOLO . Traffic sign detection
1 Introduction Traffic sign is meant to be one of the most critical elements in transport systems because it provides instructive or warning messages like road conditions and real-time traffic conditions for vehicles and pedestrians. Complying with traffic sign lawfully can greatly prevent traffic accidents and reduce congestion. For human beings, identifying the traffic sign is an easy task. However, for self-driving cars, locating and classifying the traffic sign accurately and quickly remains an incredible challenge. Therefore, traffic sign detection in autonomous vehicles has been catching the attention from the computer * Zunkai Huang [email protected] * Yongxin Zhu [email protected] * Hui Wang [email protected] 1
Shanghai Advanced Research Institute, Chinese Academy of Sciences, Shanghai 201210, People’s Republic of China
2
University of Chinese Academy of Sciences, Beijing 100049, People’s Republic of China
vision community incessantly for several decades [1–3]. Generally, traditional visual approaches for ob
Data Loading...
