Traffic Sign Recognition Based on the YOLOv3 Algorithm

Abstract

Traffic sign detection is an essential component of an intelligent transportation system, since it provides critical road traffic data for vehicle decision-making and control. To solve the challenges of small traffic signs, inconspicuous characteristics, and low detection accuracy, a traffic sign recognition method based on improved (You Only Look Once v3) YOLOv3 is proposed. The spatial pyramid pooling structure is fused into the YOLOv3 network structure to achieve the fusion of local features and global features, and the fourth feature prediction scale of 152 × 152 size is introduced to make full use of the shallow features in the network to predict small targets. Furthermore, the bounding box regression is more stable when the distance-IoU (DIoU) loss is used, which takes into account the distance between the target and anchor, the overlap rate, and the scale. The Tsinghua–Tencent 100K (TT100K) traffic sign dataset’s 12 anchors are recalculated using the K-means clustering algorithm, while the dataset is balanced and expanded to address the problem of an uneven number of target classes in the TT100K dataset. The algorithm is compared to YOLOv3 and other commonly used target detection algorithms, and the results show that the improved YOLOv3 algorithm achieves a mean average precision (mAP) of 77.3%, which is 8.4% higher than YOLOv3, especially in small target detection, where the mAP is improved by 10.5%, greatly improving the accuracy of the detection network while keeping the real-time performance as high as possible. The detection network’s accuracy is substantially enhanced while keeping the network’s real-time performance as high as possible.

ETSR-YOLO: An improved multi-scale traffic sign detection algorithm based on YOLOv5

“…Fan et al [ 21 ] enhanced detection speed by adopting DenseNet as the backbone network for YOLOv3. Gong et al [ 22 ] modified YOLOv3’s network header to create 152 × 152 feature maps, improving the detection performance of small traffic signs. Song et al [ 23 ] proposed a Chinese traffic sign detection algorithm that enhances detection accuracy by optimizing the anchor boxes and SPP network of YOLOv4.…”

TSDet: A new method for traffic sign detection based on YOLOv5‐SwinT

“…A typical example is the traffic sign detection methods based on the YOLO framework [28][29][30]. For instance, Zhang et al [31] proposed a traffic sign detection network called ReYOLO.…”

“…A typical example is the traffic sign detection methods based on the YOLO framework [28–30]. For instance, Zhang et al.…”

Sensor Fusion and Advanced Controller for Connected and Automated Vehicles

“…The Special Issue of Sensors aims at reporting on some of the recent innovative studies on the perception, decision-making, planning, and control layers of autonomous vehicles and vehicle platoons with the help of advanced sensor techniques. In the perception layer and state estimation layer, a traffic sign recognition method [ 1 ] was proposed, and a rail micro-crack detection method [ 2 ] was designed to improve the perception of small traffic targets and rail flaw detection. In addition, an efficient and high-precision estimation framework for a Four-Wheel Independently Actuated (FWIA) autonomous vehicle was focused on [ 3 ], and an efficient measurement method for brake pressure change rate was reported [ 4 , 5 ].…”

“…Aiming at the challenges of small traffic signs, inconspicuous features, and low detection accuracy, a traffic sign recognition method based on the improved You Only Look Once v3 YOLOv3 was proposed [ 1 ]. Integrating the spatial pyramid pooling structure into the YOLOv3 network structure can realize the fusion of local features and global features, and a fourth feature prediction scale of 152 × 152 is applied to make full use of the shallow features in the network to predict small targets.…”