基于YOLOv5的交通标志检测

Traffic sign detection based on YOLOv5

交通标志检测是自动驾驶技术的重要组成部分,但是在实际应用中面对复杂的真实场景,交通标识检测容易受到诸多复杂因素的影响,如背景环境、天气因素等,使得在检测过程中误检和漏检的概率增加,而传统目标检测算法很难满足对检测精度和速度的要求.为此,本文使用真实道路场景TT100k数据集作为研究对象,利用YOLOv5s算法对模型进行迭代训练,结果显示mAP可以达到82.3%,对于数据集中禁令标志的检测其准确度最高可以达到91.4%.实验结果表明,YOLOv5算法对于交通标志的检测具有较高的识别精度及良好的检测效果.

Traffic sign detection is an important part of automatic driving technology.However,in practical applications,in the face of complex real scenes,traffic sign detection is easily affected by many complex factors,such as background environment and weather factors,which increases the probability of false detection and missed detection in the detection process.However,the traditional target detection algorithm is difficult to meet the requirements of detection accuracy and speed In order to solve these problems,this paper uses the TT100k data set of real road scenes in China as the research object,and uses the Yolov5s algorithm to train the model iteratively.The results show that the mAP can reach 82.3%,and the highest accuracy can reach 91.4% for the detection of prohibition signs in the data set.The experimental results show that Yolov5 algorithm has high recognition accuracy and good detection effect for traffic signs detection.