本文提出了一种基于Faster R-CNN的交通标志检测方法。首先通过

In this paper, traffic sign detection method based Faster R-CNN is. Firstly, the image data set by collecting traffic sign on both sides of the road, and then use the corresponding pre-trained VGG16 Faster R-CNN parameter model is initialized, last Faster R-CNN model training. Tests showed that it mAP up 85.67%, to meet the actual needs of accuracy.

In this paper, a traffic sign detection method based on Faster R-CNN is proposed. The data set is first established by collecting images of traffic signs on both sides of the highway, then the corresponding parameters of the Faster R-CNN model are initialized using pre-trained VGG16, and finally the Faster R-CNN model is trained. The experiment shows that the mAP can reach 85.67 percent, which can meet the actual accuracy demand.

This paper presents a traffic sign detection method based on Faster R-CNN. Firstly, the data set is established by collecting the images of traffic signs on both sides of the highway. Then the corresponding parameters of Faster R-CNN model are initialized using pre-trained VGG16. Finally, the Faster R-CNN model is trained. Experiments show that its mAP can reach 85.67%, which can meet the actual accuracy requirements.<br>