Road traffic sign detection and recognition algorithm based on improved YOLOv5

According to the road traffic sign detection and recognition algorithm based on the improved YOLOv5, an ECA (Efficiency Channel Attention) attention mechanism is introduced on the basis of a YOLOv5 network model, the ECA attention mechanism is fused into a C3 module to form a C3-ECA module, the performance of a deep convolutional network is improved, the complexity of the model is reduced, PANet is replaced with BiFPN (Bidirectional Feature Patterned Network), and therefore, the detection and recognition efficiency of the road traffic sign is improved. Bidirectional fusion of deep and shallow layer features from top to bottom and from bottom to top is realized, and the detection performance of a network model is remarkably improved. 本发明提出一种基于改进YOLOv5的道路交通标志检测与识别算法，在YOLOv5网络模型的基础上引入了ECA（Efficient Channel Attention）注意力机制，将其融入C3模块之中组成C3-ECA模块，提高了深度卷积网络的性能，降低了模型复杂度，并且将PANet（Path Aggregation Network）替换为BiFPN（Bidirectional Feature Pyramid Network），实现了自上而下与自下而上的深层与浅层特征的双向融合，显著提升了网络模型的检测性能。