Aiming at the shortcomings of traffic sign detection algorithm, such as high network complexity, large amount of calculation and difficult to be applied at the edge. A lightweight traffic sign target detection algorithm based on YOLOv5 is proposed. By increasing the attention mechanism and using the fusion of CBAM and CA, the anti-interference ability of the detection model is strengthened; Through FPGM pruning, the model is compressed to reduce the amount of calculation and improve the reasoning speed; Through the integration design of software and hardware, YOLOv5s model and hardware are integrated to form a complete set of mobile intelligent traffic sign target detection system; The results show that the accuracy of the model is improved by 2.8% after adding multiple attention mechanisms. In the case of extreme pruning, the model is only 0.54MB. Under the environment of Jetson Nano (20W), the detection speed is up to 21 frames / s, which meets the real-time traffic sign detection.