For the issues such as license plate positioning difficulties,low acc uracy,and slow detection speed of the target detection algorithm of current traffic monitoring system in the forest environment in fog,rain,snow and other adverse weather conditions,this paper proposed a new license plate detection method,which used you only look once v5 (YOLOv5) as the base model.Firstly,this paper experimented wi th the K-means++ method and made cluster analysis of the label information of the instances to ob tain the new anchor frame sizes.Next,the convolutional block attention module (CBAM) attention mechanism was incorporated into the featur e extraction network to extract more feature information of the detection target.Finally,CIoU was c hosen as a loss function to improve the detection frame localisation accuracy.In terms of the pre-proce ssing,the possible interference generated by the camera during image acquisition was simulated and the images were scripted using OpenCV-Python to increase the robustness of the algorithm for de tection in complex environments in forest areas.As shown by the experimental analysis results,the mean average precision@0.5(mAP@0.5) and the mean average precision@0.5∶0.95(mAP@0.5∶0.95) of the improved model reached 99.5% and 86.7%,respectively.In addition,the detection speed reached 128 fps,but the model size was only 14 M.Compared with YOLOv5 and o ther leading target detection algorithms,this method had improved accuracy,real-time performance and broad deployability.