Identification of Construction Vehicles under High Voltage Transmission Line Based on Improved YOLOv5s

To achieve accurate recognition of construction vehicles under high-voltage transmission lines and eliminate the influence of external environmental influences on the recognition effect, it is essential to use effective target detection methods to achieve real-time detection of construction vehicles under transmission line scenes. Some current methods have good results in target detection accuracy, but they do not meet the lightweight requirements and cannot respond in time. To address the existing problems, based on YOLOv5s, we propose a new target detection method. The method not only improves detection accuracy but also has a small model computation, which facilitates lightweight deployment and response. A guided filtering algorithm is first used to noise-reduce the images in the dataset and enhance the texture features of the images. Then based on YOLOv5s, the receptive field module (RFB) is embedded after the spatial pyramid pooling module (SPPF) to enhance the feature extraction capability of the network. The experimental results show that compared with the traditional YOLOv5s algorithm, the improved algorithm improves the detection accuracy by 4.6%, and the recognition effect is significantly improved, which verifies the effectiveness of the proposed new algorithm.