基于改进YOLOv8的电力作业人员安全带检测

摘要:

正确穿戴安全带是预防电力作业人员高空坠落的重要措施。针对在电力现场中作业人员是否穿戴安全带检测效率低以及实效性差的问题，提出了一种基于改进YOLOv8的电力作业人员安全带检测方法。该算法在网络中添加SE注意力机制，提升了模型的识别能力；引入了加权双向特征金字塔网络结构进行特征融合，提高特征学习能力，并降低模型复杂度；采用WIoUv3损失函数替换了原先的CIoU损失函数，进一步提高了模型的检测精度以及对于不同环境的适应能力。实验结果表明，该算法的mAP平均精度均值达到96.5%，识别效果明显提升，并且优于其他经典目标检测模型，验证了新算法的有效性。

Abstract:

Wearing seat belts correctly is an important measure to prevent electrical workers from falling from heights. In order to solve the problem of low efficiency and poor effectiveness of whether workers wear seat belts in power sites, a safety belt detection method for power workers based on improved YOLOv8 was proposed. The algorithm adds SE attention mechanism to the network to improve the recognition ability of the model. At the same time, a weighted bidirectional feature pyramid network structure is introduced to perform feature fusion, improve feature learning ability, and reduce the complexity of the model. The WIoUv3 loss function is used to replace the original CIoU loss function, which further improves the detection accuracy and adaptability of the model to different environments. Experimental results show that the average accuracy of the proposed algorithm reaches 96.5%, and the recognition effect is significantly improved, which is better than other classical object detection models, which verifies the effectiveness of the new algorithm.