摘要
Abstract
The detection of foreign objects on transmission lines is crucial for ensuring the safe operation of the power system.In order to improve the efficiency of foreign object recognition in transmission lines,the YOLOv3 Tiny model is improved.Firstly,in the head network,depthwise separable convolution is used instead of standard convolution,normalization,and activation function structures to separate spatial and channel correlations,reduce convolution computation,and improve recognition speed;secondly,the EIoU loss function considering distance loss and height and width loss is introduced to replace the original loss function,enabling the model to find the optimal point between bounding box prediction and category prediction,thereby improving the detection performance of the algorithm.The ablation experiment verifies the effectiveness of these improvements,and the results show that the improved model maintains high accuracy while increasing the detection rate(FPS)by 2.02 times,reducing the number of parameters by 74.17％,and significantly reducing the computational resource requirements.The algorithm performs well in resource constrained environments and has practical application value.关键词
输电线异物检测/目标检测算法/YOLOv3-Tiny/损失函数Key words
foreign object detection on transmission lines/object detection algorithm/YOLOv3-Tiny/loss function分类
信息技术与安全科学
