提出了一种基于改进的You Only Look Once version 5(YOLOv5)模型的高压线路障碍物检测方法,旨在提高检测的准确性和效率.首先,改进了原始的YOLOv5模型,引入了注意力机制以增强对障碍物特征的识别能力.其次,针对高压线路障碍物的多样性和复杂性,构建了一种包含大规模数据集的系统,确保了模型训练的全面性.结果显示,与传统方法和未改进的YOLOv5模型相比,所提方法在高压线路障碍物检测任务上表现出更高的准确率和更快的处理速度.
Research on High-Voltage Line Obstacle Detection Algorithm Based on Improved YOLO Algorithm
This paper proposes a high-voltage line obstacle detection method based on the improved You Only Look Once version 5(YOLOv5)model,aiming to improve the accuracy and efficiency of detection.Firstly,this paper improves the original YOLOv5 model and introduces an attention mechanism to enhance the ability to identify obstacle features.Secondly,in response to the diversity and complexity of high-voltage line obstacles,we built a system containing a large-scale data set to ensure the comprehensiveness of model training.The results show that compared with traditional methods and the unimproved YOLOv5 model,our proposed method exhibits higher accuracy and faster processing speed on high-voltage line obstacle detection tasks.
YOLOv5 modelattention mechanismhigh-voltage line detectionobstacle detection
万方数据
维普
