计算机与现代化 ›› 2025, Vol. 0 ›› Issue (09): 67-72.doi: 10.3969/j.issn.1006-2475.2025.09.010

• 图像处理 • 上一篇    下一篇

基于复杂环境的绝缘子缺陷检测

  


  1. (1.南京工程学院电力工程学院,江苏 南京 211100; 2.南京工程学院计算机工程学院,江苏 南京211100)
  • 出版日期:2025-09-24 发布日期:2025-09-24
  • 作者简介: 作者简介:季星宇(2001—),男,江苏苏州人,硕士研究生,研究方向:智能电网的图像识别,故障检测,E-mail: 578403376@qq.com; 黄陈蓉(1963—),女,江苏南京人,教授,研究方向:图像边缘检测,图像分割,E-mail: huangcr@njit.edu.cn; 姚军财(1979—),男,湖北黄冈人,教授,硕士生导师,博士,研究方向:图像和视频处理技术,目标跟踪和检测,计算机视觉及其交叉学科,E-mail: yaojcnj@njit.edu.cn; 王凯(1999—),男,江苏无锡人,硕士研究生,研究方向:故障检测,E-mail: 791424012@qq.com; 顾铭杰(1999—),男,江苏盐城人,硕士研究生,研究方向:电力目标检测,E-mail: 1271569459@qq.com。
  • 基金资助:


        基金项目:国家自然科学基金资助项目(61301237); 江苏省自然科学基金面上项目(BK20201468)

Insulator Defect Detection Based on Complex Environment


  1. (1. School of Electric Power Engineering, Nanjing University of Engineering, Nanjing 211100, China; 
    2. School of Computer Engineering, Nanjing University of Engineering, Nanjing 211100, China)
  • Online:2025-09-24 Published:2025-09-24

摘要: 摘要:如今无人机已经广泛应用于电力巡检中,但由于绝缘子缺陷图像背景复杂,缺陷尺寸较小,并会出现闪络、自爆、破损等多情况损坏,大大限制了检测速度和精度,针对以上问题,提出一种基于改进YOLOv5的复杂环境的绝缘子缺陷检测方法。首先,采用改进后的特征提取网络C2FNet,在保证轻量化的同时获得更加丰富的特征信息。其次,采用具有多尺度信息的Res2Net模块,改善梯度传播和训练效果。最后,设计具有自适应融合的动态目标检测头3-DyHead,动态调整网络结构和参数。实验结果表明,本文方法平均精确度达到了94.2%,与原模型相比平均精确度提升了4.1百分点,查准率P和召回率R分别提升了3.2百分点和4.0百分点,绝缘子闪络、防震锤、破损的平均精确度分别提升了11.0百分点、2.0百分点和6.5百分点。


关键词: 关键词:绝缘子检测, C2FNet, Res2Net, 自适应融合, 3-DyHead

Abstract:
Abstract: Nowadays, drones have been widely used in power inspection. However, due to the complex background of insulator defect images, small defect sizes, and the occurrence of multiple damage situations such as flashover, self explosion, and breakage, the detection speed and accuracy are greatly limited. To address these issues, a complex environment insulator defect detection method based on improved YOLOv5 is proposed. Firstly, an improved feature extraction network C2FNet is adopted to obtain richer feature information while ensuring lightweight. Secondly, the Res2Net module with multi-scale information is adopted to improve gradient propagation and training performance. Finally, a dynamic object detection head 3-DyHead with adaptive fusion is designed to dynamically adjust the network structure and parameters. The experimental results show that the average accuracy of this method has reached 94.2%, which is 4.1 percentage points higher than the original model. The precision P and recall R have increased by 3.2 percentage points and 4.0 percentage points, respectively. The average accuracy of insulator flashover, hammer, and defect has increased by 11.0 percentage points, 2.0 percentage points and 6.5 percentage points.

Key words: Key words: insulator testing, C2FNet, Res2Net, adaptive fusion, 3-DyHead

中图分类号: