基于改进SSD的高空绝缘子缺陷检测算法

High-altitude insulator defect detection algorithm based on improved SSD

俞俊 ¹武丽 ²付相为 ¹张征浩 ¹葛彩成¹

扫码查看

作者信息

1. 南京信息工程大学电子与信息工程学院,南京 210000
2. 南京信息工程大学电子与信息工程学院,南京 210000;无锡学院电子信息工程学院,江苏无锡 214000
折叠

摘要

针对高空绝缘子背景复杂、缺陷目标较小等因素造成小目标检测效果不佳,提出了一种改进的基于残差融合注意力模块的SSD(Residual Fusion Attention Module,RFAM-SSD)小目标检测算法.首先将网络一分为二,主干网络通过残差网络的多重特征循环融合模块(ResNet Multiple Cycle Fea-ture Fusion Module,RES-MFCFM)得到特征层,经过卷积得到六个预测特征层,分支网络通过卷积得到对应的六个特征层,两分支通过RFAM,得到最终的六个特征层来检测目标;设计Focal-IOU Loss来代替原损失,提高检测效果.实验表明改进后的算法mAP为92.4％,比原始SSD算法提高了 7.2％,且满足实时检测需求,表明该检测算法对于绝缘子缺陷小目标有较好的检测效果.

Abstract

According to the poor detection effect caused by the complex background of high-altitude insula-tors and the small defect target,an improved small-target detection algorithm based on Residual Fusion At-tention Module(RFAM-SSD)is proposed.Firstly,the network is divided into two parts.The backbone network obtains the feature layer through the Multiple Cycle Feature Fusion Module(ResNet Multiple Cycle Feature Fusion Module,RES-MFCFM)of the residual network,and six predictive feature layers are ob-tained through convolution.The branch network obtains the corresponding six feature layers through convo-lution,while the two branches obtain the final six feature layers through RFAM to detect the target.Focal-1OU Loss is designed to replace the original loss and improve the detection effect.Experiments show that the mAP of the improved algorithm is 92.4％,which is 7.2％higher than that of the original SSD algorithm,and meets the real-time detection requirements,indicating that the detection algorithm has a good detection effect on small insulator defect targets.

关键词

小目标检测/SSD算法/残差网络/注意力机制/特征融合

Key words

small target detection/SSD algorithm/residual network/attention mechanism/feature fusion

引用本文复制引用

出版年

2024

信息技术

黑龙江省信息技术学会中国电子信息产业发展研究院　中国信息产业部电子信息中心

信息技术

CSTPCD

影响因子：0.413

ISSN：1009-2552

段落导航