融合多尺度特征与位置信息的输电线路山火检测算法  

Mountain Fire Detection Algorithm of Transmission Line Based on Multi-scale Features and Coordinate Information

作　　者：陈冬 周浩[1] 袁国武[1] 杨凌宇 成秋艳 任莹 马仪 CHEN Dong;ZHOU Hao;YUAN Guowu;YANG Lingyu;CHENG Qiuyan;REN Ying;MA Yi(School of Information Science and Engineering,Yunnan University,Kunming 650504,China;Power Science Research Institute of Yunnan Power Grid LTD,Kunming 650217,China)

机构地区：[1]云南大学信息学院,昆明650504 [2]云南电网有限责任公司电力科学研究院,昆明650217

出　　处：《计算机科学》2024年第S02期248-254,共7页Computer Science

基　　金：云南省科技重大专项(202202AD080004)。

摘　　要：针对输电线路山火中烟雾火灾目标存在尺度多变、背景复杂的特点,会导致检测精度低和误检的问题,文中在YOLOv5的基础上提出了一种融合多尺度特征与位置信息的输电线路山火检测算法。首先,为解决尺度多变的问题,针对SPPF(快速空间金字塔池化层)级联式结构在尺度融合时只关注局部特征信息的问题,提出了一种层级式与CSP(跨阶段局部网络)结构相结合的尺度融合模块SPPCSPC(跨阶段空间金字塔池化层),能有效提取并融合多层次、不同尺度的全局特征信息,提高算法对烟雾火灾目标的检测能力。其次,为解决误检问题,提出了新的颈部网络PCANet(路径位置聚合网络)既融合浅层和深层特征图,又分别从特征图的竖直和水平两个方向将目标的位置信息融入通道中,增强算法对烟雾火灾目标位置信息的关注度,降低复杂背景的干扰。在输电线路烟雾火灾数据集上进行实验,所提算法的mAP_(50)提高了1.6%,mAP_(50:95)提高了1.5%,Recall提高了2.4%,有效提高了烟雾火灾目标检测精度,减少了误检现象发生,能够更好地完成检测任务。Due to variable scale and complex background of smoke fire targets in transmission line mountain fires,it may lead to low detection accuracy and false alarms.To address these issues,this paper proposes an improved YOLOv5 object detection algorithm.To tackle the problem of scale variability,aiming at the problem that SPPF cascade structure only focuses on local feature information during scale fusion,this paper proposes spatial pyramid pooling cross stage partial conva(SPPCSPC),which combines hierarchical and cross stage partial networks(CSP)structures,effectively extracting and fusing multi-level and different scale global feature information to enhance the model’s ability to detect smoke and fire targets.Secondly,to solve the misdetection problem,this paper proposes a new neck network PCANet(path coordinate aggregation network).While integrating the backbone network and neck feature maps,it integrates the target’s position information from the vertical and horizontal directions of the feature maps into the channels respectively.This enhances the model’s attention to smoke fire target position information and reduces interference from complex backgrounds.Experiments are conducted on a transmission line smoke fire dataset to evaluate the proposed model.The proposed algorithm achieves an increase of 1.6% in mAP_(50),1.5% in mAP_(50:95),and 2.4% in recall,respectively,which effectively improve the detection accuracy of smoke and fire target,reduce the occurrence of false detection,and can better complete the detection task.

关 键 词：输电线路 烟雾火灾检测 多尺度特征融合 位置信息 YOLOv5 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]