摘要
以无人机搭载的红外传感器获取的视频为数据源,研究了基于红外图像的电力设备异常发热检测,实现了输电线故障缺陷位置的自动诊断和定位。首先从红外视频中抽取红外序列图像帧并对抽取的红外图像进行自动拼接和帧间差分,进而确定输电线的主方向,并对输电线区域进行定位;然后根据红外图像和输电线故障诊断标准对提取的输电线进行故障诊断,最终实现了输电线故障的定位和自动诊断。实践证明该方法具有较高的自动化程度和效率。
An automatic electrical equipment abnormal heat diagnosis method using infrared image sequence adopted from unmanned helicopter（UHV） is proposed in this paper, which automatically identifies the locations of device fault for further manual inspection. Firstly, sequent image frames are extracted from infrared video. Then, each image frame is stitched with its neighbors by conjugate features, and frame difference is performed on the stitched frames to eliminate the background. The principal direction of power line is calculated based on the non-background different frame. Finally, a model based template matching schedule is adopted to identify the power line area. Finally intelligent diagnosis and location of power line fault is realized by infrared image and related national standards. Experimental studies validate the effectivity and robustness of the proposed method in highly cluttered power line inspection scene.
出处
《电网技术》
EI
CSCD
北大核心
2014年第5期1334-1338,共5页
Power System Technology
基金
国家重点基础研究发展计划项目(97 3项目)(2012CB725301)
国家自然科学基金项目(41071268)
中国南方电网公司重点科技项目(K-GD2013-030)~~
关键词
无人机
红外图像处理
输电线提取
故障诊断
unmanned helicopter
infrared image processing
power line extraction
fault diagnosis
