摘要
自动化、智能化是变电站发展的趋势,越来越多的自动巡检机器人被应用于巡检工作中。红外图像检漏技术是常用的电气设备SF_(6)气体检漏手段,既可以针对局部查找漏点,又可以快速大范围巡检。因此,对红外检漏自动定位算法的研究有其必要性。本文研究一种SF_(6)红外检漏视频定位算法,它既可用于变电站自动巡检机器人的红外检漏自动定位,亦可用于手持式红外图像检漏仪辅助定位。首先将红外图像灰度化,并利用高斯模糊技术消除噪声干扰;其次通过阈值变换在首帧图像中识别设备区域,得到掩膜;然后基于漏点位置相对固定且持续的特性,使用累积帧间差分法获取红外图像的累积变化;最后根据掩膜排除不在设备区域的干扰,对SF_(6)泄漏点进行定位和统计。实验表明,该算法能准确定位SF_(6)漏点位置,运算速度快且抗干扰性好。
Automation and intelligence are the development trend of substation.More and more automatic inspection robots are used in inspection work.Infrared image leakage detection technique is widely used in detecting SF_(6)gas leakage in electrical equipment.It can be used in local leakage searching and large-scale patrolling.Therefore,it is necessary to study the automatic location algorithm of infrared leakage detection.First,gray processing is applied to infrared image and Gauss blur is used to eliminate the noise.Threshold transformation is used to recognize the device area in the first frame image and obtain mask.Then,accumulated interframe difference method is used to get the cumulative changes in infrared images.Finally,the interferences outside the device area are removed,and the leakage points are located.The experiment shows that the proposed method has fast computing speed and good anti-interference ability,which makes it suitable for real-time video processing.This method can be used in both automatic substation patrol robot and handheld infrared image leakage detector.
作者
黄柏
杨帆
邓剑平
毕凡
HUANG Bo;YANG Fan;DENG Jianping;BI Fan(Electric Power Test and Research Institute,Guangzhou Power Supply Bureau Co.,Ltd,Guangzhou 510410)
出处
《电气技术》
2022年第7期104-108,共5页
Electrical Engineering
