双视电力巡检机器人夜视图像的小波降噪研究  

Research on Wavelet Threshold Denoising of Night Vision Image of a Double View Inspection Robot

作　　者：王林 陈瑶 马浩 张云婷 WANG Lin;CHEN Yao;MA Hao;ZHANG Yun-ting(State Grid(Jiaxing)Comprehensive Energy Services Co.Ltd.,Jiaxing 314000;Jiaxing Hengchuang Power Equipment Co.Ltd.,Jiaxing 314000,China;Jiaxing Hengchuang Power Group Co.Ltd.,Jiaxing 314000,China)

机构地区：[1]国网(嘉兴)综合能源服务有限公司,浙江嘉兴314000 [2]浙江嘉恒科创有限公司,浙江嘉兴314000 [3]嘉兴恒创电力集团有限公司,浙江嘉兴314000

出　　处：《电气开关》2024年第2期86-88,共3页Electric Switchgear

摘　　要：选取电力系统中应用广泛的双视巡检机器人,分析双目视觉机器人夜视巡检图像,针对视觉系统采集的电力设备图像进行中值滤波、特征提取、聚类分析。通过构造模糊阈值函数,提出小波阈值夜视图像降噪改进方法。通过数据分析,夜视图像降噪效果明显,为利用图像分析设备状态提供支撑。In this paper,the binocular inspection robot widely used in power system is selected,and the night inspection images of binocular vision robot are analyzed.The images of power equipment collected by vision system are subjected to median filtering,feature extraction and cluster analysis.By constructing fuzzy threshold function,an improved wavelet threshold night vision image denoising method is proposed.Through data analysis,the noise reduction effect of night vision image is obvious,which provides support for using image analysis equipment status.

关 键 词：双目视觉 夜视图像 模糊函数 小波降噪 

分 类 号：TM71[电气工程—电力系统及自动化]