基于特征匹配的变电站巡检机器人设备识别  

Identification of Substation Inspection Robot Equipment Based on Feature Matching

作　　者：张云翔 李智诚 ZHANG Yunxiang;LI Zhicheng(Shenzhen Power Supply Bureau Co.Ltd.,Shenzhen 518022,China)

机构地区：[1]深圳供电局有限公司,广东深圳518022

出　　处：《工业加热》2022年第12期62-65,70,共5页Industrial Heating

基　　金：广东省科技攻关项目(20205899R01)。

摘　　要：目前变电站巡检机器人设备识别技术存在设备识别准确率低、设备识别效果差和指针定位精度低的问题,提出基于特征匹配的变电站巡检机器人设备识别方法。对图像进行定位、去噪、图像分割、二值化以及细化的预处理,提高识别精度,降低外界干扰;在Hough变换的基础上计算设备指针的详细斜率,完成指针定位,以此生成设备读数,与相应的阈值进行比较,在特征匹配下完成设备识别和录入,实现变电站巡检机器人设备识别。实验结果表明,所提方法的仪表设备识别准确率高、仪表设备识别效果好,且仪表指针的定位精度高。At present,the recognition technology of substation inspection robot equipment has the problems of low equipment recognition accu⁃racy,poor equipment recognition effect and low pointer positioning accuracy.A method for substation inspection robot equipment recognition based on feature matching is proposed.Perform preprocessing of image positioning,denoising,image segmentation,binarization and refine⁃ment to improve recognition accuracy and reduce external interference;calculate the detailed slope of the device pointer based on the Hough transform to complete pointer positioning to generate the device The readings are compared with the corresponding thresholds,and equipment i⁃dentification and input are completed under feature matching to realize the identification of substation inspection robot equipment.The experi⁃mental results show that the proposed method has a high recognition accuracy of the instrument and equipment,a good recognition effect of the instrument and equipment,and the positioning accuracy of the instrument pointer is high.

关 键 词：特征匹配 变电站 预处理 设备读数 设备识别 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]