基于图像处理的变电站指针式仪表读数研究

为解决变电站人工巡检指针式仪表工作强度大和读数误差高的问题,提出了一种基于图像处理技术的变电站指针式仪表自动检测系统。针对读入的图片实现图像灰度化、图像二值化;采用改进的Hough变换提取图像中的直线、计算指针角度和显示指针读数;并利用OpenCV函数库建立了指针式仪表图像的模型库,导入新的图片可选择不同的模型库。实验结果表明,该指针式仪表自动判读系统可以准确识别出指针的读数,且读数精度远大于人工读数。

Substation Pointer Instrument Reading Based on Image Processing

In order to solve the working strength and high reading error of substation manual checking pointer instrument,this paper puts forward an automatic reading system for the substation pointer instrument based on image processing.Firstly image graying and binarization processing of the reading pictures is performed.Secondly the improved Hough transform is used to extract the straight lines of the image for the pointer calculation and the display of pointer readings.And the image model libraries of the pointer instrument are established based on the OpenCV function library for the system to choose a different model library based on the new image.The experimental results show that the pointer instrument automatic interpretation system can accurately identify the pointer readings with better accuracy than artificial reading.