变压器输变电设备边缘检测的改进算法  

Improved algorithm for edge detection of transformer transmission and transformation equipment

作　　者：葛遵辉 李凌烜 Ge Zunhui;Li Lingxuan(School of Electrical&Control Engineering,Heilongjiang University of Science&Technology,Harbin Heilongjiang,150022)

机构地区：[1]黑龙江科技大学电气与控制工程学院,黑龙江哈尔滨150022

出　　处：《电子测试》2021年第8期55-56,15,共3页Electronic Test

摘　　要：针对传统的Prewitt算子在输变电设备图像边缘检测中,存在检测边缘较粗、定位不准确和人为选取阈值,从而会造成边缘点误判等缺点,不能正确反映边缘检测质量的实际优劣情况,本文提出一种基于改进的Prewitt算子上,用高斯滤波进行图像预处理,运用OUTS大津获取动态阈值法确定最优阈值来减少边缘点的误判。基于Matlab2016a应用此改进方案对变压器输变电设备进行图像边缘检测并且与其他传统算法进行对比。通过对结果的分析表明,该算法能有效地抑制噪声,解决边缘选取粗略的缺点,有效提高边缘检测准确性。In view of the traditional Prewitt operator in the image edge detection of power transmission equipment,there are some disadvantages such as thick detection edges,inaccurate positioning and artificial threshold selection,which will cause misjudgment of edge points,and cannot accurately reflect the actual quality of edge detection.In this case,this paper proposes an improved Prewitt operator,using Gaussian filtering for image preprocessing,using OUTS Otsu to obtain dynamic threshold method to determine the optimal threshold to reduce the false judgment of edge points.Based on Matlab2016a,this improved scheme is applied to image edge detection of transformer power transmission and transformation equipment and compared with other traditional algorithms.The analysis of the results shows that the algorithm can effectively suppress noise,solve the shortcomings of rough edge selection,and effectively improve the accuracy of edge detection.

关 键 词：变压器输变电设备 边缘检测 PREWITT算子 MATLAB 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TM41[自动化与计算机技术—计算机科学与技术]