摘要
绝缘子是输电线路中的重要组成部分,它是否正常运行影响到电力系统的安全。文章提出了利用机器视觉的方法提取绝缘子图像的特征值,为绝缘子检测和识别提供信息。首先利用像素统计法对实际绝缘子图像进行分割,然后利用灰度共生矩阵提取图像的纹理特征,利用二值图像提取不变矩特征。最后,利用图像局部性质,计算图像的几何特征值,提取边界轮廓。
Because the insulator, which affects the safety of electric power system, is important component of transmission line, this paper proposes the computer vision method to extract the characteristic values of insulator images, which can provide information for insulator detection and recognition. The pixel statistical method is first used to segment the actual insulator, the Gray level co-occurrence matrix to extract the textural features of insulator images, and the binary image to extract the invariant moment features. And then, the local properties of the images is used to calculate the geometric features and to pick up the boundary contour.
出处
《电测与仪表》
北大核心
2013年第12期37-41,共5页
Electrical Measurement & Instrumentation
关键词
绝缘子
特征提取
阈值分割
纹理
不变矩
insulator
feature extraction
threshold segment
texture
invariant moment
