基于链码和纹理分析的输电线断股检测

Broken Strands Detection of Transmission Line Based on Chain Code and Texture Analysis

为了解决基于数字图像的输电线断股的检测问题,提出了基于Freeman链码和纹理分析法相结合的输电线断股检测方法,并根据断股输电线特点提出了链码断股识别准则。对待识别图片进行灰度化、滤波和边缘检测处理;运用Radon变换对输电线进行定位,获得输电线在图像中的位置;分别运用Freeman链码和纹理分析法对输电线的外围和内部进行判断,对断股位置和掉落电线进行检测,从而实现了输电线断股和形变的定位,并对断股缺陷的程度进行判别。实验结果表明,该方法可以准确定位输电线断股位置。

In order to solve the problem of the detection of the broken strands of transmission line based on digital image, a method of broken strands detection based on Freeman chain code and texture analysis is proposed. The image is processed by gray scale, filtering and edge detection. Then, the position of transmission line is obtained by Radon transformation. Freeman chain code and texture analysis method are used to determine the external and internal of the transmission line to locate the position of the fault and the wire is dropped out, so the position of the fault and deformation of the transmission line is realized, and the degree of the fault is distinguished. Experi-mental results show that the method can accurately locate the position of the transmission line.