基于优化Gabor滤波器的输电导线断股图像检测

Image Detection for Broken Strand Faults of Transmission Conductor Based on Optimized Gabor Filter

输电导线在运行中可能发生损伤、断股现象,对其进行及时检测和诊断对于保证电网的安全运行具有重要意义。现代输电线路状态检修一般采用直升机或者巡线机器人搭载视频检测装置的方式。文中提出应用Gabor滤波器对输电导线视频图像数据进行处理,实现输电导线断股的图像检测,即通过计算完好导线图像与滤波器卷积能量的输出响应、应用小生境遗传算法寻找最优的Gabor滤波器参数和分割阈值,最后通过检测图像的滤波和能量的二值化处理得到检测结果。实验结果表明,该算法能够较好地提取输电导线断股处的故障信息,具有适用性广、识别能力强、检测速度快等优点。

Overhead transmission lines generally operate in the wild harsh environment and can be damaged by different kinds of reasons,so that detecting broken strands timely is an important mean to insure safety operation of transmission lines.The modern condition-based maintenance of transmission line uses the way of helicopter or inspection robot with visual detectors.An image detection method with Gabor filter for broken strands of transmission conductors is proposed.In the proposed scheme,output response of filter convolution energy and the normal transmission conductor images are calculated firstly,then the optimal parameters of Gabor filter and image segmentation threshold value are calculated with niche genetic algorithm（NGA）.And finally the detection results are obtained through filtering and energy binary processing of test image.Experimental results show that the proposed algorithm is able to efficiently detect the broken strand faults of transmission conductor accurately and rapidly.