基于无人机的输电网故障跳线联板识别

发布时间：2018-04-29 14:40

本文选题：红外图像 + 跳线联板　；参考：《液晶与显示》2016年12期

【摘要】：跳线联板是输电网中重要设备,其是否存在故障对输电网正常运行具有很大的影响。但由于现有的算法是对输电网中所有的故障用统一的方法进行识别,没有对各类故障输电设备进行专门的研究,导致故障跳线联板识别率低。为了高效识别红外视频图像中故障跳线联板,首先针对输电线的红外图像特征,采用改进的OTSU阈值分割图像对红外图像进行分割;其次,采用漫水法滤波分离各个连通域,运用形态学滤去小区域,填充大区域内的孔洞;最后,提取连通域的骨架,并从骨架图像中提取出USFPF特征,通过该特征识别的故障跳线联板。实验结果表明,识别故障跳线联板准确率为85.71%,漏检率为14.28%,误识别率为2.8%。该方法能够较好地识别故障跳线联板,具有较好的鲁棒性。
[Abstract]:Jumper connection board is an important equipment in transmission network. Whether its fault exists or not has a great influence on the normal operation of transmission network. However, because the existing algorithm is to identify all the faults in transmission network by the unified method, there is no special research on all kinds of fault transmission equipment, which leads to the low recognition rate of the fault jumper board. In order to efficiently identify the fault jumper coupling in infrared video image, firstly, the improved OTSU threshold is used to segment the infrared image according to the infrared image feature of transmission line; secondly, the diffuse water filter is used to separate each connected region. The small area is filtered out by morphology and the holes in the large area are filled. Finally, the skeleton of the connected domain is extracted, and the USFPF feature is extracted from the skeleton image, and the fault jumper coupling is identified by the feature. The experimental results show that the accuracy of fault jumper coupling is 85.71, the missing detection rate is 14.28 and the error recognition rate is 2.8. This method can identify the fault jumper coupling well and has good robustness.
【作者单位】：北京理工大学光电学院光电成像系统与技术教育部重点实验室;天津航天中为数据系统科技有限公司天津市智能遥感信息处理技术企业重点实验室;南方电网科学研究院有限责任公司;
【基金】：南方电网直升机重大专项资助项目(No.K-KY2014-500)~~
【分类号】：TM75;TP391.41

【相似文献】