基于改进像素聚类算法的输电线路冰区识别方法

发布时间：2018-03-23 07:13

本文选题：像素　切入点：聚类　出处：《现代电子技术》2017年20期 　论文类型：期刊论文

【摘要】：为了提高输电线路的防灾减灾能力,特别是防治冰雪灾害的能力,需要进行输电线路冰区自动识别方法和系统设计,提出基于改进像素聚类算法的输电线路冰区识别方法。对气象卫星遥感监测采集的输电线路覆盖区域的遥感图像进行图像处理,提取图像中的冰覆盖像素特征点,结合边缘轮廓检测方法进行输电线路覆盖区域的冰区边缘轮廓检测,对冰区覆盖区域的像素特征进行模糊C均值聚类处理,实现区域分割识别。仿真结果表明,采用该方法进行输电线路冰区识别的准确率较高,视觉直观效果较好,便于指导施工抢修人员进行现场排除和巡线检修。
[Abstract]:In order to improve the ability of disaster prevention and mitigation of transmission lines, especially the ability of preventing and controlling ice and snow disasters, it is necessary to carry out automatic identification methods and system design for ice areas of transmission lines. An improved pixel clustering algorithm is proposed to identify the ice area of transmission lines. The remote sensing images of transmission lines covered by meteorological satellite are processed and the ice-covered pixel feature points are extracted from the images. Combined with the edge contour detection method, the edge contour of the ice area covered by the transmission line is detected, the pixel feature of the covered area is processed by fuzzy C-means clustering, and the region segmentation and recognition is realized. The simulation results show that, Using this method, the accuracy of ice area identification of transmission lines is high, the visual effect is good, and it is convenient to guide the construction rush repair personnel to carry out the spot elimination and the inspection and repair of the lines.
【作者单位】：贵州电网有限责任公司输电运行检修分公司设备部;贵州电力设计研究院;
【基金】：国家公益性行业科研专项经费项目(GYHY201006033)
【分类号】：TM75;TP751

【相似文献】