摘要
青藏高原铁路沿线太阳辐射强、气温低、空气干燥稀薄、大风沙尘等独特的环境条件会影响接触网装置图像的采集、绝缘子类型的选择等,这些情况都会给实际情况中绝缘子的定位产生很大的影响,文章针对青藏高原铁路沿线特殊地理气候环境下复杂背景的接触网装置巡检图像提出了一种基于几何矩轮廓检测和多边形拟合的绝缘子定位算法。实验验证表明,算法很好地克服了训练样本库方法、特征匹配方法不能应用于实际的青藏铁路沿线接触网支持装置图像绝缘子定位的问题,为以后接触网支持装置图像绝缘子的检测、预警等其他进一步的处理作好了准备。
The unique environmental conditions of Qinghai Tibet Plateau Railway along the solar radiation intensity, low temperature, air drying thin, wind, sand and dust will affect the contact net device image acquisition, insulator type selection, these will give the positioning of the insulator in the actual situation to have the very big influence. In this paper for along the Qinghai Tibet railway, special geographical and climatic environment under complex background of contact net device inspection image presents a kind of insulator location algorithm based on geometric moments contour detection and polygonal fitting. Experiments show that the algorithm well overcomes the library method of training samples, feature matching methods can not be applied to actual along the Qinghai Tibet railway contact net support device insulator image positioning, for contact net support device insulator image detection, warning other further processing ready.
出处
《无线互联科技》
2016年第8期33-36,共4页
Wireless Internet Technology
关键词
绝缘子定位
HU矩
多边形拟合
图像形态学变换
insulator location
Hu moment
polygon fitting
image morphological transformation
