摘要
针对接触网安全巡检系统中利用支柱号牌进行智能缺陷分析定位的问题,提出了一种支柱号牌识别算法。该算法采用霍夫直线检测实现支柱定位,根据号牌特点实现号牌定位、号牌字符分割算法,最后采用HOG特征和支持向量机(SVM)实现字符识别。该算法识别率较高,对智能缺陷定位具有重要意义。
With regard to issues on intellectual analysis and location of defects by pole number identification in safety patrol inspection for overhead contact system, a pole number identification and calculation method is put forward in this paper. In this calculation, the pole location is realized by Hoff straight line inspection, and the pole number plate location and segmentation calculation of the number plate are realized on the basis of characteristics of number plates, and identification of characters is realized by adopting of HOG features and supporting vector machine(SVM). This calculation method has a higher identification rate, and has a great significance to intellectual location of defects.
出处
《电气化铁道》
2016年第6期34-38,共5页
Electric Railway
关键词
接触网安全巡检
图像分析
支柱定位
号牌识别
OCS safety patrol
image analysis
pole location
number plate identification