多示例的输电线路区域车辆识别算法

Based on Multi-example Transmission Line Regional Vehicle Identification Algorithm

输电线路周围区域出现的挖掘机等特种车辆是引起输电线路损害的高风险因素,因此成为输电线智能视频分析系统重点监控的对象。本文提出了一种多示例学习方法,采用局部加权的Citation-kNN算法,以训练样本分布为依据,为选举集合中的元素设置局部权值。通过实验验证该方法,能够有效地区分出输电线路区域内的车辆类型。

The special vehicles such as excavators, which appear in the area around the transmission line, are the high risk factors cau- sing transmission line damage. Therefore, they are the key monitoring objects of the intelligent video analysis system for transmission lines. This paper presents a multi instance learning method, Citation -kNN （CkNN） algorithm using a locally weighted, in full con- sideration of the distribution of training samples, according to the distribution of samples to vote for the elements of the set the corre- sponding weights. Experimental results show that this method can effectively distinguish the types of vehicles in the transmission line area.