摘要
为了解决在野外复杂背景下架空高压线路中类圆部件的缺陷诊断问题,提出了一种基于曲线聚类的类圆部件的识别算法,即利用高压电路部件间的位置关系、类圆部件对象的曲线凸凹方向特征和曲线密度等来识别类圆部件对象。提出的方法成功地应用于架空高压线路的直升机巡检系统中,可排除大部分背景干扰,能可靠精确地识别类圆曲线对象。
In order to solve the defect diagnosis of relate similar circle components of the high-voltage line in the complex outdoor background category, an effective identification algorithm of similar round parts based on the curve clustering method is proposed in this paper. These similar circular curve objects are identified by the cluster analysis in which those relations that the parts position relationship between the various components of overhead high voltage transmission lines which combines with the direction characteristics of convex, concave curves, the curve density of the objects. The proposed method is successfully applied into the helicopter overhead high voltage line inspection system which could effectively eliminate most of the background interference and would gain reliable and accurate identification of similar circular curve object.
出处
《电视技术》
北大核心
2012年第7期143-148,共6页
Video Engineering
关键词
高压架空电力线系统
直升机巡检
类圆对象
HESSIAN矩阵
曲线识别
聚类分析
high-voltage overhead power line system
helicopter inspection
similar round object
Hessian matrix
curve recognition
cluster analysis
