南京理工大学学报（自然科学版）2017，Vol.41Issue(6)：708-713,6.DOI:10.14177/j.cnki.32-1397n.2017.41.06.007
基于改进椭圆拟合与非线性支持向量机的配电设备螺栓带电检测
Online detection for bolts of power distribution equipments based on improved ellipse fitting and nonlinear support vector machine
摘要
Abstract
A novel ellipse fitting approach based on the least square method and random sampling principle is proposed to detect bolts in complex live working scenes and solve the problems of the effects of illumination,environment and noise on traditional ellipse fitting. A sorting method based on nonlinear support vector machine is proposed to classify bolts from distractors and solve the problem of alike edge of bolts and elliptical distractors. Experimental results show that,the proposed method can detect the bolts effectively and accurately from distractors.关键词
椭圆拟合/支持向量机/配电设备/螺栓检测/带电检测/最小二乘法/随机采样原理Key words
ellipse fitting/support vector machine/power distribution equipments/bolt detection/online detection/least square method/random sampling principle分类
信息技术与安全科学引用本文复制引用
纪良,吴巍,许春山,苏鹏飞,赵伟..基于改进椭圆拟合与非线性支持向量机的配电设备螺栓带电检测[J].南京理工大学学报（自然科学版）,2017,41(6):708-713,6.基金项目
江苏省重点研发计划项目(BE2017161) （BE2017161）
江苏高校优势学科建设工程项目 （）
