基于视觉注意模型的输电线路违规检测方法研究  

Research on transmission line violation detection method based on visual attention model

作　　者：张伟奎 钱涛 余长江 邱斌 邹帅 苏宇晨 ZHANG Weikui;QIAN Tao;YU Changjiang;QIU Bin;ZOU Shuai;SU Yuchen(Xinjiang information industry co.,ltd.,Urumqi 830026,China)

机构地区：[1]新疆信息产业有限责任公司,新疆乌鲁木齐830026

出　　处：《电视技术》2018年第10期8-11,24,共5页Video Engineering

摘　　要：本研究站在视觉注意模型的基础上,制定出了一种输电线路违规检测方法,分析了违规线路和其他环境的图像差异,分类计算线路边缘直方图特征,与其他线路环境的图像梯度分布相比,违规线路差异明显。对其展开修补,并采取一定的手段来划分违规线路,使其成为不同的阶段。全面分析整合机器人自身传感器的特点,将其与违规线路检测结果相结合,对现阶段机器人的运行情况进行判定,此外根据当前补修作业流程运作模型,对状态转移函数进行补充完善,形成规划的完成需要借助状态转移等手段开展起来。此外违规线路检测实验、行为规划实验等均离不开实验室模拟线路环境的支撑,实验表明该方法的有效性。In this paper,a detection method of transmission line violation based on visual attention model is proposed.The image differences between illegal lines and other environments are analyzed,and the characteristics of the line edge histogram are classified and calculated.Compared with the image gradient distribution in other line environment,the difference between illegal lines is obvious.In the course of repair,the violation line will be divided into several stages.By combining the information of the robot s own sensor and the result of the detection of the illegal line,the current state of the robot is determined,and the state transfer function is established by repairing the operation process,and the behavior planning is completed through the state transfer.At the same time,the behavior planning and violation line detection experiments are carried out in the simulated circuit environment of the laboratory.The experiment shows the effectiveness of the method.

关 键 词：视觉注意模型 违规 输电线路 检测 图像梯度 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]