大型起重机金属结构无人机智能视觉检测系统研究  

Research on UAV Intelligent Visual Inspection System for Large Crane Metal Structure

作　　者：周前飞 丁树庆 冯月贵 张慎如 王会方 任诗波 ZHOU Qian-fei;DING Shu-qing;FENG Yue-gui;ZHANG Shen-ru;WANG Hui-fang;REN Shi-bo(Nanjing Special Equipment Inspection Institute,Nanjing 210019,China)

机构地区：[1]南京市特种设备安全监督检验研究院,南京210019

出　　处：《内燃机与配件》2018年第19期224-228,共5页Internal Combustion Engine & Parts

基　　金：国家质量监督检验检疫总局科技计划项目(2017QK139)

摘　　要：针对起重机长期承载、疲劳与腐蚀效应、材料老化、磨损、维护保养不良等造成的金属结构疲劳裂纹、表面锈蚀、局部损伤等问题,利用无人机系统技术和机器视觉理论,研究应用于高空、高温、大风等危险恶劣环境下的远程自动检测技术,代替或辅助完成人所不能、不适或力所不及的各项结构检测工作。系统采用无人机作为飞行平台,搭载高分辨率视觉传感器探测装置,实时采集显示被检测区域的图像,利用数字图像处理技术进行缺陷特征提取与识别,以求从根本上改变起重机重要检测部位在危险检测环境下的人工检验方式,进一步提高特种设备的自动化检测水平。For the crane metal structure fatigue crack,surface rust,local damage and other issues caused by long-term load,fatigue and corrosion effect,material aging,wear and tear,poor maintenance,etc,the remote automatic detection technology applied to the high altitude,high temperature,high wind and other bad dangerous environment is researched.Through the use of unmanned aerial vehicle(UAV)system technology and machine vision theory,it can replace or assist persons to accomplish the structure test work which persons cannot complete,not fitting,or beyond the reach.The system uses the UAV as a flying platform carrying a high resolution visual sensor detection device,it can real-time capture and show the images of the tested area,and use the digital image processing techniques for defect feature extraction and recognition.It can fundamentally change the manual inspection testing way for the crane's important testing parts in a dangerous environment,and further improve the level of special equipment automation testing.

关 键 词：起重机械 金属结构检测 无人机 视觉检测 多尺度几何分析 支持向量机 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]