摘要
分析了矿井外因火灾监测方法,提出了基于可见光和红外图像的矿井外因火灾识别方法。①温度和烟雾等传感器矿井外因火灾监测方法需多点布置,成本高、维护工作量大。②可见光图像矿井外因火灾监测方法具有监测范围广、使用简单、成本低、受距离影响小、分辨率高、色彩信息丰富、边缘纹理信息清晰、火焰结构信息便于提取等优点,但受矿灯、巷道灯、车灯及红色物体等影响大。③红外图像矿井外因火灾监测方法具有监测范围广、使用简单、可在浓烟等能见度低情况下监测火灾等优点,但监视距离、矿尘、环境温度和湿度、物体表面发射率等均会影响温度测量,近距离的作业人员、白炽灯及机电设备等均会影响火灾监测。④基于可见光和红外图像的矿井外因火灾识别方法提高了火灾识别准确率:当可见光图像和红外图像均识别出火灾时,才判定为火灾,发出火灾报警信号;当仅有单一可见光图像或红外图像识别出火灾时,不判定为火灾,发出疑似火灾报警信号。⑤通过试验验证了基于可见光和红外图像的矿井外因火灾识别方法的可行性。
Mine exogenous fire monitoring methods were analyzed,and a mine exogenous fire identification method based on visible light and infrared image was proposed.① Mine exogenous fire monitoring methods based on sensors for temperature,smoke and others need to arrange multiple points,so cost is high,and maintenance workload is large.② Mine exogenous fire monitoring method based on visible light image has advantages of large monitoring range,simple use,low cost,small influence by distance,high resolution ratio,rich color information,clear edge texture information,easy extraction of flame structure information,etc.But it is influenced greatly by cap-lamp lights,roadway lights and red objects.③ Mine exogenous fire monitoring method based on infrared image has advantages of large monitoring range and simple use,and can monitor fire in case of low visibility with smoke or others.But monitoring distance,mine dust,ambient temperature or humidity and emissivity of object surface all affect temperature measurement.Close-range workers,incandescent lamps and electromechanical equipment will affect fire monitoring.④ The mine exogenous fire identification method based on visible light and infrared image improves accuracy of fire identification.When both visible light image and infrared image identify fire,it is determined to be a fire,and a fire alarm signal is issued.When there is only a single visible light image or infrared image to identify fire,it is not determined to be a fire,and a suspected fire alarm signal is issued.⑤ The feasibility of the mine exogenous fire identification method based on visible light and infrared image is verified by experiments.
作者
孙继平
孙雁宇
范伟强
SUN Jiping;SUN Yanyu;FAN Weiqiang(China University of Mining and Technology (Beijing),Beijing 100083,China)
出处
《工矿自动化》
北大核心
2019年第5期1-5,21,共6页
Journal Of Mine Automation
基金
国家重点研发计划资助项目(2016YFC0801800)
关键词
矿井火灾
外因火灾
火灾识别
可见光图像
红外图像
图像识别
mine fire
exogenous fire
fire identification
visible light image
infrared image
image recognition
作者简介
孙继平(1958-),男,山西翼城人,教授,博士,博士研究生导师,中国矿业大学(北京)信息工程研究所所长;获国家科技进步二等奖3项;作为第1完成人获省部级科技进步特等奖和一等奖8项;作为第1完成人主持制定中华人民共和国煤炭行业和安全生产行业标准28项;主持制定《煤矿安全规程》第十一章“监控与通信”;作为第1作者或独立完成著作12部;被SCI和EI检索的第1作者或独立完成论文80余篇;作为第1发明人获国家授权发明专利60余项;作为国务院煤矿事故调查专家组组长参加了10起煤矿特别重大事故调查工作;E-mail:sjp@cumtb.edu.cn。
