基于林格曼黑度的烟囱排放自动监测分析系统的研究与实现
|
摘要
大中型联合钢铁企业的生产制造流程长,能源资源消耗大,在区域环境中所占的污染负荷也较大。随着国家节能减排实施纲要的落实,企业环境保护不仅是建设环境友好型企业需要,更是企业的生命线,所以必然要加强对污染源的监控。但钢厂区域范围大,烟囱林立,分布广泛,靠人来监控烟囱的排放显然不能做到及时有效,因此开发烟囱排放林格曼黑度自动监测系统十分必要。
烟囱排放林格曼黑度监测系统的作用是对厂范围内的所有烟囱排放进行图像监控,并对拍摄到的图像进行林格曼黑度分析、报警,对各烟囱排放的历史数据进行统计分析,帮助环保管理人员寻找重点大气污染源,与生产经营信息对照后,还可以找出超标排放的原因,实施源头治理。
烟囱排放林格曼黑度自动监测系统由现场视频设备、光纤传输网络、监控中心和系统监控软件等部分组成。现场视频设备(摄像头)拍摄到的现场图像,通过局域光纤网络传输到监控中心的系统监控软件,被监控软件显示、分析和处理。该系统在图像数据的传输上充分利用了计算机网络技术,集成了网络的远程控制、通信、信息处理优势,并应用先进的计算机图象处理技术,对烟囱排放的林格曼黑度进行计算机分析和自动判级。
本文论述了烟囱排放林格曼黑度自动监测系统实现的几个关键技术:研究了以网络视频服务器为核心的总体框架结构,采用了适应数字图像传输的IP多播网络传输技术,从客户角度对系统监控软件的功能模块配置进行了系统论述,并对其中的两个关键模块(视频设备控制模块和自动轮巡模块)进行了详细论述。本文着重研究了数字图像处理技术在烟尘图像林格曼黑度测量上的运用,通过对各种图像预处理方法、边缘检测方法、图像分割方法进行实验比较,最终确定了最佳阀值分割法作为烟尘图像检测的基本算法。计算机分析林格曼黑度时,首先由最佳阀值分割法从数字图像中分割出烟尘图像,然后统计烟尘图像的加权平均灰度值即为烟囱排放的林格曼黑度值,再按照用户设定的黑度值区间划分黑度等级。
本软件已在钢铁企业应用,实践证明系统运行稳定,计算机分析的林格曼黑度值和黑度级的符合我们肉眼观测到的黑度级,满足现场环保管理要求。
Large and middle scale unites Iron and steel enterprise's manufacturing flow to be long, the energy resources consume in a big way, occupies the pollution load in the region environment to be also big. Reduces the row of implementation summary along with the national energy conservation realization, the enterprise environment protection not only constructs the environment friendly enterprise need, but also enterprise's lifeline. Therefore,It is bound to strengthen the monitoring of pollution sources. But the steel mill area coverage is big, the chimney stands in great numbers, the distribution is widespread, monitors chimney's emissions depending on the human not to be able to achieve obviously real-time effective. Therefore, it is very essential that the design and development automatic monitoring of Ringelmann blackness of chimney emissions.
The chimney emissions Ringelmann blackness monitoring system's function is carries on the image frequency monitoring to the factory scope all chimney emissions, and to the picture which monitors carries on the Ringelmann blackness analysis, the warning,discharges the historical data to various chimneys to carry on the statistical analysis, helps the environmental protection administrative personnels to seek for the key atmospheric source of pollution, after production operation information comparison, but may also discover the reason for the exceeding the allowed figure emissions, in order to implement source governance.
The chimney emissions Ringelmann blackness automatic monitoring system is composed by the on-site video equipment, fiber-optic transmission network, monitoring control center and video surveillance software. The scene material which the on-site video equipment (camera) observes transmits the monitors software through the confined network, the software shows the monitored, analysis and processing. The system of image data transmission make full use of the computer Network Technology. It is Integrated the network remote control, communications, information processing advantages and application of advanced computer image processing technology, on the chimney’s Ringelmann blackness testing and computer automatic class.
This article discusses the system implementation of some key technologies: Network Video Server as the core of the overall framework, IP multicast network transmission technology and module settings of monitoring software. It describes the two key modules. The focus attention in the article is digital image processing technique and how to apply on Ringelmann blackness arithmetic testing by result of image analysis. Through experiments to determine the optimal threshold partition image as a smoke detection algorithm.
This system has been applied in the iron and steel enterprise, the practice had proven that the system is stable and credible. Computer analysis results consistents with the people’s determine. it has reached a pre-set requirements of us.
烟囱排放林格曼黑度监测系统的作用是对厂范围内的所有烟囱排放进行图像监控,并对拍摄到的图像进行林格曼黑度分析、报警,对各烟囱排放的历史数据进行统计分析,帮助环保管理人员寻找重点大气污染源,与生产经营信息对照后,还可以找出超标排放的原因,实施源头治理。
烟囱排放林格曼黑度自动监测系统由现场视频设备、光纤传输网络、监控中心和系统监控软件等部分组成。现场视频设备(摄像头)拍摄到的现场图像,通过局域光纤网络传输到监控中心的系统监控软件,被监控软件显示、分析和处理。该系统在图像数据的传输上充分利用了计算机网络技术,集成了网络的远程控制、通信、信息处理优势,并应用先进的计算机图象处理技术,对烟囱排放的林格曼黑度进行计算机分析和自动判级。
本文论述了烟囱排放林格曼黑度自动监测系统实现的几个关键技术:研究了以网络视频服务器为核心的总体框架结构,采用了适应数字图像传输的IP多播网络传输技术,从客户角度对系统监控软件的功能模块配置进行了系统论述,并对其中的两个关键模块(视频设备控制模块和自动轮巡模块)进行了详细论述。本文着重研究了数字图像处理技术在烟尘图像林格曼黑度测量上的运用,通过对各种图像预处理方法、边缘检测方法、图像分割方法进行实验比较,最终确定了最佳阀值分割法作为烟尘图像检测的基本算法。计算机分析林格曼黑度时,首先由最佳阀值分割法从数字图像中分割出烟尘图像,然后统计烟尘图像的加权平均灰度值即为烟囱排放的林格曼黑度值,再按照用户设定的黑度值区间划分黑度等级。
本软件已在钢铁企业应用,实践证明系统运行稳定,计算机分析的林格曼黑度值和黑度级的符合我们肉眼观测到的黑度级,满足现场环保管理要求。
Large and middle scale unites Iron and steel enterprise's manufacturing flow to be long, the energy resources consume in a big way, occupies the pollution load in the region environment to be also big. Reduces the row of implementation summary along with the national energy conservation realization, the enterprise environment protection not only constructs the environment friendly enterprise need, but also enterprise's lifeline. Therefore,It is bound to strengthen the monitoring of pollution sources. But the steel mill area coverage is big, the chimney stands in great numbers, the distribution is widespread, monitors chimney's emissions depending on the human not to be able to achieve obviously real-time effective. Therefore, it is very essential that the design and development automatic monitoring of Ringelmann blackness of chimney emissions.
The chimney emissions Ringelmann blackness monitoring system's function is carries on the image frequency monitoring to the factory scope all chimney emissions, and to the picture which monitors carries on the Ringelmann blackness analysis, the warning,discharges the historical data to various chimneys to carry on the statistical analysis, helps the environmental protection administrative personnels to seek for the key atmospheric source of pollution, after production operation information comparison, but may also discover the reason for the exceeding the allowed figure emissions, in order to implement source governance.
The chimney emissions Ringelmann blackness automatic monitoring system is composed by the on-site video equipment, fiber-optic transmission network, monitoring control center and video surveillance software. The scene material which the on-site video equipment (camera) observes transmits the monitors software through the confined network, the software shows the monitored, analysis and processing. The system of image data transmission make full use of the computer Network Technology. It is Integrated the network remote control, communications, information processing advantages and application of advanced computer image processing technology, on the chimney’s Ringelmann blackness testing and computer automatic class.
This article discusses the system implementation of some key technologies: Network Video Server as the core of the overall framework, IP multicast network transmission technology and module settings of monitoring software. It describes the two key modules. The focus attention in the article is digital image processing technique and how to apply on Ringelmann blackness arithmetic testing by result of image analysis. Through experiments to determine the optimal threshold partition image as a smoke detection algorithm.
This system has been applied in the iron and steel enterprise, the practice had proven that the system is stable and credible. Computer analysis results consistents with the people’s determine. it has reached a pre-set requirements of us.
引文
[1]国家环境保护总局,国家质量监督检验检疫总局.GB13271-2001锅炉大气污染物排放标准[S].北京:中国标准出版社,2001.
[2]杜丽娜.烟道粉尘浓度在线监测.东北大学硕士论文.2005.P1～P11.
[3]国家环境保护总局.HJ/T398-2007,固定污染源排放烟气黑度的测定林格曼烟气黑度图法[S].北京:2007.
[4]东松,乔立杰,李洪滨,任秀兰,李明亮.粉尘浓度的激光测量[J],红外与激光,1997,27(l):39-41
[5]KoiehiLinoya,KeishiGotoh,KoHigashitani,PowderTeehnologyHandbook[J],NewYork.1991:23-27.
[6]袁杰,沈庆宏,都思丹等.烟气黑度多媒体联网监测系统[J].环境污染与防治,2003,25(3):183-185.
[7]郑炯,朗文鹏.工业烟尘黑度自动监测系统的研制[J].上海大学学报,1995,(06).
[8]林劲.基于网络的视频监控系统软件的设计与实现.中国硕士学位论文全文数据库,2006,(04).
[9]张晓春.计算机图像识别的烟气自动监测[J]环境监测管理与技术, 2003,(05).
[10]张虹,任玉升.利用高斯混合模型进行烟气黑度智能检测[J].环境污染与防治,2008-7,30(7):82-84.
[11]宋克国,周清梅.影响烟气林格曼黑度观测的相关因素探讨[J].山东环境,2005.(5):47
[12]蔡焕兴.烟气黑度监测与结果计算[J].上海环境科学, 1993,(09):32-33
[13]王耀南等编著.计算机图像处理与识别技术[M].高等教育出版社.2001.P55-211
[14]王润生.图像理解[M].国防科技大学出版社, 1995.P1-5
[15]沈庭芝,方子文编著.数字图像处理及模式识别[M].北京理工大学出版社, 1998.P11-33.
[16]飞思科技产品研发中心编著.MATLAB6.5辅助图像处理[M].电子工业出版社.2003.(1):P117-119
[17]彭仪普.傅立叶变换原理及其在数字图像处理中的应用[J].铁路航测,2000(3):4-5
[18]张才华.沃尔什函数在信号综合中的应用研究[J].淅江工贸职业技术学院学报,2005,5(2):68-69
[19]李弼程,等编著.智能图像处理技术[M ].北京:电子工业出版社, 2004.
[20]冯清枝.基于直方图修正的图像增强技术[ J ].广东公安科技,2004,(2):49-51
[21] GB9078-1996.工业炉窑大气污染物排放标准[S]
[22]孙宏琦,施维颖,巨永锋.利用中值滤波进行图像处理[J].长安大学学报,2003-3,23(2):104-106.
[23]孙增国,韩崇昭.基于Laplacian算子的图像增强[J].计算机应用研究,2007(1):222-223.
[24]吕俊.基于快速Kirsch与边缘点概率分析的边缘提取[J].计算机应用,2001-2,21(2):33-35.
[25]王訃堂,杨善林.基于Gauss-Laplace算子的灰度图低胸边缘检测[J].计算机工程与应用,2003,26:132-134.
[26]王磊,莫玉龙,戚飞虎.基于Canny理论的边缘提取改善方法[ J ].中国图像图形学报,1996, 7 ( 3 ):191-193.
[27]杨振亚,白治江,王成道.自适应Canny边缘检测算法[ J ].上海海运学院学报, 2003, 24 (4):373 -377.
[28]吴宇岚.实时图象理解系统中符号提取的设计与实现和小波变换在图象处理中的应用[D].浙江大学学位论文,1996,1.
[29]史册.对一种快速边缘跟踪算法的讨论[J].小型微型计算机系统,2000,21(6):641-645.
[30]葛彭.一种快速边缘跟踪与提取的新算法[J].微电子学与计算机,2005,22(8):14-17.
[31]王新成.高级图象处理技术[M]:中国科学技术出版社,2000.
[32]章毓晋.图象分割[M].北京:科学出版社, 2001.
[33]Nikhil R. Pal and Sankar K. Pal, A review on image segmentation techniques. Pattern Rcognition 1993;26(9):1277-1294.
[34]吴一全,朱兆达.图像处理中阈值选取方法30年(1962-1992)的进展(一).数据采集与处理, 1993, 9(3): 193~201.
[35]吴一全,朱兆达.图像处理中阈值选取方法30年(1962-1992)的进展(二).数据采集与处理, 1993, 9(4): 268-281.
[36]赵荣椿,迟耀斌,朱重光.图像分割技术进展:中国体视学与图像分析,1998, 3(2): 121-128.
[37]王秋让.基于自动门限化的图像分割以目标提取方法研究[D],西北工业大学博士论文.
[38] Rosenfeld A, Torre P De La. Histogram concavity analysis as an aid in threshold selection. IEEE Trans, 1983; SMC-13(2): 231-235.
[39] Otsu N. Discriminant and least square threshold selection. In: Proc 4IJCPR, 1978:592-596.
[40] Kittler J, Illingworth J. On threshold selection using clustering criteria. IEEE Trans. 1985; SMC-15: 652-655.
[41]于新文.几种图像分割算法在棉铃虫图像处理中的应用[C].中国农业大学学报,2001,6
[42] Pun T. A new method for gray-level picture thresholding using the entropy of histogram. Signal Processing, 1980; 2: 223-237.
[43] Kapur J N, Sahoo P K, Wong A K C. A new method for gray-level picture thresholding using the entropy of the histogram. Computer Vision, Graphics and Image Processing, 1985; 29: 273-285.
[44] Kittler J, Illingworth J. Minimum error thresholding. Pattern Recognition, 1986; 19(1): 41-47.
[45] Nakagawa Y, Rosenfeld A. Some experiments on variable thresholding. Pattern Recognition, 1979,11: 191-204.
[46]Duff M J B.Computing structures for Image Proeessing,Addisor.Wesley Publishing Company,USA,1983:31-102.
[47]Dvaid M.kroenke,”Database Processing: Fandamentals, Design, and Implementation”, Prentice Hall, Inc,1995.
[48] STAUFFER C,GRIMSON W. Learning patterns of activity usingreal-time tracking[J].IEEETransactions on Pattern Recognitionand Machine Intelligence (TPAMI),2000,22(8):747-757.
[49]OIKAWAT,TOMIZAWAM,DEGAWAS.Newmonitoring system forthermal power plantsusing digital image processing and sound analysis[J].ControlEngineeringPractice,1997,5 (1) :75-78..
[50] A grawal, Vishw aniD. Design of mixed-signal systems for testa-bility Integration. The VL S I Journal, 1999, 26 (1) : 145-147.
[2]杜丽娜.烟道粉尘浓度在线监测.东北大学硕士论文.2005.P1～P11.
[3]国家环境保护总局.HJ/T398-2007,固定污染源排放烟气黑度的测定林格曼烟气黑度图法[S].北京:2007.
[4]东松,乔立杰,李洪滨,任秀兰,李明亮.粉尘浓度的激光测量[J],红外与激光,1997,27(l):39-41
[5]KoiehiLinoya,KeishiGotoh,KoHigashitani,PowderTeehnologyHandbook[J],NewYork.1991:23-27.
[6]袁杰,沈庆宏,都思丹等.烟气黑度多媒体联网监测系统[J].环境污染与防治,2003,25(3):183-185.
[7]郑炯,朗文鹏.工业烟尘黑度自动监测系统的研制[J].上海大学学报,1995,(06).
[8]林劲.基于网络的视频监控系统软件的设计与实现.中国硕士学位论文全文数据库,2006,(04).
[9]张晓春.计算机图像识别的烟气自动监测[J]环境监测管理与技术, 2003,(05).
[10]张虹,任玉升.利用高斯混合模型进行烟气黑度智能检测[J].环境污染与防治,2008-7,30(7):82-84.
[11]宋克国,周清梅.影响烟气林格曼黑度观测的相关因素探讨[J].山东环境,2005.(5):47
[12]蔡焕兴.烟气黑度监测与结果计算[J].上海环境科学, 1993,(09):32-33
[13]王耀南等编著.计算机图像处理与识别技术[M].高等教育出版社.2001.P55-211
[14]王润生.图像理解[M].国防科技大学出版社, 1995.P1-5
[15]沈庭芝,方子文编著.数字图像处理及模式识别[M].北京理工大学出版社, 1998.P11-33.
[16]飞思科技产品研发中心编著.MATLAB6.5辅助图像处理[M].电子工业出版社.2003.(1):P117-119
[17]彭仪普.傅立叶变换原理及其在数字图像处理中的应用[J].铁路航测,2000(3):4-5
[18]张才华.沃尔什函数在信号综合中的应用研究[J].淅江工贸职业技术学院学报,2005,5(2):68-69
[19]李弼程,等编著.智能图像处理技术[M ].北京:电子工业出版社, 2004.
[20]冯清枝.基于直方图修正的图像增强技术[ J ].广东公安科技,2004,(2):49-51
[21] GB9078-1996.工业炉窑大气污染物排放标准[S]
[22]孙宏琦,施维颖,巨永锋.利用中值滤波进行图像处理[J].长安大学学报,2003-3,23(2):104-106.
[23]孙增国,韩崇昭.基于Laplacian算子的图像增强[J].计算机应用研究,2007(1):222-223.
[24]吕俊.基于快速Kirsch与边缘点概率分析的边缘提取[J].计算机应用,2001-2,21(2):33-35.
[25]王訃堂,杨善林.基于Gauss-Laplace算子的灰度图低胸边缘检测[J].计算机工程与应用,2003,26:132-134.
[26]王磊,莫玉龙,戚飞虎.基于Canny理论的边缘提取改善方法[ J ].中国图像图形学报,1996, 7 ( 3 ):191-193.
[27]杨振亚,白治江,王成道.自适应Canny边缘检测算法[ J ].上海海运学院学报, 2003, 24 (4):373 -377.
[28]吴宇岚.实时图象理解系统中符号提取的设计与实现和小波变换在图象处理中的应用[D].浙江大学学位论文,1996,1.
[29]史册.对一种快速边缘跟踪算法的讨论[J].小型微型计算机系统,2000,21(6):641-645.
[30]葛彭.一种快速边缘跟踪与提取的新算法[J].微电子学与计算机,2005,22(8):14-17.
[31]王新成.高级图象处理技术[M]:中国科学技术出版社,2000.
[32]章毓晋.图象分割[M].北京:科学出版社, 2001.
[33]Nikhil R. Pal and Sankar K. Pal, A review on image segmentation techniques. Pattern Rcognition 1993;26(9):1277-1294.
[34]吴一全,朱兆达.图像处理中阈值选取方法30年(1962-1992)的进展(一).数据采集与处理, 1993, 9(3): 193~201.
[35]吴一全,朱兆达.图像处理中阈值选取方法30年(1962-1992)的进展(二).数据采集与处理, 1993, 9(4): 268-281.
[36]赵荣椿,迟耀斌,朱重光.图像分割技术进展:中国体视学与图像分析,1998, 3(2): 121-128.
[37]王秋让.基于自动门限化的图像分割以目标提取方法研究[D],西北工业大学博士论文.
[38] Rosenfeld A, Torre P De La. Histogram concavity analysis as an aid in threshold selection. IEEE Trans, 1983; SMC-13(2): 231-235.
[39] Otsu N. Discriminant and least square threshold selection. In: Proc 4IJCPR, 1978:592-596.
[40] Kittler J, Illingworth J. On threshold selection using clustering criteria. IEEE Trans. 1985; SMC-15: 652-655.
[41]于新文.几种图像分割算法在棉铃虫图像处理中的应用[C].中国农业大学学报,2001,6
[42] Pun T. A new method for gray-level picture thresholding using the entropy of histogram. Signal Processing, 1980; 2: 223-237.
[43] Kapur J N, Sahoo P K, Wong A K C. A new method for gray-level picture thresholding using the entropy of the histogram. Computer Vision, Graphics and Image Processing, 1985; 29: 273-285.
[44] Kittler J, Illingworth J. Minimum error thresholding. Pattern Recognition, 1986; 19(1): 41-47.
[45] Nakagawa Y, Rosenfeld A. Some experiments on variable thresholding. Pattern Recognition, 1979,11: 191-204.
[46]Duff M J B.Computing structures for Image Proeessing,Addisor.Wesley Publishing Company,USA,1983:31-102.
[47]Dvaid M.kroenke,”Database Processing: Fandamentals, Design, and Implementation”, Prentice Hall, Inc,1995.
[48] STAUFFER C,GRIMSON W. Learning patterns of activity usingreal-time tracking[J].IEEETransactions on Pattern Recognitionand Machine Intelligence (TPAMI),2000,22(8):747-757.
[49]OIKAWAT,TOMIZAWAM,DEGAWAS.Newmonitoring system forthermal power plantsusing digital image processing and sound analysis[J].ControlEngineeringPractice,1997,5 (1) :75-78..
[50] A grawal, Vishw aniD. Design of mixed-signal systems for testa-bility Integration. The VL S I Journal, 1999, 26 (1) : 145-147.