基于视觉检测的连铸中间包钢水液位测量方法的研究
|
摘要
中间包钢水液位是保证连铸生产顺利进行的重要工艺参数之一。中间包钢水液位的实时测量与控制对于稳定浇注、防止漏钢事故、提高铸坯质量及增加中间包内衬耐材寿命等均有直接影响。特别是在停浇末期,根据准确的钢水液位信息,尽可能地减少中间包内钢水的剩余量而又保证不卷渣,从而提高钢水收得率,增加企业效益。因此,准确地测量中间包钢水液位具有实际意义。
中间包钢水液位测量是一个至今未能解决的难题,由于中间包内钢水上方存在厚度不定的保护渣(覆盖剂、钢渣和钢水中上浮的杂质),以及高温、多尘、强干扰等恶劣环境均给钢水液位的测量带来困难,使得超声、涡流、微波和视觉等方法一直不能应用于中间包钢水液位的测量。
针对中间包钢水液位测量难题,本文研究一种基于视觉检测的钢水液位测量方法,其研究工作包括机理模型的建立与分析、温度信息的合理提取、温度梯度的微弱目标定位、测量系统误差校正及现场试验。主要研究内容与创新工作如下：
(1)基于视觉检测的中间包钢水液位测量方法
本文测量钢水液位的方法是利用一支耐火材料制成的测量棒插入中间包钢水中感知温度,由于中间包内从上至下存在具有不同温度变化速率的空气层、保护渣层、钢水层三层介质,热平衡后测量棒轴向温场分布在空气层-保护渣层以及保护渣层-钢水层分界面处会出现温度梯度的突变点,利用这一特性,通过视觉检测定位两个分界面位置,从而得到钢水液位高度。
为从理论上分析本论文方法的可行性,基于有限元分析建立了测温棒三维非稳态传热模型,并进行了数值计算,通过现场实际温度测量数据与模型计算结果比较对模型进行验证。数值计算结果分析表明,测量棒纵向温度场分布在保护渣层-钢水层分界面以及空气层-保护渣层分界面处均存在明显的温度梯度的突变点,为本文的钢水液位测量方法提供了理论依据。
通过测量系统的参数对液位测量结果的影响分析,并结合现场实际条件,确定系统的测量参数。视觉系统空间分辨率最低约为2mm/pixel,最高可达0.4mm/pixel,满足测量精度要求。
(2)测量棒温度信息的可靠提取及温梯度分界而定位
针对常规的图像分割方法对于本文测量图像在边沿定位精度与干扰消除级间存在矛盾,不能较好地解决测量棒与背景、噪声的分离问题,提出和研究一种改进的粒子群ostu分割方法,该方法利用目标和背景的最大类间方差作为判据,实现了测量棒和背景噪声的分离,同时用与信息熵相结合的自适应步长改进粒子群搜索方法,实现最优分割阈值的获取,阈值搜索速度提高到迭代法的2.4倍,最终实现测量棒温度信息的可靠提取。
针对在复杂噪声及流渣干扰下保护渣层-钢水层分界面的微弱特征被掩盖而无法可靠定位的难题,提出和研究一种基于图像信息与机理模型相结合的微弱目标定位方法。以机理模型模拟计算得到的目标特征判据,抑制图像数据中的复杂噪声及流渣干扰等形成的伪特征,实现微弱目标的可靠搜索、提取和定位,分界面定位准确率达到95%左右。
(3)液位测量系统校正与现场试验
提出和研究一种改进的单目视觉畸变校正方法,该方法是在实现控制点的自动、准确、鲁棒提取的基础上,建立畸变图像与原始图像的多项式映射模型,并结合双曲线插值运算对视觉系统进行校正,从而减小钢水液位测量偏差。视觉畸变导致液位测量最大偏差约20mm,校正后偏差减小到0.4mm。同时,针对系统动态测量中的参数变化问题,利用基准参照物的物象关系实现参数的校正,校正后,参数变化对测量影响偏差小于2.5mm。为验证测量准确性,利用手动插入金属棒验证测量钢水液位,本文的液位测量方法与手动验证方法相比较,测量统计偏差为4.3mm。
机理模型分析与现场试验结果表明：本文方法可实现准确可靠的中间包钢水液位测量,测量原理可推广到其他类似的熔融金属液位测量场合,具备很好的应用前景。
Molten steel liquid level in tunish is an important information during continuous casting production. And during pouring, the real-time measurement of molten steel liquid level has many effects on controlling pouring stability, improving quality of casting billet, preventing breakout and increasing the lifetime of refractory material on tundish and so on. Especially, on premise of guaranteeing casting quality, the accurate molten steel liquid level information can be applied to reduce the surplus molten steel in the tundish as much as possible, hence, metallurgical enterprise should benefit from increasing of steel yielding rate. Hence, it is of practical significance to study the liquid level measurement of molten steel in the tundish.
At present, molten steel liquid level measurement in tundish is a difficult problem which has not been solved. It is found that not only the mould power, including covering flux and impurities between the slags and the molten steel, but also the tundish in high-temperature, multiple-dust and strong interference, cause difficulties to the liquid level measurement in tundish. Hence, some methods, such as ultrasonic, eddy current, microwave and vision and so on, can not be applied to molten steel liquid level measurement.
Because it is hard to apply the liquid level measurement of tundish for practical engineering, a new molten steel liquid level measuring method based on vision detection was put forward in this paper. In addition, the article also has research on setting up and analyzing of the mechanism model, extracting accurate temperature, locating temperature gradient of feebleness objective, correcting distortion on measuring system and carrying on the experiment in the metallurgical field and so on. In summary, some works and innovation are presented as followed:
1. Molten steel liquid level measuring method based on vision detection
The principle of the method, researched in this paper, is that a measuring bar which is made of refractory material was inserted into molten steel in tundish to percept for the temperature.And as air layer, flux slag layer and molten metal layer which have different temperature change velocity to each other, locates in the tundish, after heat balance, the temperature distribution of the measuring bar will appears of catastrophe point at the interface between the air layer, the flux slag layer and the molten metal layer. Then, the molten steel liquid level can be obtained from the two interfaces by visual detection.
In order to analyzing the feasibility of the method in theory, a numerical calculation is carried out through the3d unsteady heat-transferring modeling, based on the finite element analysis. Then, compared with the result of the temperature measured on the filed, the calculating temperature on the modeling is showed that there are some catastrophe points of longitudinal temperature distrubition in the interface between the air layer, the flux slag layer and the molten metal layer. Therefore, the method provides the theoretical basis for measuring the molten steel liquid level in the papers.
Besides, after analyzing the measuring parameters of the system at actual condition, the resolution of system spatial, varied from0.4mm/pixel to2mm/pixel, meets accuracy requirements of metallurgical field.
2. Extracting temperature of measuring bar reliably and locating interface of temperature gradient
The conventional image segmentation method is inapplicable to deal with the separation between measuring bar, background and noise. However, according to the characteristics of measuring images, the improved segmentation method on ostu image of particle swarm, given in this paper, can realize the optimal segmentation between them. Simultaneously, the optimal segmentation threshold must be obtained reasonably and quickly from the improved adaptive step searching for particle swarm method which connects with information entropy. Finally, as the speed of searching for threshold can be2.4times as high as the iterative method, the improved segmentation method will achieve to extracting the temperature on the measuring bar feasibly.
As feebleness objective, interface between the flux slag layer and the molten metal layer covered up by complicated noise and interference may not be located reliably, the thesis establishes feebleness objective location method based on image information and mechanism modeling. And according to the criterion for objection feature, obtained from simulating through the mechanism model, the false feature forming for complicated noise, interference in image can be restrained. So the feebleness objective is searched, extracted and located. And the accuracy rate about the interface location is up to about95%.
3. Correcting molten steel liquid level measuring system and test
An improved correcting method of monocular vision distortion is put forward in this paper. It is used to not only extract the control point automatically and accurately but also build up the modeling, which connected with the bilinear interpolation operation, of polynomial mapping between distorted images and original ones. Considering correcting precision and complexity of operation, the five degree polynomial is the best of all. After analyzing the result, it is found that the error caused by distortion is about20mm between the measuring liquid level and actual ones before distortion correction, but0.4mm after distortion correction. Simultaneously, it uses the relationship between reference object and its image correct measuring paramaters which may be vary in dynamic measuring. Deviation of the measuring result is less than2.5mm after paramater-correction. In order verify accuracy of the molten steel liqiud level measuring method, compared the level measuring results with hand-measured results, it is found that the average deviation is about4.3mm.
As well, the results obtained from the mechanism modeling and test in the industrial field suggest that the method proposed in this paper can reliably be used to liquid level measure on the spot. In addition, in order to meet to industrial application, the measure principle can also popularize to other molten metal liquid level. So molten steel liquid level measure method based on temperature has wide application prospect and popularization value.
中间包钢水液位测量是一个至今未能解决的难题,由于中间包内钢水上方存在厚度不定的保护渣(覆盖剂、钢渣和钢水中上浮的杂质),以及高温、多尘、强干扰等恶劣环境均给钢水液位的测量带来困难,使得超声、涡流、微波和视觉等方法一直不能应用于中间包钢水液位的测量。
针对中间包钢水液位测量难题,本文研究一种基于视觉检测的钢水液位测量方法,其研究工作包括机理模型的建立与分析、温度信息的合理提取、温度梯度的微弱目标定位、测量系统误差校正及现场试验。主要研究内容与创新工作如下：
(1)基于视觉检测的中间包钢水液位测量方法
本文测量钢水液位的方法是利用一支耐火材料制成的测量棒插入中间包钢水中感知温度,由于中间包内从上至下存在具有不同温度变化速率的空气层、保护渣层、钢水层三层介质,热平衡后测量棒轴向温场分布在空气层-保护渣层以及保护渣层-钢水层分界面处会出现温度梯度的突变点,利用这一特性,通过视觉检测定位两个分界面位置,从而得到钢水液位高度。
为从理论上分析本论文方法的可行性,基于有限元分析建立了测温棒三维非稳态传热模型,并进行了数值计算,通过现场实际温度测量数据与模型计算结果比较对模型进行验证。数值计算结果分析表明,测量棒纵向温度场分布在保护渣层-钢水层分界面以及空气层-保护渣层分界面处均存在明显的温度梯度的突变点,为本文的钢水液位测量方法提供了理论依据。
通过测量系统的参数对液位测量结果的影响分析,并结合现场实际条件,确定系统的测量参数。视觉系统空间分辨率最低约为2mm/pixel,最高可达0.4mm/pixel,满足测量精度要求。
(2)测量棒温度信息的可靠提取及温梯度分界而定位
针对常规的图像分割方法对于本文测量图像在边沿定位精度与干扰消除级间存在矛盾,不能较好地解决测量棒与背景、噪声的分离问题,提出和研究一种改进的粒子群ostu分割方法,该方法利用目标和背景的最大类间方差作为判据,实现了测量棒和背景噪声的分离,同时用与信息熵相结合的自适应步长改进粒子群搜索方法,实现最优分割阈值的获取,阈值搜索速度提高到迭代法的2.4倍,最终实现测量棒温度信息的可靠提取。
针对在复杂噪声及流渣干扰下保护渣层-钢水层分界面的微弱特征被掩盖而无法可靠定位的难题,提出和研究一种基于图像信息与机理模型相结合的微弱目标定位方法。以机理模型模拟计算得到的目标特征判据,抑制图像数据中的复杂噪声及流渣干扰等形成的伪特征,实现微弱目标的可靠搜索、提取和定位,分界面定位准确率达到95%左右。
(3)液位测量系统校正与现场试验
提出和研究一种改进的单目视觉畸变校正方法,该方法是在实现控制点的自动、准确、鲁棒提取的基础上,建立畸变图像与原始图像的多项式映射模型,并结合双曲线插值运算对视觉系统进行校正,从而减小钢水液位测量偏差。视觉畸变导致液位测量最大偏差约20mm,校正后偏差减小到0.4mm。同时,针对系统动态测量中的参数变化问题,利用基准参照物的物象关系实现参数的校正,校正后,参数变化对测量影响偏差小于2.5mm。为验证测量准确性,利用手动插入金属棒验证测量钢水液位,本文的液位测量方法与手动验证方法相比较,测量统计偏差为4.3mm。
机理模型分析与现场试验结果表明：本文方法可实现准确可靠的中间包钢水液位测量,测量原理可推广到其他类似的熔融金属液位测量场合,具备很好的应用前景。
Molten steel liquid level in tunish is an important information during continuous casting production. And during pouring, the real-time measurement of molten steel liquid level has many effects on controlling pouring stability, improving quality of casting billet, preventing breakout and increasing the lifetime of refractory material on tundish and so on. Especially, on premise of guaranteeing casting quality, the accurate molten steel liquid level information can be applied to reduce the surplus molten steel in the tundish as much as possible, hence, metallurgical enterprise should benefit from increasing of steel yielding rate. Hence, it is of practical significance to study the liquid level measurement of molten steel in the tundish.
At present, molten steel liquid level measurement in tundish is a difficult problem which has not been solved. It is found that not only the mould power, including covering flux and impurities between the slags and the molten steel, but also the tundish in high-temperature, multiple-dust and strong interference, cause difficulties to the liquid level measurement in tundish. Hence, some methods, such as ultrasonic, eddy current, microwave and vision and so on, can not be applied to molten steel liquid level measurement.
Because it is hard to apply the liquid level measurement of tundish for practical engineering, a new molten steel liquid level measuring method based on vision detection was put forward in this paper. In addition, the article also has research on setting up and analyzing of the mechanism model, extracting accurate temperature, locating temperature gradient of feebleness objective, correcting distortion on measuring system and carrying on the experiment in the metallurgical field and so on. In summary, some works and innovation are presented as followed:
1. Molten steel liquid level measuring method based on vision detection
The principle of the method, researched in this paper, is that a measuring bar which is made of refractory material was inserted into molten steel in tundish to percept for the temperature.And as air layer, flux slag layer and molten metal layer which have different temperature change velocity to each other, locates in the tundish, after heat balance, the temperature distribution of the measuring bar will appears of catastrophe point at the interface between the air layer, the flux slag layer and the molten metal layer. Then, the molten steel liquid level can be obtained from the two interfaces by visual detection.
In order to analyzing the feasibility of the method in theory, a numerical calculation is carried out through the3d unsteady heat-transferring modeling, based on the finite element analysis. Then, compared with the result of the temperature measured on the filed, the calculating temperature on the modeling is showed that there are some catastrophe points of longitudinal temperature distrubition in the interface between the air layer, the flux slag layer and the molten metal layer. Therefore, the method provides the theoretical basis for measuring the molten steel liquid level in the papers.
Besides, after analyzing the measuring parameters of the system at actual condition, the resolution of system spatial, varied from0.4mm/pixel to2mm/pixel, meets accuracy requirements of metallurgical field.
2. Extracting temperature of measuring bar reliably and locating interface of temperature gradient
The conventional image segmentation method is inapplicable to deal with the separation between measuring bar, background and noise. However, according to the characteristics of measuring images, the improved segmentation method on ostu image of particle swarm, given in this paper, can realize the optimal segmentation between them. Simultaneously, the optimal segmentation threshold must be obtained reasonably and quickly from the improved adaptive step searching for particle swarm method which connects with information entropy. Finally, as the speed of searching for threshold can be2.4times as high as the iterative method, the improved segmentation method will achieve to extracting the temperature on the measuring bar feasibly.
As feebleness objective, interface between the flux slag layer and the molten metal layer covered up by complicated noise and interference may not be located reliably, the thesis establishes feebleness objective location method based on image information and mechanism modeling. And according to the criterion for objection feature, obtained from simulating through the mechanism model, the false feature forming for complicated noise, interference in image can be restrained. So the feebleness objective is searched, extracted and located. And the accuracy rate about the interface location is up to about95%.
3. Correcting molten steel liquid level measuring system and test
An improved correcting method of monocular vision distortion is put forward in this paper. It is used to not only extract the control point automatically and accurately but also build up the modeling, which connected with the bilinear interpolation operation, of polynomial mapping between distorted images and original ones. Considering correcting precision and complexity of operation, the five degree polynomial is the best of all. After analyzing the result, it is found that the error caused by distortion is about20mm between the measuring liquid level and actual ones before distortion correction, but0.4mm after distortion correction. Simultaneously, it uses the relationship between reference object and its image correct measuring paramaters which may be vary in dynamic measuring. Deviation of the measuring result is less than2.5mm after paramater-correction. In order verify accuracy of the molten steel liqiud level measuring method, compared the level measuring results with hand-measured results, it is found that the average deviation is about4.3mm.
As well, the results obtained from the mechanism modeling and test in the industrial field suggest that the method proposed in this paper can reliably be used to liquid level measure on the spot. In addition, in order to meet to industrial application, the measure principle can also popularize to other molten metal liquid level. So molten steel liquid level measure method based on temperature has wide application prospect and popularization value.
引文
1.高文芳,刘成信,马勤学.连铸中间包水力模型研究[J],钢铁,1997,32：740-742.
2. Da-peng Tan, Pei-yu Li, Li Xu, etal. Steel water continuous casting slag detection system based on VQ[C], Proceedings of 2006 IEEE International Conference on Systems,2006, 1315-1319.
3.张瑞,张婕宇,王波等.中间包液面波动的物理模拟[J],中国稀土学报,2004,22：273-277.
4. Kazuya Asano, Takayuki Kaji, Hidemi Aoki, etal. Robust molten steel level control for continuous casting[C], Proceedings of The 35th IEEE International Conference on Decision and Control,1996,1245-1250.
5. Hiroshi Kitada, Osamu Kondo, Hiroshi Kusachi. H∞ control of molten steel level in continuous caster[J], IEEE Transaction on Control Systems Technology,1998,6(2): 200-207.
6. Miguel A. Barron, Jesus Gonzales, Filepe Vargas. An implicit controller for a steel continuous caster[C], Proceedings of The 41st IEEE International Conference on Decision and Control,2002,2486-2491.
7. Miguel A. Barron, Ricardo Aguilar,Jesus Gonzales. Stabilization of the molten steel level in the mold of a continuous casting machine by means of vacuum[J], IEEE Transaction on Industry Application,2000,36(3):861-864.
8.刘明辉.基于探极与非线性PID液位检测控制系统研究[D],沈阳：东北大学,2005.
9. Pardeshi R. Mathematical modeling of the tundish of a single-strand slab caster[J]. ISIJ International,2004,44(9):1534-1540.
10. Uemura T, Nashiwa H, Hashio M, etal. Automatic system of continuous slab casting machine[J], Sumitomo Search,1974,12:7-18
11. Shiraiwa T, Sakamoto Y, Kobayashi S, etal. Automatic control of casting speed in ingot casting[J], Automatica,1981,17(4):613-618.
12. Martin H. Mansel E, Schinkels G. Effects by using automatic casting equipment with memory-programmable control[J]. LEW-Nachrichten,1989,18(39):30-31.
13. Takawa Takeshi, Takamoto Tsutomu. Ohashi Yasutake, etal. Automatic casting method of twin belt type thin slab caster[J]. Tetsu-To-Hagane/Journal of the Iron and Steel Institute of Japanl989,75(2):275-281.
14. Niu Jungao, Su Shuhong, Liu Yan, etal. Influence of MgO-CaO based spraying material and CaO based cover flux for tundish on steel cleanness[J], Iron and Steel,2002,37(8): 25-27.
15.任开春,涂亚庆.20余种液位测量方法分析[J],工业仪表与自动化装置,2003,5：12-16.
16.唐耀庚.同位素检测在钢水液位测量中的应用[J],传感器技术,2001,20(3)：39-41.
17. Gao Feiyan. A novel isotopic mould measurement system for molten steel level based on real time algorithm[J], Journal of Test and Measurement Technology,2007,22(6): 505-509.
18. Kohler M, Hult M, Arnold D, etal. Reference measurements of low levels of 60Co in steel[J],2004,61(2-3):207-211.
19. Hoftetter K J, Beals D M, Mclane G A, etal. In situ measurements of 137Cs in the Steel Creek watershed at SRS[J], Transactions of the American Nuclear Society,2001,85: 133-135.
20. Sekeni Y. Ultrasonic liquid level detector for steel tank, measures frequency characteristic profile of electric impedance by monitoring impedance change of ultrasonic transducer accompanying sweep of continuous sine wave applied to liquid [P]:JP2009109296-A.
21. Chaki S, Bourse G. Guided ultrasonic waves for non-destructive monitoring of the stress levels in prestressed steel strands[J], Ultrasonics,2009,49(2):162-171.
22. An S I, Kwon J H. Apparatus and method for measuring level of molten steel by eddy current sensor [P]:KR2001087620-A.
23. Katakin R A, Pokrovskiy A D. A study of the eddy-current method of testing the level of molten steel in the thin-slab crystallization process[J], Russian Journal of Nondestructive Testing,2009,45(8):542-548.
24宋东飞.涡流检测在板坯连铸钢水液位的应用[J],冶金设备,2003,5：50-52.
25. Lalla Robert, Muller Roland, Strutt Bernd. Measuring molten steel level by using microwave method [P]:Endress+Hauser GmbH+Co, EP98108121.9.
26. Kawata Y, Kusaca T, Manabe C. etal. A precision microwave rangefinder and its application to level measurement in the steel industry [J]. Industrial Metrology,1990,1(1): 19-32.
27. Imai Takashi. New technology development of microwave level-meter for ironmaking and steelmaking processes[J], Research and Development Kobe Steel Engineering Reports, 2007,57(3):16-20.
28.王渠东,徐建伦.移动电极法金属液位渣厚自动测试[P]:CN97106380.X.
29.朱丽叶,方志宏,吴惕华等.采用CCD传感器进行薄带连铸熔池液位的测量[J],自动化仪表, 2005,25(12)：30-34.
30. Arakawa Denki. Surface level measuring device processes visual image obtained from image pick-up which photographs L-shaped refraction and V-shaped reflection of light on molten metal surface to measure surface level[P]:JP2000230850-A.
31. Kadota Junichi, Kohama Kikuo. Accurate steel melt level measuring method in a mould-by dipping a steel rod into mould which is then coated with mould powder and a steel melt layer then measuring change in rod dimensions[P]:NKK CORP, JP8168860-A.
32. Cockerell R A, Graebe S F, Elsley G, etal. Design study for continuous caster tundish weight control and ladle steel flow estimate[J], Automation in the Steel Industry,1998: 235-240.
33.王双起.中包称重控制系统的抗干扰措施[J],可编程控制器与工厂自动化,2009：60-61.
34.向祖权,茅云生,张建华.潜艇废气排放气泡视觉测量系统[J],造船技术,2006, 3：35-37.
35. A. R. Conway. Autonomous Control of an Unstable Helicopter using Carrier Phase GPS Only[M], USA:Stanford University,1995.
36.张广军,周富强.基于双圆特征的无人机着陆位置姿态视觉测量方法[J],航空学报,2005,26(3)：344-348.
37.孔明,汤强晋,王式民.锅炉煤粉仓粉量立体视觉测量研究[J],锅炉研究,2006,37(2)：21-25,75.
38.刘伟,徐斌.卫星气浮平台视觉测量系统研究[J],微计算机信息,2008,28(4)：124-126.
39. S. S. Cory, O. Shakernia, S. Shankar. A Vision System for Landing an Unmanned Aerial Vehicle[C]. Proceedings of 2001 IEEE International Conference of Robotics and Automations.2001.21-26.
40.许俐.机器视觉测量大米留胚率的研究[J].江苏理工大学学报.1997.18(6)：8-11.
41.王克奇.木质材料表面缺陷计算机视觉测量的识别方法[J],林业科学,1996,32(1)：92-96.
42.苏建,崔乃斌,刘玉梅等.汽车整车尺寸视觉测量系统的研究[J],公路交通科技,2007,24(4)：145-149.
43.李新华,宋承祥,刘弘.双目视觉测量在乒乓球运动速度测量中的应用[J],计算机科学,2008,35(3)：256-257.
44. Yue Hong, Li Kai, Zhao Haiwen. Vision-based pipeline girth-welding robot and image processing of weld seam[J], Industrial Robot,2009,36(3):284-289.
45.刘鹏,王滨,张文明等.基于视觉传感器的焊缝图像采集与分析[J],现代焊接,2008,71：13-15.
46.邹怡容,吴哲明,曾凯.基于焊缝纹理特征的视觉识别方法[J],焊接学报,2009,30(5)：13-16.
47. Tokuhiro A T, Vadakattu S. Inspecting a research reactor's control rod surface for pitting using a machine vision[J],2005,42(11):994-1000.
48. Changhyun Park, Seho Choi, Sangchul Won. Vision-based inspection for periodic defects in steel wire rod production[J], Optical Engineering,2010,49(1):202-212.
49. Jong-Wook Kim, Jong Pil Yun, Sang Woo Kim, etal.Automatic detection of cracks in raw steel block using Gabor filter optimized by univariate dynamic encoding algorithm for searches (uDEAS)[J], NDT&E International,2009,42(5):389-397.
50. Jong Pil Yun, SungHoo Choi, Boyeul Seo, etal. Visual surface inspection of scale-covered steel billets[C], Proceedings of the Tenth IASTED International Conference on Computer Graphics and Imaging,2008,64-69
51.吴晓鹏.林介邦,唐辉等.基于机器视觉的铸坯表面缺陷检测系统的研制[J].武钢技术,2010,48(1)：49-53.
52. Tang Bo, Kong Jian-Yi, Wang Xing-Dong. Surface inspection system of steel strip based on machine vision[C]. Proceedings of 2009 1st International Workshop on Database Technology and Applications,2009,259-262.
53. Fan Zhansuo, Tang bing, Hao Junyu. Application of cut-to-length by using a pickup-camera to capture images and measure the length in the re-cutting procedure of steel billet[J]. Control & Automation,2005,1:21-22.
54. Tao Liu. Yi Zheng; Zong-yi Wang. etal. Dynamic image processing for visual robot cutting steel plate[C],2008 Second International Symposium on Intelligent Information Technology Application,2008,716-720.
55. Cheng Peng-fei, Yang Fan, Yan Li-qiang, etal. Automatic counting system of steel rods based on machine vision[J], Metallurgical Industry Automation,2008,32(1):22-26.
56. Qi Sui-ping, Zhang Hong-jian, Li Xiu-juan, etal. Counting steel rods online using LQV neural network in real-time images[C], Proceedings of the 2006 IEEE International Conference on Information Acquisition,2006,956-960.
57.李俊飞.一种棒材计数装置的研究[J],计算机测量与控制,2004,12(12)：1217-1218.
58. Sarychev A F, Galibuzov I P, Epp P A. Proceddings int workshop om industrial applications of machine intelligence & vision[J], Metallurgist,1989,32(7-8):236.
59. Graham K J, Krishnapisharody K, Irons G A, etal. Ladle eye area measurement using multivariate image analysis[J], Canadian Metallurgical Quarterly,2007,46(4):397-406.
60.杨庆军,王绍纯.计算机单目视觉检测连铸结晶器熔钢液面高度[J],北京科技大学学报,1999,21(5)：498-501.
61. Pfob C, Niel K S, Roessler R. Flushing analysis by machine vision and fuzzy logic at molten steel for the automation process[C], Proceedings of the SPIE-The International Society for Optical Engineering,2006,81-85.
62. LI Dingyi, WANG Yongran. CAO Yongguo, etal. Physico-mathematical model of heat transfer for molten steel temperature in continuous casting [J], Iron & Steel,1996,11: 17-21.
63. So R M C, Sommer T P. Explicit algebraic heat-flux model for the temperature field[J], International Journal of Heat and Mass Transfer,1996,39(3):455-465.
64. Guo Zhipeng, Xiong Shoumei, Cho Sang-Hyun, etal. Development of an inverse heat transfer model and its application in the prediction of the interfacial heat flux[J], Acta Metallurgica Sinica,2007,43(6):607-611.
65. Mikrovas AC, Argyropoulos S A, Sommerville I D. Heat transfer characteristics of molten slags [J], Ironmaking and Steelmaking,1991,18(3):169-181.
66. Solhed Henrik, Jonsson Lage. An investigation of slag floatation and entrapment in a continuous-casting tundish using fluid-flow simulations, sampling and physical metallurgy [J]. Scandinavian Journal of Metallurgy,2003,32(1):15-32.
67.陶文铨.传热学[M],西安：西北工业大学,2006：24-105.
68.徐平坤,董应榜.刚玉耐火材料[M],北京：冶金工业出版社,1999,263-292.
69.王诚训,孙炜明,张义先等.钢包用耐火材料[M],北京：冶金工业出版社：2003,31-51.
70.朱立光,张玉文,张淑会.钢水覆盖剂传热及熔融结构的数学模型[J],炼钢,1999,4(2)：29-32.
71. Lee Jye-Long, Chen J K. Pan YT, etal. Evaluation of transformation latent heat in C-Mn steels[J], ISIJ International,1999,36(3):281-287.
72. Shiina Y. Inagaki T.Study on the efficiency of effective thermal conductivities on melting characteristics of latent heat storage capsules[J], International Journal of Heat and Mass Transfer,2005,48(2):373-383.
73.周建兴,刘瑞祥.凝固过程数值模拟中的潜热处理方法[J],铸造,2001(07)：34-39.
74.谢植,次英,孟红记等.基于在线黑体空腔理论的钢水连续测温传感器的研制[J],仪器仪表学报,2005,26(5)：446-448.
75.谢植,高魁明.工业辐射温度测量[M],沈阳：东北大学出版社,1993,113-129.
76.冈萨雷斯.数字图像处理Matlab版[M],北京：电子工业出版社,2005.
77.罗希平,田捷,诸葛婴.图像分割方法综述[J],模式识别与人工智能,1999,12(3)：300-312.
78. Navon E, Miller O, Averbuch A. Color Image Segmentation Based on Adaptive Local Thresholds[J].Image and Vision Computing,2005,23 (1):69-85.
79.冉玉梅,王洪国,杨玉会.基于自适应局部阈值的彩色图像分割[J],计算机与现代化,2010,173(1)：109-112.
80. Ety Navon, OferMiller, Am ir Averbuch. Color image seg-mentation based on adaptive local thresholds[J]. Imageand Vision Computing.2005,23(1):69-85.
81. PavlidisT, Liow Y Integrating region growing and edge detection[J]. IEEE Transactions on PatternAnalysis and Machine Intelligence,1990,12(3):225-233.
82. Krishnan Nallaperumal, Justin Varghese, etal. Iterative adap-tive switching median filter[C]. Proceedings of IEEE Conference on Industrial Electronics and Applications. Singapore,2006,1-6.
83.牟春洁,张国英.基于区域边界生长的图像分割方法[J],北京石油化工学院学报,2009,17(4)：8-12.
84. Ikonomatakis N. Plataniotis K N, Zervakis M, etal. Region growing and region merging image segmentation[J].Digital Signal Proeessing Proeeedings,1997, (1),1299-1302.
85.陈小光,封举富.基于多特征判别分析的指纹图像分割方法[J],北京大学学报,2009,(1)：1-5.
86. DORIN C, PETER M. Mean shift:A robust approach toward feature space analysis[J], IEEE Transactions on Pattern Analysis and Machine Intelligence,2002,24(5):603-619.
87.李光,王朝英,侯志强.基于K均值聚类与区域合并的彩色图像分割算法[J],计算机应用,2010,30(2)：354-358.
88. SHI J, MALIK J. Normalized cuts and image segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2000,22(8):888-905.
89. Yen J C, Chang F J, Chang S. A new criterion for automatic mutilevel thresholding[J], IEEE Transaction on image processing,1995,4(3):370-377.
90. Pikaz A, Averbuch A. Digital image thresholding based on topological stable state[J], Pattern Recognition,1996,29(5):829-843.
91. Papamarkos N, Gatos B. A new approach for multilevel threshold selection. Graphic Models and Image Processing,1994,56:357-370.
92. Saboo P, Wilkins C, Yeager J. Threshold selection using Renyi's entropy. Pattern Recognition,1997,30(1):71-84.
93. Brink A D. Thresholding of digital image using two-dimensional entropies[J]. Pattern Recognition,1992,25(8):803-808.
94. Chen H D, Chen J R, Li J G. Theshold selection based on fuzzy C-partition entropy approach. Pattern Recognition,1998,31(7):857-870.
95.蔡梅艳,吴庆宪,姜长生.采用进化规划和最大熵的目标图像分割算法[J],光电与控制.2008,15(4)：13-16.
96. de Albuquerque M P, Esquef I A, de Albuquerque M P. Image segmentation using non-extensive relative entropy[J], IEEE Latin America Transactions,2008,6(5): 477-483.
97. Mayszko Dariusz, Stepaniuk Jarosaw. Adaptive rough entropy clustering algorithms in image segmentation[J]. Fundamenta Informaticae,2010,98(2):199-213.
98.齐敏,邓林.弋丹等.灰关联空间中基于最大熵阈值的医学彩色图像分割算法[J],中国科学技术大学学报.2007,37(12)：1543-1545.
99. Nalwa V S. Binford T O. On detecting edges[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,1986,8(6):699-714.
100. Haralick R M. Digital step edges from zero crossing of second directional derivatives[J], IEEE Transactions on Pattern Analysis and Machine Intelligence,1984,6(1):58-68.
101. Kang Mu, Xu Qinggong, Wang Baoshu. A Roberts' adaptive edge detection method[J], Journal of Xi'an Jiaotong University,2008,42(10):1240-1244.
102. Alam M S, Iftekharuddin K M, Karim M A. Polarization-encoded optical shadow casting: Edge detection using Roberts operator[J], Microwave and Optical Technology Letters, 1993,6(3):190-193.
103.杨振亚,王勇,王成道.LOG算子边缘检测方法的改进方案[J],计算机应用与软件,2004,21(9)：87-90.
104. Sheng-Chieh Huang, Liang-Gee Chen, Hao-Chieh Chang. A novel image compression algorithm by using Log-Exp transform[J], Proceedings of the 1999 IEEE International Symposium on Circuits and Systems VLSI,1999,17-20.
105. Hoi Steven C H, Lyu Michael R, Jin Rong. A unified log-based relevance feedback scheme for image retrieval [J], IEEE Transactions on Knowledge and Data Engineering, 2006,18(4):509-523.
106. Huiling Zhang, Baichuan Lu, Yong Yu. The method of Ostu used in the video detection of vehicles[C],2008 International Conference on Intelligent Computation Technology and Automation,2008,547-551.
107. Jiang Hui-yan, Cui Xiao-liang, Zhou Xiao-jie, etal. Image segmentation based on improved dual fast marching method[J], Journal of System Simulation,2008,20(3): 803-806.
108.韩建勋,饶欣.复杂系统建模——高维特征空间变量法[J],自动化学报,1991,17(2)：160-165.
109.郭文忠,陈国龙,陈水利等.基于粒子群优化的分类规则挖掘方法及其应用[J],集美大学学报,2008；3：34-38.
110.李峻金,向阳,芦英明等.粒子群聚类算法综述[J],计算机应用研究,2009,26(12)：4423-4427.
111. Frigu I H, Krishnaram R. A robust competitive clustering algorithm with application in computer vision[J]. IEEE Transaction on Pattern Analysis and Machine Intelligence, 1999.21(5):450-465.
112.高尚,杨静宇.群智能算法及其应用[M],北京：中国水利水电出版社,2006：79-81.
113.吴琳丽,赵海娜,汪涛等.新的全局-局部最优最小值粒子群优化算法[J],计算机应用,2009,29(12)：3270-3272.
114.郑春颖,郑全弟,王晓丹等.基于试探的变步长自适应粒子群算法[J],计算机科学,2009,36(11)：193-195.
115.赵峙江,张田文,张志宏.一种基于视觉模型与连通域统计的闽值分割新算法[J],电子学报,2005,33(5)：793-797.
116. BieniekA, Moga A. An efficientwatershed algorithm based on connect components[J]. Pattern Recognition,2000,33(6):907-916.
117.张恒,胡文龙,丁赤飙.基于快速连通域分析的目标特征提取算法[J],计算机工程与应用,2009,45(29)：230-232.
118. Lifeng He, Yuyan Chao, Suzuki K. A run-based two-scan labeling algorithm[J], IEEE Transaction on image processing,2008,17(5):131-142.
119.宋斌.一种新的图像连通域快速标号算法[J],电子测量技术,2009,32(9)：67-68.
120.徐正光,鲍东来,张利欣.基于递归的二值图像连通域像素标记算法[J],计算机工程,2006,32(24)：186-188.
121.管业鹏,童林夙等.基于数码相机的三维立体空间几何位置的摄影测量[J],电子学报,2002,30(6)：849-852.
122. Cheng Yumin. Advances and review on moving least-square methods[J], Computer Aided Engineering,2009,18(2):5-11.
123. Hu Juan-li, Deng Jia-bin, Hu Chang. An algorithm for circle curve fitting based on the constrained least square model represented by mosaic observation points[C], Proceedings of the 2009 5th International Conference on Wireless Communications, Networking and Mobile Computing,2009,4.
124. Melter R A, Stojmenovic I, Zunic J. A new characterization of digital lines by least square fits[J], Pattern Recognition Letters,1993,14(2):83-88.
125. Bo Chang, Kanefsky M. Adaptive average filters for image processing[C], Twenty-Sixth Annual Allerton Conference on Communication, Control, and Computing,1998, 459-460.
126. Ray B K. Ray K S. Corner detection using iterative Gaussian smoothing with constant window size[J]. Pattern Recognition,1995.28(11):1765-1781.
127. Tay Meng Keat Christopher, Laugier Christian. Modelling smooth paths using Gaussian processes[C], Field and Service Robotics:Results of the 6th International Conference, 2008,381-390.
128. Luo Hai-Bo, Shi Ze-Lin, Hui Ying, etal. Real-time large window -sized 2D median filter based on multi-Phased grouping and sorting network[J], Infrared and Laser Engineering, 2008,37(5):935-939.
129. Wang X, Wang D. On the max/median filter[J], IEEE Transactions on Acoustics, Speech and Signal Processing,1990,38(8):1473-1475.
130.周俊峰.高精度薄板带板凸度激光检测的误差分析与精度控制研究[D],长沙：中南大学,2006.
131.廖士中,董明,高培焕.工业摄像机几何畸变校正控制点的自动生成[J],计算机应用研究,2001,(10)：30-32.
132. Fusiello A E, Trucco A V. Rectification with unconstrained stereo geometry[C], Proceedings of the British Machine Vision Conference, Colchester,1997:400-409.
133. Fusiello A E, Trucco A V. A compact algorithm for rectification of stereo pairs [J], Machine Vision and Applications,2000,22(12):16-22.
134. Cooke E. Rectification[C], An Overview, HHI Report'99. Berlin,1999,201-202.
135.郭永刚,葛庆平,冯平等.一种基于控制点自动提取的图像畸变校正算法[J],计算机工程与应用,2007,43(2)：214-216.
136. Shuang K, Dong S P. A progressive approach method for improving the accuracy of binocular vision measurement[J], Opto-Electronic Engineering,2000,27(3):30-33.
137. Ouellet J N, Hebert P. Developing assistant tools for geometric camera calibration assessing the quality of input images[J]. Pattern Recognition,2004.4:80-83.
138. Yang C H, Lai S H, etal. Reliable image matching via modified Hausdorff distance with normalized gradient consistency measure[C], Proceedings of 3rd International Conference on Information Technology:Research and Education, Taiwan,2005: 158-161.
139. Manzar M A, Cheema T A,Jalil A, etal. New image matching technique based on hyper-vectorisation of grey level sliced binary image[J], Image Processing,2008,2(6): 337-351.
140. Shieh Hsin-Jang, Huang Po-Kai. Trajectory tracking of piezoelectric positioning stages using a dynamic sliding-mode control[J], IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control,2006,53(10):1872-1882.
141. Zhaohua Hu, Xin Fan, Yaoliang Song, etal. Joint trajectory tracking and recognition based on bi-directional nonlinear learning[J], Image and Vision Computing,2009,27(9): 1302-1312.
2. Da-peng Tan, Pei-yu Li, Li Xu, etal. Steel water continuous casting slag detection system based on VQ[C], Proceedings of 2006 IEEE International Conference on Systems,2006, 1315-1319.
3.张瑞,张婕宇,王波等.中间包液面波动的物理模拟[J],中国稀土学报,2004,22：273-277.
4. Kazuya Asano, Takayuki Kaji, Hidemi Aoki, etal. Robust molten steel level control for continuous casting[C], Proceedings of The 35th IEEE International Conference on Decision and Control,1996,1245-1250.
5. Hiroshi Kitada, Osamu Kondo, Hiroshi Kusachi. H∞ control of molten steel level in continuous caster[J], IEEE Transaction on Control Systems Technology,1998,6(2): 200-207.
6. Miguel A. Barron, Jesus Gonzales, Filepe Vargas. An implicit controller for a steel continuous caster[C], Proceedings of The 41st IEEE International Conference on Decision and Control,2002,2486-2491.
7. Miguel A. Barron, Ricardo Aguilar,Jesus Gonzales. Stabilization of the molten steel level in the mold of a continuous casting machine by means of vacuum[J], IEEE Transaction on Industry Application,2000,36(3):861-864.
8.刘明辉.基于探极与非线性PID液位检测控制系统研究[D],沈阳：东北大学,2005.
9. Pardeshi R. Mathematical modeling of the tundish of a single-strand slab caster[J]. ISIJ International,2004,44(9):1534-1540.
10. Uemura T, Nashiwa H, Hashio M, etal. Automatic system of continuous slab casting machine[J], Sumitomo Search,1974,12:7-18
11. Shiraiwa T, Sakamoto Y, Kobayashi S, etal. Automatic control of casting speed in ingot casting[J], Automatica,1981,17(4):613-618.
12. Martin H. Mansel E, Schinkels G. Effects by using automatic casting equipment with memory-programmable control[J]. LEW-Nachrichten,1989,18(39):30-31.
13. Takawa Takeshi, Takamoto Tsutomu. Ohashi Yasutake, etal. Automatic casting method of twin belt type thin slab caster[J]. Tetsu-To-Hagane/Journal of the Iron and Steel Institute of Japanl989,75(2):275-281.
14. Niu Jungao, Su Shuhong, Liu Yan, etal. Influence of MgO-CaO based spraying material and CaO based cover flux for tundish on steel cleanness[J], Iron and Steel,2002,37(8): 25-27.
15.任开春,涂亚庆.20余种液位测量方法分析[J],工业仪表与自动化装置,2003,5：12-16.
16.唐耀庚.同位素检测在钢水液位测量中的应用[J],传感器技术,2001,20(3)：39-41.
17. Gao Feiyan. A novel isotopic mould measurement system for molten steel level based on real time algorithm[J], Journal of Test and Measurement Technology,2007,22(6): 505-509.
18. Kohler M, Hult M, Arnold D, etal. Reference measurements of low levels of 60Co in steel[J],2004,61(2-3):207-211.
19. Hoftetter K J, Beals D M, Mclane G A, etal. In situ measurements of 137Cs in the Steel Creek watershed at SRS[J], Transactions of the American Nuclear Society,2001,85: 133-135.
20. Sekeni Y. Ultrasonic liquid level detector for steel tank, measures frequency characteristic profile of electric impedance by monitoring impedance change of ultrasonic transducer accompanying sweep of continuous sine wave applied to liquid [P]:JP2009109296-A.
21. Chaki S, Bourse G. Guided ultrasonic waves for non-destructive monitoring of the stress levels in prestressed steel strands[J], Ultrasonics,2009,49(2):162-171.
22. An S I, Kwon J H. Apparatus and method for measuring level of molten steel by eddy current sensor [P]:KR2001087620-A.
23. Katakin R A, Pokrovskiy A D. A study of the eddy-current method of testing the level of molten steel in the thin-slab crystallization process[J], Russian Journal of Nondestructive Testing,2009,45(8):542-548.
24宋东飞.涡流检测在板坯连铸钢水液位的应用[J],冶金设备,2003,5：50-52.
25. Lalla Robert, Muller Roland, Strutt Bernd. Measuring molten steel level by using microwave method [P]:Endress+Hauser GmbH+Co, EP98108121.9.
26. Kawata Y, Kusaca T, Manabe C. etal. A precision microwave rangefinder and its application to level measurement in the steel industry [J]. Industrial Metrology,1990,1(1): 19-32.
27. Imai Takashi. New technology development of microwave level-meter for ironmaking and steelmaking processes[J], Research and Development Kobe Steel Engineering Reports, 2007,57(3):16-20.
28.王渠东,徐建伦.移动电极法金属液位渣厚自动测试[P]:CN97106380.X.
29.朱丽叶,方志宏,吴惕华等.采用CCD传感器进行薄带连铸熔池液位的测量[J],自动化仪表, 2005,25(12)：30-34.
30. Arakawa Denki. Surface level measuring device processes visual image obtained from image pick-up which photographs L-shaped refraction and V-shaped reflection of light on molten metal surface to measure surface level[P]:JP2000230850-A.
31. Kadota Junichi, Kohama Kikuo. Accurate steel melt level measuring method in a mould-by dipping a steel rod into mould which is then coated with mould powder and a steel melt layer then measuring change in rod dimensions[P]:NKK CORP, JP8168860-A.
32. Cockerell R A, Graebe S F, Elsley G, etal. Design study for continuous caster tundish weight control and ladle steel flow estimate[J], Automation in the Steel Industry,1998: 235-240.
33.王双起.中包称重控制系统的抗干扰措施[J],可编程控制器与工厂自动化,2009：60-61.
34.向祖权,茅云生,张建华.潜艇废气排放气泡视觉测量系统[J],造船技术,2006, 3：35-37.
35. A. R. Conway. Autonomous Control of an Unstable Helicopter using Carrier Phase GPS Only[M], USA:Stanford University,1995.
36.张广军,周富强.基于双圆特征的无人机着陆位置姿态视觉测量方法[J],航空学报,2005,26(3)：344-348.
37.孔明,汤强晋,王式民.锅炉煤粉仓粉量立体视觉测量研究[J],锅炉研究,2006,37(2)：21-25,75.
38.刘伟,徐斌.卫星气浮平台视觉测量系统研究[J],微计算机信息,2008,28(4)：124-126.
39. S. S. Cory, O. Shakernia, S. Shankar. A Vision System for Landing an Unmanned Aerial Vehicle[C]. Proceedings of 2001 IEEE International Conference of Robotics and Automations.2001.21-26.
40.许俐.机器视觉测量大米留胚率的研究[J].江苏理工大学学报.1997.18(6)：8-11.
41.王克奇.木质材料表面缺陷计算机视觉测量的识别方法[J],林业科学,1996,32(1)：92-96.
42.苏建,崔乃斌,刘玉梅等.汽车整车尺寸视觉测量系统的研究[J],公路交通科技,2007,24(4)：145-149.
43.李新华,宋承祥,刘弘.双目视觉测量在乒乓球运动速度测量中的应用[J],计算机科学,2008,35(3)：256-257.
44. Yue Hong, Li Kai, Zhao Haiwen. Vision-based pipeline girth-welding robot and image processing of weld seam[J], Industrial Robot,2009,36(3):284-289.
45.刘鹏,王滨,张文明等.基于视觉传感器的焊缝图像采集与分析[J],现代焊接,2008,71：13-15.
46.邹怡容,吴哲明,曾凯.基于焊缝纹理特征的视觉识别方法[J],焊接学报,2009,30(5)：13-16.
47. Tokuhiro A T, Vadakattu S. Inspecting a research reactor's control rod surface for pitting using a machine vision[J],2005,42(11):994-1000.
48. Changhyun Park, Seho Choi, Sangchul Won. Vision-based inspection for periodic defects in steel wire rod production[J], Optical Engineering,2010,49(1):202-212.
49. Jong-Wook Kim, Jong Pil Yun, Sang Woo Kim, etal.Automatic detection of cracks in raw steel block using Gabor filter optimized by univariate dynamic encoding algorithm for searches (uDEAS)[J], NDT&E International,2009,42(5):389-397.
50. Jong Pil Yun, SungHoo Choi, Boyeul Seo, etal. Visual surface inspection of scale-covered steel billets[C], Proceedings of the Tenth IASTED International Conference on Computer Graphics and Imaging,2008,64-69
51.吴晓鹏.林介邦,唐辉等.基于机器视觉的铸坯表面缺陷检测系统的研制[J].武钢技术,2010,48(1)：49-53.
52. Tang Bo, Kong Jian-Yi, Wang Xing-Dong. Surface inspection system of steel strip based on machine vision[C]. Proceedings of 2009 1st International Workshop on Database Technology and Applications,2009,259-262.
53. Fan Zhansuo, Tang bing, Hao Junyu. Application of cut-to-length by using a pickup-camera to capture images and measure the length in the re-cutting procedure of steel billet[J]. Control & Automation,2005,1:21-22.
54. Tao Liu. Yi Zheng; Zong-yi Wang. etal. Dynamic image processing for visual robot cutting steel plate[C],2008 Second International Symposium on Intelligent Information Technology Application,2008,716-720.
55. Cheng Peng-fei, Yang Fan, Yan Li-qiang, etal. Automatic counting system of steel rods based on machine vision[J], Metallurgical Industry Automation,2008,32(1):22-26.
56. Qi Sui-ping, Zhang Hong-jian, Li Xiu-juan, etal. Counting steel rods online using LQV neural network in real-time images[C], Proceedings of the 2006 IEEE International Conference on Information Acquisition,2006,956-960.
57.李俊飞.一种棒材计数装置的研究[J],计算机测量与控制,2004,12(12)：1217-1218.
58. Sarychev A F, Galibuzov I P, Epp P A. Proceddings int workshop om industrial applications of machine intelligence & vision[J], Metallurgist,1989,32(7-8):236.
59. Graham K J, Krishnapisharody K, Irons G A, etal. Ladle eye area measurement using multivariate image analysis[J], Canadian Metallurgical Quarterly,2007,46(4):397-406.
60.杨庆军,王绍纯.计算机单目视觉检测连铸结晶器熔钢液面高度[J],北京科技大学学报,1999,21(5)：498-501.
61. Pfob C, Niel K S, Roessler R. Flushing analysis by machine vision and fuzzy logic at molten steel for the automation process[C], Proceedings of the SPIE-The International Society for Optical Engineering,2006,81-85.
62. LI Dingyi, WANG Yongran. CAO Yongguo, etal. Physico-mathematical model of heat transfer for molten steel temperature in continuous casting [J], Iron & Steel,1996,11: 17-21.
63. So R M C, Sommer T P. Explicit algebraic heat-flux model for the temperature field[J], International Journal of Heat and Mass Transfer,1996,39(3):455-465.
64. Guo Zhipeng, Xiong Shoumei, Cho Sang-Hyun, etal. Development of an inverse heat transfer model and its application in the prediction of the interfacial heat flux[J], Acta Metallurgica Sinica,2007,43(6):607-611.
65. Mikrovas AC, Argyropoulos S A, Sommerville I D. Heat transfer characteristics of molten slags [J], Ironmaking and Steelmaking,1991,18(3):169-181.
66. Solhed Henrik, Jonsson Lage. An investigation of slag floatation and entrapment in a continuous-casting tundish using fluid-flow simulations, sampling and physical metallurgy [J]. Scandinavian Journal of Metallurgy,2003,32(1):15-32.
67.陶文铨.传热学[M],西安：西北工业大学,2006：24-105.
68.徐平坤,董应榜.刚玉耐火材料[M],北京：冶金工业出版社,1999,263-292.
69.王诚训,孙炜明,张义先等.钢包用耐火材料[M],北京：冶金工业出版社：2003,31-51.
70.朱立光,张玉文,张淑会.钢水覆盖剂传热及熔融结构的数学模型[J],炼钢,1999,4(2)：29-32.
71. Lee Jye-Long, Chen J K. Pan YT, etal. Evaluation of transformation latent heat in C-Mn steels[J], ISIJ International,1999,36(3):281-287.
72. Shiina Y. Inagaki T.Study on the efficiency of effective thermal conductivities on melting characteristics of latent heat storage capsules[J], International Journal of Heat and Mass Transfer,2005,48(2):373-383.
73.周建兴,刘瑞祥.凝固过程数值模拟中的潜热处理方法[J],铸造,2001(07)：34-39.
74.谢植,次英,孟红记等.基于在线黑体空腔理论的钢水连续测温传感器的研制[J],仪器仪表学报,2005,26(5)：446-448.
75.谢植,高魁明.工业辐射温度测量[M],沈阳：东北大学出版社,1993,113-129.
76.冈萨雷斯.数字图像处理Matlab版[M],北京：电子工业出版社,2005.
77.罗希平,田捷,诸葛婴.图像分割方法综述[J],模式识别与人工智能,1999,12(3)：300-312.
78. Navon E, Miller O, Averbuch A. Color Image Segmentation Based on Adaptive Local Thresholds[J].Image and Vision Computing,2005,23 (1):69-85.
79.冉玉梅,王洪国,杨玉会.基于自适应局部阈值的彩色图像分割[J],计算机与现代化,2010,173(1)：109-112.
80. Ety Navon, OferMiller, Am ir Averbuch. Color image seg-mentation based on adaptive local thresholds[J]. Imageand Vision Computing.2005,23(1):69-85.
81. PavlidisT, Liow Y Integrating region growing and edge detection[J]. IEEE Transactions on PatternAnalysis and Machine Intelligence,1990,12(3):225-233.
82. Krishnan Nallaperumal, Justin Varghese, etal. Iterative adap-tive switching median filter[C]. Proceedings of IEEE Conference on Industrial Electronics and Applications. Singapore,2006,1-6.
83.牟春洁,张国英.基于区域边界生长的图像分割方法[J],北京石油化工学院学报,2009,17(4)：8-12.
84. Ikonomatakis N. Plataniotis K N, Zervakis M, etal. Region growing and region merging image segmentation[J].Digital Signal Proeessing Proeeedings,1997, (1),1299-1302.
85.陈小光,封举富.基于多特征判别分析的指纹图像分割方法[J],北京大学学报,2009,(1)：1-5.
86. DORIN C, PETER M. Mean shift:A robust approach toward feature space analysis[J], IEEE Transactions on Pattern Analysis and Machine Intelligence,2002,24(5):603-619.
87.李光,王朝英,侯志强.基于K均值聚类与区域合并的彩色图像分割算法[J],计算机应用,2010,30(2)：354-358.
88. SHI J, MALIK J. Normalized cuts and image segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2000,22(8):888-905.
89. Yen J C, Chang F J, Chang S. A new criterion for automatic mutilevel thresholding[J], IEEE Transaction on image processing,1995,4(3):370-377.
90. Pikaz A, Averbuch A. Digital image thresholding based on topological stable state[J], Pattern Recognition,1996,29(5):829-843.
91. Papamarkos N, Gatos B. A new approach for multilevel threshold selection. Graphic Models and Image Processing,1994,56:357-370.
92. Saboo P, Wilkins C, Yeager J. Threshold selection using Renyi's entropy. Pattern Recognition,1997,30(1):71-84.
93. Brink A D. Thresholding of digital image using two-dimensional entropies[J]. Pattern Recognition,1992,25(8):803-808.
94. Chen H D, Chen J R, Li J G. Theshold selection based on fuzzy C-partition entropy approach. Pattern Recognition,1998,31(7):857-870.
95.蔡梅艳,吴庆宪,姜长生.采用进化规划和最大熵的目标图像分割算法[J],光电与控制.2008,15(4)：13-16.
96. de Albuquerque M P, Esquef I A, de Albuquerque M P. Image segmentation using non-extensive relative entropy[J], IEEE Latin America Transactions,2008,6(5): 477-483.
97. Mayszko Dariusz, Stepaniuk Jarosaw. Adaptive rough entropy clustering algorithms in image segmentation[J]. Fundamenta Informaticae,2010,98(2):199-213.
98.齐敏,邓林.弋丹等.灰关联空间中基于最大熵阈值的医学彩色图像分割算法[J],中国科学技术大学学报.2007,37(12)：1543-1545.
99. Nalwa V S. Binford T O. On detecting edges[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,1986,8(6):699-714.
100. Haralick R M. Digital step edges from zero crossing of second directional derivatives[J], IEEE Transactions on Pattern Analysis and Machine Intelligence,1984,6(1):58-68.
101. Kang Mu, Xu Qinggong, Wang Baoshu. A Roberts' adaptive edge detection method[J], Journal of Xi'an Jiaotong University,2008,42(10):1240-1244.
102. Alam M S, Iftekharuddin K M, Karim M A. Polarization-encoded optical shadow casting: Edge detection using Roberts operator[J], Microwave and Optical Technology Letters, 1993,6(3):190-193.
103.杨振亚,王勇,王成道.LOG算子边缘检测方法的改进方案[J],计算机应用与软件,2004,21(9)：87-90.
104. Sheng-Chieh Huang, Liang-Gee Chen, Hao-Chieh Chang. A novel image compression algorithm by using Log-Exp transform[J], Proceedings of the 1999 IEEE International Symposium on Circuits and Systems VLSI,1999,17-20.
105. Hoi Steven C H, Lyu Michael R, Jin Rong. A unified log-based relevance feedback scheme for image retrieval [J], IEEE Transactions on Knowledge and Data Engineering, 2006,18(4):509-523.
106. Huiling Zhang, Baichuan Lu, Yong Yu. The method of Ostu used in the video detection of vehicles[C],2008 International Conference on Intelligent Computation Technology and Automation,2008,547-551.
107. Jiang Hui-yan, Cui Xiao-liang, Zhou Xiao-jie, etal. Image segmentation based on improved dual fast marching method[J], Journal of System Simulation,2008,20(3): 803-806.
108.韩建勋,饶欣.复杂系统建模——高维特征空间变量法[J],自动化学报,1991,17(2)：160-165.
109.郭文忠,陈国龙,陈水利等.基于粒子群优化的分类规则挖掘方法及其应用[J],集美大学学报,2008；3：34-38.
110.李峻金,向阳,芦英明等.粒子群聚类算法综述[J],计算机应用研究,2009,26(12)：4423-4427.
111. Frigu I H, Krishnaram R. A robust competitive clustering algorithm with application in computer vision[J]. IEEE Transaction on Pattern Analysis and Machine Intelligence, 1999.21(5):450-465.
112.高尚,杨静宇.群智能算法及其应用[M],北京：中国水利水电出版社,2006：79-81.
113.吴琳丽,赵海娜,汪涛等.新的全局-局部最优最小值粒子群优化算法[J],计算机应用,2009,29(12)：3270-3272.
114.郑春颖,郑全弟,王晓丹等.基于试探的变步长自适应粒子群算法[J],计算机科学,2009,36(11)：193-195.
115.赵峙江,张田文,张志宏.一种基于视觉模型与连通域统计的闽值分割新算法[J],电子学报,2005,33(5)：793-797.
116. BieniekA, Moga A. An efficientwatershed algorithm based on connect components[J]. Pattern Recognition,2000,33(6):907-916.
117.张恒,胡文龙,丁赤飙.基于快速连通域分析的目标特征提取算法[J],计算机工程与应用,2009,45(29)：230-232.
118. Lifeng He, Yuyan Chao, Suzuki K. A run-based two-scan labeling algorithm[J], IEEE Transaction on image processing,2008,17(5):131-142.
119.宋斌.一种新的图像连通域快速标号算法[J],电子测量技术,2009,32(9)：67-68.
120.徐正光,鲍东来,张利欣.基于递归的二值图像连通域像素标记算法[J],计算机工程,2006,32(24)：186-188.
121.管业鹏,童林夙等.基于数码相机的三维立体空间几何位置的摄影测量[J],电子学报,2002,30(6)：849-852.
122. Cheng Yumin. Advances and review on moving least-square methods[J], Computer Aided Engineering,2009,18(2):5-11.
123. Hu Juan-li, Deng Jia-bin, Hu Chang. An algorithm for circle curve fitting based on the constrained least square model represented by mosaic observation points[C], Proceedings of the 2009 5th International Conference on Wireless Communications, Networking and Mobile Computing,2009,4.
124. Melter R A, Stojmenovic I, Zunic J. A new characterization of digital lines by least square fits[J], Pattern Recognition Letters,1993,14(2):83-88.
125. Bo Chang, Kanefsky M. Adaptive average filters for image processing[C], Twenty-Sixth Annual Allerton Conference on Communication, Control, and Computing,1998, 459-460.
126. Ray B K. Ray K S. Corner detection using iterative Gaussian smoothing with constant window size[J]. Pattern Recognition,1995.28(11):1765-1781.
127. Tay Meng Keat Christopher, Laugier Christian. Modelling smooth paths using Gaussian processes[C], Field and Service Robotics:Results of the 6th International Conference, 2008,381-390.
128. Luo Hai-Bo, Shi Ze-Lin, Hui Ying, etal. Real-time large window -sized 2D median filter based on multi-Phased grouping and sorting network[J], Infrared and Laser Engineering, 2008,37(5):935-939.
129. Wang X, Wang D. On the max/median filter[J], IEEE Transactions on Acoustics, Speech and Signal Processing,1990,38(8):1473-1475.
130.周俊峰.高精度薄板带板凸度激光检测的误差分析与精度控制研究[D],长沙：中南大学,2006.
131.廖士中,董明,高培焕.工业摄像机几何畸变校正控制点的自动生成[J],计算机应用研究,2001,(10)：30-32.
132. Fusiello A E, Trucco A V. Rectification with unconstrained stereo geometry[C], Proceedings of the British Machine Vision Conference, Colchester,1997:400-409.
133. Fusiello A E, Trucco A V. A compact algorithm for rectification of stereo pairs [J], Machine Vision and Applications,2000,22(12):16-22.
134. Cooke E. Rectification[C], An Overview, HHI Report'99. Berlin,1999,201-202.
135.郭永刚,葛庆平,冯平等.一种基于控制点自动提取的图像畸变校正算法[J],计算机工程与应用,2007,43(2)：214-216.
136. Shuang K, Dong S P. A progressive approach method for improving the accuracy of binocular vision measurement[J], Opto-Electronic Engineering,2000,27(3):30-33.
137. Ouellet J N, Hebert P. Developing assistant tools for geometric camera calibration assessing the quality of input images[J]. Pattern Recognition,2004.4:80-83.
138. Yang C H, Lai S H, etal. Reliable image matching via modified Hausdorff distance with normalized gradient consistency measure[C], Proceedings of 3rd International Conference on Information Technology:Research and Education, Taiwan,2005: 158-161.
139. Manzar M A, Cheema T A,Jalil A, etal. New image matching technique based on hyper-vectorisation of grey level sliced binary image[J], Image Processing,2008,2(6): 337-351.
140. Shieh Hsin-Jang, Huang Po-Kai. Trajectory tracking of piezoelectric positioning stages using a dynamic sliding-mode control[J], IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control,2006,53(10):1872-1882.
141. Zhaohua Hu, Xin Fan, Yaoliang Song, etal. Joint trajectory tracking and recognition based on bi-directional nonlinear learning[J], Image and Vision Computing,2009,27(9): 1302-1312.