Image processing method for thermal control of the lined objects
/ 5
- Details
- Category: Information technologies, systems analysis and administration
- Last Updated on 14 January 2015
- Published on 14 January 2015
- Hits: 4326
Authors:
V.A. Yemelyanov, Cand. Sci. (Tech.), Sevastopol Banking Institute of the Banking University of the National Bank of Ukraine, SBI BU NBU, Senior Instructor of the Information Technologies and Systems Department, Sevastopol, Crimea
Abstract:
Purpose. To develop the thermograms processing method of the lining objects to determine their lining burnout locations.
Methodology. The method of converting a local adaptive contrast and filters of Prewitt, Sobel, Roberts, and Canny for preprocessing thermogram images have been applied. The neural network for thermogram recognition has been used.
Findings. The main stages of the image processing method for lining objects thermal control have been described. The thermogram image processing technique of the moved mixers and wagons with liquid iron has been proposed and described. The approach for improving the thermal images quality by adaptive transform local contrast has been proposed. The approach to identification of the thermogram informative areas by filtration has been studied. The comparative results of thermogram image filtering to separate the burnout areas from the image background have been shown. The algorithm for vectorizing the thermogram images to highlight burnout areas on the filtered image has been developed. The neural networks choice to solve the problem for thermogram image recognition of the lining objects has been substantiated. The thermogram image processing results of the moved mixers and wagon with liquid iron to determine their technical condition have been described.
Originality. The thermogram image processing method of the lining objects for thermal control which based on a combination of the neural networks and classical image processing methods and which allows diagnosing the lining objects condition (determining burnout areas) has been developed.
Practical value. The practical value of these results is that the provisions of this scientific work allowed carrying out technical diagnostics of the lining objects by determining their lining burnout areas.
References:
1. Živčák, J., Hudák, R., Madarász, L. and Rudas, I.J. (2013), Methodology, Models and Algorithms in Thermographic Diagnostics, Springer.
2. Gerasimos, R. (2010), Intelligent Industrial Systems: Modeling, Automation and Adaptive Behavior, Industrial Systems Institute & National Technical University of Athens, Greece.
3. Czichos, H. (2013), Handbook of Technical Diagnostics, Springer.
4. Головко В. Мониторить „здоровье“ футеровки конвертеров будут лазерные сканеры. / В. Головко // Металлург. – 2011. – №34. – С. 2–3.
Golovko, V. (2011), “Laser scanners will monitor the “health” of converter lining”, Metallurg, Vol. 34, pp. 2–3.
5. Модернизация и комплексное оснащение современным оборудованием предприятий металлургии / Г.С. Суков, Ю.Н. Белобров, Н.Н. Попов, В.А. Дзержинский // Металлургия: Тенденции развития. – 2008. –№3. – С.4–7.
Sukov, G.S., Belobrov, Iu.N., Popov, N.N. and Dzerginskii, V.A. (2008), “Modernization and integrated by modern equipment in metallurgy”, Metallurgiya: Tendentsyi Razvitiya, Vol. 3, pp. 4–7.
6. Petrou, M. (2010), Image Processing: The Fundamentals, Wiley.
7. Gonzalez, R.S. and Woods, R.E. (2002), Digital Image Processing, Prentice, USA.
8. Емельянов В.А. Моделирование нейронных сетей распознавания металлографических изображений для диагностики состояния сталей / В.А. Емельянов // Электротехнические и компьютерные системы – 2013. – №12(88) – С. 125–131.
Yemelyanov, V.A. (2013), “Neural networks modeling for metallographic image recognition to diagnose steels condition”, Electrotekhnicheskiye i Kompyuternye Sistemy, Vol. 12(88), pp. 125–131.
9. Haykin, S. (2008), Neural Networks and Learning Machines, (3rd Edition), Prentice Hall..
 2014_6_yemelyanov
|
|
 2015-01-13  443.63 KB  986 |
|
Archive