1. Contactless measurement of charge temperature in heating furnaces using digital photopyrometry Introduction Temperature is a basic parameter of engineering processes, during which heat exchange occurs between the environment the material under consideration. The exact knowledge and control of the value of metal surface temperature during the heating-up process provides the capability to carry out the process properly. The surface temperature of charge being heated has a direct effect on heat consumption, steel loss for scale, CO2 emission, and scale adhesion to the substrate. The correct progress of the technological process is made possible by selecting the appropriate heating technology. The selection of the appropriate heating technology allows the proper progress and quality of the technological process being conducted. The final result of a correctly completed process is a charge with permissible temperature differences on its cross-section. It can be stated with certainty that the outcome of heating furnace operation is determined by the heating technology, and for a given technology - by the capacity. The technology is understood as the variation of heated charge surface temperature in time. In actual heating processes, it is the furnace chamber temperature that is regarded as the temperature of the charge being heated. The furnace chamber temperature is measured using a thermocouple. It is a basic parameter that controls the operation of the furnace. It defines also the value of the flux of heat taken up by the charge. For this parameter is considered to be credible, there should be a precise relationship between oven temperature and the actual surface temperature of heating charge. However, the furnace temperature also depends on many factors, including: gas temperature, heat transfer conditions, chamber geometry, etc. Therefore, the temperature of the furnace chamber, although it is a fundamental parameter determining the heating processes is not sufficiently objective for the proper control of the process. This is the reason for unceasing theoretical analyses and laboratory tests on the possibility of taking a direct measurement of charge temperature during the charge heating process. Using contact methods for the direct measurement of charge temperature during the heating process seems to be infeasible. This is so because of the need for a measuring device to come into direct contact with the material, whose temperature is to be measured. It is therefore of purpose to search for the solution among the contactless methods. Often, a thermovision camera is used for this purpose. The principle of thermovision relies on the recording of the thermal radiation of material being examined in the infrared radiation range. Electromagnetic radiation of wavelengths between visible light and radio waves is recorded. The advantage of thermovision is that it enables recording to be done at a temperature as low as several kelvins. A major limitation on the use of thermovision is the high price of the device (a thermovision camera). Thermovision measurements may also be burdened with a measurement error resulting from an incorrectly selected emissivity. A method that could become an alternative to thermovision in the future is digital photopyrometry. Laboratory tests are being currently conducted into the possibility of using digital photopyrometry for measuring the temperature of heated steel charge. The digital photopyrometry is also a contactless, partially automated method which is aided by a suitable computer program. However, its principle of operation relies on the recording of the image of material being heated with a still digital camera. Recording of the image is done within the wavelength range of visible light. Testing methodology The basis of measurement in digital photopyrometry is the recording of the image of a radiating sample with its simultaneous storing in the camera’s memory. The core to this measurement is the recording of the spectral emission density, ecλ. The digital camera has a CCD converter, instead of the traditional photographic plate, and a processor that processes the data and enables them to be stored in the camera’s memory. Data are recorded in the RGB (Red, Green, Blue) mode as a 24-bit colour image. Then, using an appropriate computer program, the image can be transformed into an 8-bit multi-shade bitmap that can be red out in the greyscale. 256 greyscale levels (from 0-black to 255-white) are recorded. This transformation enables the temperature of any object to be represented as a function of the absolute greyscale level. The greyscale level is determined by the computerized analysis of digital photographs using suitable graphic software. The greyscale level determination is a basis for developing a temperature characteristics. Such a characteristics will define the dependence of the surface temperature of the heat radiation-emitting object being photographed on the average greyscale level of the photograph showing the object under examination. Recently, the temperature read-out process has been automated by developing a relevant computer program. This program automatically transforms the colour image of material examined into a greyscale image. Moreover, it enables a quick temperature readout by indicating the coordinates of the point (location), at which temperature is being measured, with the computer mouse. The greyscale level is then automatically read out and subsequently converted into the value of temperature. An example digital photo of an examined sample is shown in Fig.1. Fig. 1. Digital photograph of an examined sample at a temperature of 1150  C. The measuring stand Tests were carried out on a steel sample. The specimen was heated up directly in the chamber of an electric-gas oven. The reference sample temperature was measured using an NiCr-Ni thermocouple and red out under stationary heat flow conditions, with the simultaneous photographic recording of the image of the specimen being examined. The distance of the camera from the test specimen was approx. 1.5m. A digital reflex camera was used for the tests. The technical specification of the camera is as follows: ISO 80; S= 1/60 s, 1/125 s, 1/250 s; A=4. For comparison, the recording of test specimen temperature was also done using a thermovision camera. The distance of the thermovision camera from the test specimen was approx. 1.5m. The emissivity, on the other hand, was measured by adjusting the temperature measured with the camera to the temperature value indicated by the thermocouple. The measured value of ε was ε=0.9. The temperature, as measured with the still camera, was red out by two methods. The first method was by manual reading out the average greyscale level of for the assumed surface. The value of the greyscale level readout was then converted into the temperature value. The temperature was also red out using the authors’ computational program. Using the still digital camera, attempts were also made to measure the temperatures of materials, such as graphite and chamotte. On this basis, temperature characteristics were drawn up. Kamila Hałaczkiewicz, Marian Kieloch, Agnieszka Klos Fig. 2. Schematic of the measuring stand Determination of the form of temperature characteristics The effect of temperature on the greyscale level of the digital photograph The effect of temperature on the greyscale level of the digital photo is illustrated in Figure 3. This dependence was determined for three exposure time settings, namely S=1/60s, S=1/125s and S=1/250s. Fig. 3. Effect of temperature on the greyscale level of a digital photo for carbon steel; ISO 80, A=4 ISO 80, A=4 It has been established that the effect of temperature on the greyscale level can be described with the following equation: where: a, b, c, d – constants, r – greyscale level, t – temperature,  C. On the basis of measurement results, e.g. for S=1/60s, the following relationship is obtained: The calculation of the value of temperature from the developed relationship is difficult. Therefore, relationship (1) is represented in the following form: The effect of the greyscale level of the object examined on the value of temperature is represented graphically in Fig. 4. Fig. 4. Effect of the exposure time on the behaviour of the relationship of t = f(r); ISO =80, A = 4 The values of the constants in relationship (3) have been established by using the results of measurements carried out for developing the correlation between the temperature value and the greyscale level. Thus, relationship (3) for S=1/60s is obtained in the form, as below: In the identical manner, the relationship under consideration can be developed for other conditions of conducting measurements. Measurement results and their analysis Comparison of temperature measurement results obtained by three methods Based on the measurements carried out, comparison was made of temperature values obtained using the thermovision camera and the still digital camera with the temperature values obtained with a thermocouple. The measurement results obtained by the three different methods are shown in Fig. 7. The temperature values obtained using the digital camera much less differ from the reference temperatures compared with the temperature values measured with the thermovision camera. In the case of the still digital camera, a larger error is observable in lower temperature ranges. By contrast, for the thermovision camera, the error increases with the increase in the temperature of the sample. This error is caused by increasing emissivity of the object with the increase in temperature. Comparison of the accuracy of red out temperatures The direct result of the measurement is a digital photograph of the object examined. The results of reading out the surface temperature by the manual method (t1) and by the automatic method (t2), and the results of measurements of the reference temperature (t0) are illustrated in Fig. 8. Fig. 7. Results of temperature measurements done by three methods. Fig. 8. Comparison of the accuracy of red out temperatures. It can be found from the imaged analysis of the measurement results that the temperatures red out from the digital photographs only little deviate from the model temperatures, which is indicative of high measuring accuracy of the method used, which is digital photopyrometry. An attempt to automate digital photopyrometry by the development of a suitable computer program will enable a faster temperature readout and enhance the accuracy of measurements. Temperature characteristics for different solid materials To compare the forms of temperature characteristics, temperature measurements were taken for different materials. The results of measurements for chamotte, steel, graphite and chamotte L6 are shown in Fig. 9. Fig. 9. Temperature characteristics for several solid materials Summary The accurate measurement of temperature is necessary for the process to be run correctly. Therefore, it seems appropriate to search for new methods and possibilities for measuring this parameter. Digital photography may become one of such methods in the future, as indicated by the results of the laboratory tests presented in this paper. From the tests carried out, the following conclusions can be drawn:  I t has been established that the dependence of the temperature value on the greyscale level of the digital photograph is most accurately described by the equation:  T he form of the calibration curve depends on the sensitivity of the CCD matrix, the exposure, as controlled by the shutter opening time and the lens diaphragm, the filter used, and the type of material examined.  C omparison of the measurement results obtained using the thermovision camera with the results obtained with the still digital camera shows that digital photopyrometry yields results similar to those provided by thermovision in terms of accuracy. The use of a still digital camera for temperature measurement is easier and requires less financial outlays compared to thermovision.  T he analysis of measurement results shows that temperature measurement using a still digital camera are taken with high accuracy. In addition, the use of a suitable computer program enables a considerable automation of the process and facilitates the readout to be taken, while enhancing the accuracy of the temperature measurement. By using the program, the punctual readout of steel surface temperature is possible.