DESCRIPTION

Field of technology

Temperature measurement; Flow visualization; Image processing Technical problem

This invention relates to an apparatus and a method for non-contact temperature measurement of high-temperature objects and flows which emit visible light due to the thermal radiation. Technical problem addressed is failure to measure surface temperature fields of said objects and flows with a good dynamical response and at a high spatial resolution. Said technological problem is encountered in a multitude of industrial processes, including, but not limited to the mineral wool production and hot rolling of metals.

State of the art

Precise temperature measurement is often required for the purpose of process monitoring and control. Hot objects and flows moving at high velocities can exhibit large spatial and temporal variations of surface temperature, consequently requiring instantaneous temperature field measurement with a good spatial resolution and dynamic response, and possibly a high sampling rate as well.

Non-contact, nonintrusive optical methods such as the infrared imaging are most commonly employed to fulfill said temperature measurement requirements. Infrared cameras allow for accurate instantaneous temperature field measurements of stationary and low velocity processes. However, temperature field measurement is problematic in processes with fast moving objects or flows where available image acquisition rate of infrared cameras becomes insufficient, causing excessive temporal averaging of acquired temperature fields. The image acquisition rate can be improved by reduction of the acquisition window size, but the spatial resolution of computed temperature fields is reduced as a result.

The present invention addresses the problem of non-contact temperature field measurements by employing visible light cameras. The term "visible light camera" refers to a camera operating in (but not necessarily limited to) the wavelength range of the light spectrum which is visible to the human eye, this is between 390nm and 700nm. The main advantages of temperature field measurements by employing visible light cameras instead of infrared cameras are superior image acquisition rates and spatial resolution as well as significantly lower cost of the imaging setup. Temperature measurements by a visible light camera are possible above approximately 600 °C when the signal-to-noise ratio of emitted light is sufficiently high for detection by the camera sensor.

Detailed description of new invention

Apparatus and method for non-contact temperature measurement which solves above referenced technical problem comprises of the following elements:

1. At least one digital camera sensitive to visible light, this is light with a wavelength between 390nm and 700nm. Said camera can be a monochromatic camera recording grayscale images, or a color camera recording color images which are then transformed to the grayscale format.

2. A computer capable of acquiring images from the camera continuously and in real time.

3. A numerical computer-aided algorithm for real-time conversion of grayscale images obtained from said cameras to scalar fields of temperature. Said algorithm is an integral part of this invention and will be presented in this section.

A hot object or flow is assumed to radiate electromagnetic waves as a gray body, this is a body with emissivity independent of the wavelength of emitted light. Radiant exitance of a body with emissivity ε is defined by Stefan-Boltzmann law - Eq. (1 ). j = εσ ₃ Τ ⁴ a _s = 5.67 \0-*Wm ^{~ l} K- (1 )

Camera sensor illuminance Ev (Eq. (2)) is proportional to the product of radiating body light efficacy (Γ)), and blackbody radiant exitance ( σ ₅ Τ ⁴ ):

E _v = C ^ (fT ⁴ (2)

Calibration constant C is a function of distance between the camera sensor and the focal point, of the emissivity of observed hot surface, and of the sensor exposure time. Said constant is determined by calibration to a body with a known temperature.

Light efficacy η (Eq. (3)) is defined as a ratio between luminous flux and total radiation power in the wavelength range 0 < λ <∞.

/z = 6.626· l(T ³⁴ / s ; k _B = 1.381 W ²ⁱ J I K ; c = 2.998· 10 ⁸ m/s

In Eq. (3) h is the Planck constant, ke is the Boltzmann constant, c is the speed of light, Β _λ is the spectral radiance and Υ(λ) is the normalized quantum efficiency of the camera (0≤ Υ(λ)≤ 1 ). Temperature dependence function η( ~Γ) can be obtained from Eq. (3) by computation over the temperature range of interest.

Aside from the formulation in Eq. (2), camera sensor illuminance E _v can also be defined in relation to the sensor output signal (normalized gray level G) - Eq. (4). Normalized gray level of each pixel in every recorded image is defined on an interval 0 ≤ G ≤ 1 where 0 corresponds to black color and 1 to white color. Depending on the bit depth n of the image, the interval of G is uniformly divided to 2" gray levels.

In Eq. (4), ts is the camera sensor exposure time, and k is the sensor sensitivity with unit [lux ^'1 -s ^~1 ] defined as the response in the normalized gray level for a given change in camera sensor illuminance.

By combining Eq. (2) and Eq. (4), the relation between the normalized gray level and the absolute temperature is obtained in Eq. (6) which must be solved iteratively due to the temperature dependent value of η.

Procedure for determining the scalar temperature fields of a process includes the following steps:

1 . Recording of the process by a camera and image transfer to a computer. The exposure time must be set properly to avoid image saturation and blurring.

2. Temperature calibration by determination of calibration constant C. C is calculated from Eq. (5) by using temperature and corresponding image gray level data of a surface with known temperature and emissivity representative of the observed process. If the camera positioning, lens type, exposure time or the emissivity of observed objects or flows changes, temperature calibration must be repeated.

3. Iterative calculation of the absolute temperature by Eq. (5) and the function η( Τ). The function η{ Τ) defines the temperature dependence of the camera sensor light efficacy and must be calculated from the sensitivity and quantum efficiency data of the particular camera.

The invention will now be described with reference to a sample embodiment illustrated in the following figures, wherein:

Fig. 1 shows a schematic drawing of the apparatus and method for non-contact temperature measurement with a visible light camera. Shown is an example of application of said apparatus and method for the monitoring of the temperature fields on the first wheel of a mineral wool spinner.

Fig. 2: A typical graph for conversion between the normalized gray level of a process image, recorded by a visible light camera, and the absolute temperature.

Said apparatus and method for non-contact temperature field measurement with the particular application to the mineral wool production process comprises of a high-speed, visible light digital camera (6) connected to a computer (7) where continuous acquisition, processing and saving of imaging data occurs. Said process begins with a stream of mineral melt (4) flowing from the reservoir (3) onto the first spinner wheel (1 ) where a thin melt film (2) is formed and scattered to mineral wool fibers (5) due to the centrifugal force.

Relation between the normalized gray level of process images, recorded by said digital camera, and the absolute temperature varies with the measurement setup (camera positioning, lens type and exposure time) and the emissivity of observed hot objects or flows. A typical graph (8) for said relation between the normalized gray level and the absolute temperature is shown in Fig. 2.