DE102014108971A1 - Calibration method and correction method for a shutterless infrared camera system and the like - Google Patents

Abstract

The invention, which relates to both a calibration and a correction method for a shutterless infrared camera system, the object of the invention to provide a method for correcting recorded infrared images (IR images) available, which does not require mechanical aids and both for infrared camera systems with temperature-stabilized as well as with temperature-stabilized sensor can be used, as well as a method by means of which also infrared camera systems with high sensor resolutions can be calibrated with significantly reduced expenditure of time. The object is achieved by determining a measurement uncertainty representing a disturbance component of the supplied signal influenced by the ambient temperature as a function of a sensor temperature and / or at least one camera temperature under defined conditions, wherein the influences of the ambient temperature are determined by correction functions as a function of the sensor temperature and / or or the camera temperature are estimated and their camera system-specific correction coefficients are stored in a memory unit of the signal processing unit for later correction of an infrared image in a correction process.

Description

The invention relates to a calibration method for a shutterless infrared camera system comprising a housing, an infrared-transmissive optics, called a sensor matrix in the following sensor, consisting of a plurality of sensor elements which detect a radiation incident in the housing and in response to each deliver a signal to a signal processing unit and respectively have an offset and a sensitivity, wherein for each sensor element, an offset correction in dependence on an adjustable in a temperature cabinet ambient temperature and controllable radiation sources is performed.

The invention further relates to a correction method for a shutterless infrared camera system comprising a housing, an infrared-transmissive optics, called a sensor matrix in the following sensor, consisting of a plurality of sensor elements which detect a radiation incident in the housing and in response to each deliver a signal to a signal processing unit, wherein for each sensor element, an offset correction is performed by a correction calculation with camera system-specific correction parameters.

The invention further relates to an infrared camera system, comprising a housing, an infrared-transmissive optical system, a sensor matrix comprising a plurality of sensor elements and a digital signal processing unit, which uses the calibration method and the correction method.

Surface temperatures and their distribution are determined by infrared cameras. Mostly uncooled microbolometers in the form of sensor matrices are used for this purpose. A sensor matrix consists of individual microbolometer sensors arranged in a large number in rows i and columns j. The individual sensors of the sensor matrix are also called sensor element or pixel, each sensor element being uniquely assigned to a position ij of the sensor matrix. Incident radiation is absorbed and the resulting temperature change causes a change in the electrical resistance within the sensor material. Cyclically, a constant electric current is sent through the individual sensors, whereby the change in resistance results in a variable signal voltage. This is further processed by digital signal processing steps, wherein finally an infrared image in the form of an infrared image (IR image) of the surface temperature of an object to be examined or a measurement scene can be mapped. In this case, a calibration of the infrared camera or the infrared camera system (IR camera) is used to calculate a radiation-proportional output signal (IR image) from a measurement signal (raw image), taking into account all the technical-physical properties of the IR camera.

The output signal voltage of a sensor element is proportional to the exchanged radiation flux. The radiation-voltage characteristics of the individual sensor elements of a sensor matrix vary within certain limits.

Infrared-transparent lenses focus areas of the object space on the sensor elements. Due to the large opening angle of the sensor elements, these detect both the radiation components of the observed scene and radiation components which originate from the interior of the infrared camera ( 1 ). The amount of interference radiation components depends on various factors, including the ambient temperature and the heat losses within the infrared camera. This dependence on external circumstances causes the amount of detected interference radiation to change unpredictably and influence the temperature measurement. For the lowest possible measurement uncertainty, the influence of the interference radiation on the output signal of each sensor element must be corrected.

The main characteristic parameters of a microbolometer sensor element are the sensitivity (gain) and the operating point (offset). Due to the technology, the individual sensor elements of a microbolometer array have individually different operating points (offsets) and sensitivities and thus different characteristic curves with respect to the incident radiation flow. By correcting this nonuniformity (also called NUC - Non-Uniformity Correction), all sensor elements can be converted to a uniform characteristic curve, a so-called standard characteristic curve. So far, the non-uniformity of the sensor element radiation flux voltage characteristics is corrected by a sensor element-dependent differential voltage is calculated from the average image value at homogeneous object radiation and constant ambient temperature. The coefficients of the individual linear differential voltage curves are determined by a characteristic comparison. This requires two infrared images of a homogeneous radiation source at different temperatures (object temperatures) covering the entire field of view of the infrared camera.

Microbolometer sensors can be operated with and without temperature stabilization. Since the two main characteristic parameters of the sensor elements, sensitivity and offset, depend on the sensor temperature, these are almost constant in the case of a temperature-stabilized microbolometer sensor. If the sensor is not temperature-stabilized, the characteristic parameters vary depending on the sensor temperature. Assuming that all sensor elements have the same operating temperature, the change in sensitivity and offset is the same for all sensor elements.

Due to heat conduction through the camera body, the sensor temperature is coupled to the ambient temperature. Thus, changes in the ambient temperature affect both the amount of absorbed radiation flux as well as the characteristic parameters, sensitivity and offset of the individual sensor elements. The sensor temperature is available during a measurement as an additional measured variable at the respective sensor element output. This makes it possible to correct the influence of the sensor temperature dependence on the signal voltage of each sensor element.

The proportion of interfering radiation at the respective signal voltage of a sensor element has hitherto been determined in most cases with the aid of an optical shutter (shutter), in order subsequently to be subtracted from the measurement signal and thus to correct the output signal. For example, in Budzier, H., "Radiometric Calibration of Uncooled Infrared Cameras", TUDpress, Dresden, 2014, ISBN 978-3-944331-45-4 , the shutter is closed, so that only the interior radiation is detected by each sensor element of the sensor matrix. This offset of the output voltage is then weighted via the shutter characteristic curve and subtracted from the measuring signal when the shutter is open. For this purpose, the so-called shutter characteristic is recorded, which represents the ratio of shutter-to-signal and shutter-to-signal as a function of the infrared camera temperature. If the infrared camera temperature changes by a certain amount of time or if a certain amount of time has passed, this process is repeated.

A second procedure is over Olbrycht, R. et al., "Thermal drift compensation method for microbolometer thermal cameras", Applied Optics, Vol. 51, no. 11, p. 1788ff., 2012 known. Instead of a shutter, an infrared filter is used, which is cyclically swung into the beam path of the infrared camera. The further method steps correspond to the correction method with shutter.

The two methods described above rely on additional, mechanically moving aids, shutters or infrared filters, which limit the design possibilities of infrared cameras and are mechanically susceptible.

Therefore, an infrared camera system and an associated correction method, which can dispense with the mechanical components, would be advantageous. In Bieszczad, G. et al., "Method of detectors offset correction in a thermovision camera with an uncooled microbolometer focal plane array", SPIE Proceedings, Vol. 7481, 74810O-1-8, 2009 , a method is disclosed which estimates the amount of spurious radiation based on the sensor temperature for each sensor element and subtracts it from the measured value, although only the temperature-dependent offset is considered. While the infrared camera is aimed at a homogeneous and constant radiation source, the ambient temperature is varied at intervals over the entire permissible range. Before the measured values are recorded, it is waited until the temperature distribution in the infrared camera has assumed a stationary state. For each sensor element, the compensation calculation for estimating the output voltage as a function of the sensor temperature is carried out as polynomial 2nd order. During infrared measurement, the sensor temperature is constantly available as a parameter. Although this method does not require additional aids, it is limited to infrared cameras with non-temperature-stabilized sensor. In addition, the compensation calculation of the correction line is performed for each pixel and takes correspondingly much time during the calibration of the infrared camera. At current sensor resolutions of 640 × 480 sensor elements or pixels, the time required is in the range of several hours. Another disadvantage is the greatly delayed reaction of the sensor temperature to a change in the ambient temperature. As a result, the measurement uncertainty increases significantly after a sudden change in the ambient temperature and only slowly decreases again.

It is therefore an object of the invention to provide a method for correcting recorded infrared images (IR images) available, which works without mechanical aids and can be used both for infrared camera systems with temperature-stabilized and temperature-stabilized sensor, and a method by which also infrared camera systems high sensor resolutions can be calibrated with significantly less time.

The object is achieved on the method side by determining a measurement uncertainty representing a noise component influenced by the ambient temperature of the supplied signal as a function of a sensor temperature and / or at least one camera temperature, the influences of the ambient temperature being determined by correction functions as a function of the sensor temperature and / or the camera temperature are estimated and their camera system-specific correction coefficients are stored in a memory unit of the signal processing unit for later correction of an infrared image.

In one embodiment of the calibration method, the calibration method is used for an infrared camera system with temperature-stabilized sensor. It is assumed that both the sensitivity of all sensor elements of the sensor matrix and the offset of all sensor elements of the sensor matrix are constant and if there are differences in the main characteristic parameters between the sensor elements, these vary only within certain limits.

In another embodiment of the invention, the calibration method for an infrared camera system with a temperature-stabilized sensor is used. Therefore, both the sensor temperature-dependent sensitivity change and the sensor temperature-dependent offset change of the sensor elements must be taken into account in a calibration process.

For an infrared camera system with temperature-stabilized sensor, the following calibration method is proposed: For the sensor matrix, an operating point is set, wherein an object temperature range and an ambient temperature range are specified. A camera-temperature-specific nonuniformity of all sensor elements is determined by means of a first characteristic adjustment of degree n, where n ≥ 2 and correction coefficient matrices a _{n, ij} with n = 0, 1, 2,... N stored in the memory unit, where Nonuniformity correction is performed at n + 1 different ambient temperatures and a constant, homogeneous object temperature.

Furthermore, a correction curve for the consideration of a camera interior radiation is determined, wherein the camera temperature is measured at m locations within the housing of the infrared camera system, with m ≥ 1, and determines the dependence of the signal voltage of the sensor on the camera temperature by means of a compensation calculation, in particular a polynomial approximation at least second order where the polynomial coefficients α _{K, m} , β _{K, m} , γ _{K, m} are stored in the memory unit. The number m and the position of the temperature measuring points within the housing of the infrared camera system are particularly important in order to determine the exact influence of the different camera temperatures within the housing on the signal voltage of the sensor.

Furthermore, an object temperature-specific nonuniformity of the infrared camera system is determined by means of a second characteristic adjustment of degree n, where n ≥ 1 and correction coefficient matrices b _{n, ij} with n = 0, 1, 2, ... n are stored in the memory unit, wherein the Nonuniformity correction is performed at n + 1 different object temperatures and a constant ambient temperature. The camera temperature is measured at the m points within the housing of the infrared camera system, with m ≥ 1, and serves as the reference camera temperature for the later correction procedure.

The first characteristic adjustment must be at least of degree n = 2, ie based on three interpolation points, and thus corresponds to a three-point correction. The second characteristic adjustment must be at least of the degree n = 1, that is, based on two interpolation points, and thus corresponds to a two-point correction. In order to reduce the measurement uncertainty of the corrected signal voltages of the sensor elements, both characteristic adjustments can be performed to the same degree (but at least to the degree n = 2) as a three-point correction.

After each calibration step, a determination should be made of sensor elements whose behavior lies outside defined limits. The output signals of these sensor elements are ignored for all further steps and replaced by averaged output signals of adjacent sensor elements. This leads to a lower measurement uncertainty of the corrected IR images.

The corrected signal voltages of the individual sensor elements can be assigned temperatures in a last signal processing step, a so-called radiometric adjustment. In this radiometric adjustment, a conversion rule is determined with the aid of calibrated radiation sources at different object temperatures in order to assign absolute object temperatures to the corrected sensor signal voltages. This conversion rule is stored in a memory unit of the signal processing unit.

For an infrared camera system with temperature-stabilized sensor, the following calibration method is proposed: For the sensor matrix, an operating point is set, wherein an object temperature range and an ambient temperature range are specified. A sensor temperature-dependent sensitivity change of all sensor elements is approximated with a second-order polynomial and correction coefficients g ₂ , g ₁ , g ₀ are stored in the memory unit, wherein a signal voltage difference change of at least two different constant object temperatures is measured as a function of the measured sensor temperature at different ambient temperatures ,

Furthermore, a sensor temperature-dependent offset change of all sensor elements is approximated with a third-order polynomial and correction coefficients o ₃ , o ₂ , o ₁ , o ₀ in the Stored storage unit, wherein at a constant object temperature as a function of the sensor temperature at different ambient temperatures, a signal voltage change is measured. After these calculation steps, the infrared camera system behaves this way, all the sensor would be temperature stabilized. The following calibration is similar to that of an infrared camera system with a temperature-stabilized sensor. A camera-temperature-specific nonuniformity of all sensor elements is determined by means of a first characteristic adjustment of degree n, where n ≥ 2 and correction coefficient matrices a _{n, ij} with n = 0, 1, 2,... N stored in the memory unit, where Nonuniformity correction is performed at n + 1 different ambient temperatures and a constant, homogeneous object temperature.

Furthermore, a correction curve for the consideration of a camera interior radiation is determined, wherein the camera temperature is measured at m locations within the housing of the infrared camera system, with m ≥ 1, and determines the dependence of the signal voltage of the sensor on the camera temperature by means of a compensation calculation, in particular a polynomial approximation at least second order where the polynomial coefficients α _{K, m} , β _{K, m} , γ _{K, m} are stored in the memory unit.

Furthermore, an object temperature-specific nonuniformity of the infrared camera system is determined by means of a second characteristic adjustment of degree n, where n ≥ 1 and correction coefficient matrices b _{n, ij} with n = 0, 1, 2, ... n are stored in the memory unit, wherein the Nonuniformity correction is performed at n + 1 different object temperatures and a constant ambient temperature. The camera temperature is measured at the m points within the housing of the infrared camera system, with m ≥ 1, and, together with the sensor temperature also measured, serves as the reference temperature for the later correction procedure.

The first characteristic adjustment must be at least of degree n = 2, ie based on three interpolation points, and thus corresponds to a three-point correction. The second characteristic adjustment must be at least of the degree n = 1, that is, based on two interpolation points, and thus corresponds to a two-point correction. In order to reduce the measurement uncertainty of the corrected signal voltages of the sensor elements, both characteristic adjustments can be performed to the same degree (but at least to the degree n = 2) as a three-point correction.

After each calibration step, a determination should be made of sensor elements whose behavior lies outside defined limits. The output signals of these sensor elements are ignored for all further steps and replaced by averaged output signals of adjacent sensor elements. This leads to a lower measurement uncertainty of the corrected IR images.

The corrected signal voltages of the individual sensor elements can be assigned temperatures in a last signal processing step, a so-called radiometric adjustment. In this radiometric adjustment, a conversion rule is determined with the aid of calibrated radiation sources at different object temperatures in order to assign absolute object temperatures to the corrected sensor signal voltages. This conversion rule is stored in a memory unit of the signal processing unit.

The raw data acquisition of the IR images and the acquisition of the respectively associated sensor and camera temperatures for the individual steps of the calibration process can be carried out in any order and are not decisive for the final provision of the correction coefficients or polynomial coefficients. Rather, it is important that the determined correction coefficients or polynomial coefficients for a subsequent correction of an infrared image are stored in a memory unit.

• – In einem ersten Korrekturschritt wird eine sensortemperaturabhängige Empfindlichkeitsänderung mit einem Polynom zweiten Grades korrigiert,
• – in einem zweiten Korrekturschritt wird eine sensortemperaturabhängige Offsetänderung mit einem Polynom dritten Grades korrigiert,
• – in einem dritten Korrekturschritt wird eine kameratemperaturspezifische Ungleichförmigkeit mittels eines Kurvenabgleichs vom Grad n, mit n ≥ 2 korrigiert,
• – in einem vierten Korrekturschritt wird der Einfluss einer Kamerainnenraumstrahlung entsprechend einer Korrekturkurve korrigiert,
• – in einem fünften Korrekturschritt wird eine objekttemperaturspezifische Ungleichförmigkeit des Infrarotkamerasystems mittels eines Kurvenvergleichs vom Grad n, mit n ≥ 1 korrigiert. Dabei werden die infrarotkamerasystemspezifische Korrekturkoeffizienten, welche in einem vorgeschalteten Kalibrierverfahren unter bekannten Bedingungen ermittelt wurden, für die Korrektur einer Infrarotaufnahme verwendet.

The object is also achieved in terms of the method in that the correction coefficients or polynomial coefficients determined in the calibration method are subsequently used by a correction method according to the invention, the correction method comprising the following steps:

• In a first correction step, a sensor temperature-dependent sensitivity change is corrected with a second-degree polynomial,
• In a second correction step, a sensor temperature-dependent offset change is corrected with a third-degree polynomial,
• In a third correction step, a camera-temperature-specific nonuniformity is corrected by means of a curve adjustment of the degree n, with n ≥ 2,
• In a fourth correction step, the influence of a camera interior radiation is corrected according to a correction curve,
• In a fifth correction step, an object temperature-specific nonuniformity of the infrared camera system is corrected by means of a curve comparison of the degree n, with n ≥ 1. The infrared camera system-specific correction coefficients, which were determined in a preceding calibration method under known conditions, are used for the correction of an infrared recording.

In one refinement of the correction method, the first correction step uses correction coefficients g ₂ , g ₁ , g ₀ , which are determined from a measured signal voltage difference change between at least two different constant object temperatures as a function of a measured sensor temperature θ _s .

In a further refinement of the correction method, the second correction step uses correction coefficients o ₃ , o ₂ , o ₁ , o ₀ , which are determined from a signal voltage change at a constant object temperature as a function of the sensor temperature θ _s at different ambient temperatures.

In a further refinement of the correction method, the third correction step uses correction coefficient matrices a _{n, ij} with n = 0, 1, 2,... N, which are determined at n + 1 different ambient temperatures and a constant, homogeneous object temperature.

In a further embodiment of the correction method, the fourth correction step uses polynomial coefficients α _{K, m} , β _{K, m} , γ _{K, m of} a correction curve and measured camera temperatures θ _{K, m} at m different locations within the housing of the infrared camera system, with m ≥ 1, where the polynomial coefficients from the dependence of the signal voltage of the sensor on the camera temperature θ _{K, m} at different positions m within the housing of the infrared camera system, with m ≥ 1 are determined.

In a further refinement of the correction method, the fifth correction step uses correction coefficient matrices b _{n, ij} with n = 0, 1, 2,... N, which are determined at n + 1 different object temperatures and a constant ambient temperature.

Preferably, the correction method can be extended by a radiometric adjustment. The signal voltages are converted into absolute object temperatures with the aid of a conversion rule, this conversion rule being determined in a radiometric adjustment as part of a calibration process of the infrared camera system using calibrated radiation sources at different object temperatures.

In one embodiment of the correction method according to the invention, the correction steps one to five are carried out for a temperature-unstabilized sensor, i. from the first correction step to the fifth correction step.

In another embodiment of the correction method according to the invention, only the correction steps three to five are carried out for a temperature-stabilized sensor, i. from the third correction step to the fifth correction step. The first two correction steps are not necessary because in an infrared camera system with a temperature-stabilized sensor it is assumed that the sensitivity and the offset of the sensor element remain constant with changing ambient temperature for all sensor elements.

On the arrangement side, the object is achieved by an infrared camera system, wherein at least two temperature measuring means measuring a camera temperature are arranged in the interior of the housing of the infrared camera system.

In one embodiment of the infrared camera system, the temperature measuring means is designed as a thermocouple and / or a thermal resistance.

In the following, the invention will be explained in more detail with reference to exemplary embodiments.

The accompanying drawings show

1 a schematic representation of the detected by the sensor elements radiation flows from the considered half space,

2 a schematic representation of the calibration of an infrared camera system,

3 a schematic sequence of a correction method for an infrared camera system with a temperature-stabilized sensor, wherein three camera interior temperatures are measured and both characteristic adjustments to degree n = 2 done and the matrix multiplications are performed element by element,

4 a schematic sequence of a correction method for an infrared camera system with a temperature-unstabilized sensor, wherein three camera interior temperatures are measured and both characteristic adjustments to degree n = 2 done and the matrix multiplications are performed element by element,

5 an overview for determining the correction coefficient matrices for a three-point correction, wherein the matrix multiplications are performed element-by-element,

6 an overview for determining the correction coefficient matrices for a two-point correction, wherein the matrix multiplications are performed element by element.

2 schematically shows an infrared camera system 11 in a temperature cabinet 1 , where the full field of view 10 of the infrared camera system 11 from a surface radiator 2 is completed. A movable black matte screen 3 can be pivoted into the field of view of the infrared camera system so that it completely covers the field of view of the infrared camera system. With this setup, the conditions for calibrating the infrared camera system 11 targeted and controlled.

In a first embodiment, the calibration method of an infrared camera system 11 be explained in more detail with temperature-stabilized sensor. In this case, the sensor of the infrared camera system 11 be maintained for example by a Peltier element at a constant temperature. Due to the constant sensor temperature it is assumed that the sensitivity and the offset of the individual sensor elements are constant and independent of the ambient temperature. During calibration, the infrared camera system is located in a temperature cabinet 1 whose temperature can be regulated. The temperature in the temperature cabinet 1 Sets the ambient temperature of the infrared camera system 11 and affects the camera interior temperature of the infrared camera system 11 ,

In a calibration step, the non-uniformity of the sensor elements with respect to the influence of different ambient temperatures, ie, different internal camera interference radiation components should be corrected so that all sensor elements react in the same way to changes in the camera interior interference radiation components. The infrared camera system 11 is doing on a surface spotlight 2 is directed with a homogeneous radiation surface with a constant temperature and thus a constant radiation Φ. Now at least three ambient temperatures T _U1 , T _U2 , T _U3 in the temperature cabinet 1 adjusted, the camera temperature of the infrared camera system 11 is coupled to the ambient temperature. Preferably, the temperatures in the temperature cabinet 1 adjusted to the values of the upper and lower limits of the allowable ambient temperature range of the infrared camera system 11 correspond. The remaining number of required temperatures should be evenly distributed over the ambient temperature measuring range. In the steady state, the uncorrected sensor element voltages U _{p, ij are} measured at the at least three interpolation points, ie at the set ambient temperatures T _U1 , T _U2 , T _U3 , the absolute values of the temperatures not being of interest here. From the measured signal voltages U _{p, ij of} all sensor elements at the respective interpolation points, the mean squared values U _{p, ij} ^{2 become} the mean value U _p and calculates the differences to the mean ΔU _{p, ij} . Using the calculation rule 5 For example, the correction coefficient matrices a _{n, ij are} calculated for a characteristic adjustment in the form of a three-point correction. Assuming that the deviations of the signal voltage curves of the sensor elements from a standard curve at different ambient temperatures and thus different camera internal radiation results as a quadratic relationship, a three-point correction is sufficient. In order to further reduce the measurement uncertainty, more than three sample points can be used to calculate the correction coefficient matrices a _{n, ij} . For this purpose, a plurality of sets of auxiliary correction coefficient matrices a _{n, ij} "are calculated for different combinations of three interpolation points, the outer interpolation points remaining the same and only the third required interpolation point being varied. The calculated auxiliary correction coefficient matrices of the same degree n (with n = 0, 1, 2) are then averaged to give the final correction coefficient matrices a _{n, ij} . These correction coefficient matrices a _{n, ij} with n = 0, 1, 2 are used in a memory unit of the signal processing unit for a later correction procedure for infrared acquisition under real measuring conditions 7 of the infrared camera system 11 stored in an FPGA, for example. After this first step of the calibration, the corrected infrared image is homogeneous for different ambient temperatures and the absolute value of the corrected signal voltages corresponds to the magnitude of the mean value of the uncorrected signal voltages of all the sensor elements.

In another calibration step, the influence of different ambient temperatures, ie different camera interior interference radiation components, based on the measured camera interior temperature, hereinafter referred to as the camera temperature to be corrected in terms of magnitude, the infrared camera system on a homogeneous radiating surface radiator 2 is directed at a constant temperature T _obj and where the ambient temperature in the temperature _cabinet 1 is changed according to a predetermined, but any temperature regime. While the temperature in the temperature cabinet 1 constantly changes, over time t at at least one point within the infrared camera system, the camera temperature θ _{K, 1} and the average value of all sensor elements U _p measured. The measurement takes place at different times i. Preferably, a plurality of temperature sensors are within the housing 4 of the infrared camera system arranged and the camera temperature can be measured at several points m = 1, 2, 3, ... simultaneously. It is assumed that an ambient temperature change has a different effect on the camera interior and the total interference radiation component can be regarded as a linear superposition of different radiation components, which can be estimated by temperature measurements at various points m within the camera.

The aim of this step is to correct the influence of different ambient temperatures and thus camera interior temperatures in terms of absolute value, so that the corrected absolute value of the signal voltage is independent of the ambient temperature. The signal voltage component U _{K, m} , caused by the camera interior radiation, is determined indirectly via the measurement of the camera interior temperatures θ _{K, m} at the points m and with a quadratic polynomial U _{K, m} (θ _{K, m} ) = α _{K, m} · θ _{K, m} ² + β _{K, m} · θ _{K, m} + γ _{K, m} approximated. This equation is overdetermined by the large number of recorded measured values and the associated camera temperatures at the respective measuring time i, so that the coefficients α _{K, m} , β _{K, m} , γ _{K, m} can be determined by a compensation calculation. These are also stored in a memory unit of the signal processing unit 7 of the infrared camera system, for example, stored in an FPGA for a later correction method of infrared recording under real measurement conditions.

In a further calibration step, the nonuniformity of the sensor elements with respect to the influence of different object radiations is to be corrected so that all sensor elements react in the same way to changes in the object radiation. For this, the infrared camera system is located in a temperature cabinet 1 , The ambient temperature is maintained at a constant value T _U , whereby the thereby adjusting camera temperatures θ _{K0, m are} measured and serve as reference camera temperatures for the correction process. The temperature of the surface radiator 2 to which the infrared camera system is directed is varied. In this case, at least two different Flächenstrahlertemperaturen, ie object temperatures T _{Obj, 1} , T _{Obj, 2 are} set. To reduce the measurement uncertainties, the number of object _temperatures T _{obj can correspond to} the number of set ambient temperatures T _U from the calibration _step for correcting the nonuniformity of the influence of different ambient temperatures, ie camera temperatures. In this case, the determination of the correction coefficients takes place analogously to the calculation rule shown in FIG 5 , In the case of a two-point correction, the sensor element voltages U _{p, ij} are again measured at the at least two reference points, ie at the set object temperatures, the absolute values of the temperatures also not being of interest here. From the measured signal voltages U _{p, ij} all sensor elements at the respective support points are the average U _p and calculates the differences to the mean ΔU _{p, ij} . Using the illustrated calculation rule in 6 For example, the correction coefficient matrices b _{n, ij are} calculated for a characteristic adjustment in the form of a two-point correction. Assuming that the deviations of the signal voltage curves of the sensor elements from a standard curve at different object temperatures result in a linear relationship, a two-point correction is sufficient. These correction coefficient matrices b _{n, ij} with n = 0, 1, 2 are for a later correction method of an infrared image under real measuring conditions in the memory unit of the signal processing unit 7 of the infrared camera system 11 stored in an FPGA, for example. After this step and the execution of the aforementioned calibration steps results in constant object radiation, a homogeneous, ambient temperature independent and thus also camera interior temperature independent infrared image.

Before the first and after each individual calibration step, those sensor elements whose behavior, i. E. For example, their output signal voltages are outside defined limits. The output signals of these sensor elements are ignored for all further steps and replaced by averaged output signals of adjacent sensor elements.

The radiometric calibration assigns temperatures to the corrected signal voltages of the individual sensor elements. For this purpose, the infrared camera system is aimed at calibrated radiation sources whose absolute object temperatures are known. The observed object temperatures should be evenly distributed over the permissible measuring range of the infrared camera system. The conversion rule from sensor signal voltage to absolute object temperature can be described by a function with the aid of a compensation calculation or by an assignment table (look-up table) and in a memory unit of the signal processing unit 7 of the infrared camera system 11 get saved.

In a second embodiment, the calibration of an infrared camera system with a temperature-stabilized sensor will be explained in detail. In contrast to an infrared camera system with a temperature-stabilized sensor which assumes that the sensor parameters (sensitivity and offset) are not influenced by the ambient temperature, in an infrared camera system with a temperature-stabilized sensor, the standard characteristic of the sensor parameters changes with the change of the ambient temperature. It is therefore imperative that a sensitivity and offset Perform calibration to perform the temperature-dependent correction of the standard characteristic curve.

For this, the infrared camera system is located in a temperature cabinet 1 whose temperature can be regulated. For sensitivity calibration, there are at least two homogeneous and constant area radiators 2 with different radiator temperatures (θ _obj1 , θ _obj2 with θ _obj1 ≠ θ _obj2 ). In the calibration arrangement ( 2 ) is outside the temperature cabinet 1 and in the beam path 10 between the surface radiators 2 and the infrared camera system 11 a black matte screen 3 that the object radiation of the surface radiator 2 on the infrared camera system 11 can interrupt and for the sensor 6 of the infrared camera system 11 a homogeneous surface radiator 2 represents. The infrared camera system 11 is on the two area radiators 2 directed, wherein the radiation of both surface radiators with only a small delay time, ie, seen at approximately the same time t _i from the infrared camera system, so that the camera temperature 8a . 8b before and after swiveling in or out the aperture 3 can be considered the same.

In a first step, the ambient temperature in the temperature cabinet 1 changed after a predetermined but arbitrary temperature regime and the aperture 3 repeated for a short time in the beam path 10 between the surface radiators 2 and the infrared camera system 11 swung in that the infrared camera system 11 alternately the homogeneous radiation surface of the aperture 3 and the surface radiator 2 sees. There are repeatedly recorded raw data sets of sensor element voltages U _{p, ij} , which can be distinguished with respect to their radiation source. From this data becomes the mean U _p calculated the sensor element voltages. Subsequently, the difference voltage of the first calculated mean value when looking at the black-matted aperture 3 and the second calculated mean value when looking at the area radiator 2 , calculated over the time t at different times t _i and stored, the ambient temperature T _U in the temperature cabinet 1 is varied, whereby the sensor temperature changes. As the sensor temperature rises, this difference voltage changes. This temperature dependence of the sensitivity can be described by a polynomial of second degree: G _V (θ _S ) = g _{2 S} ² + g _{1 S} + g ₀ . With the differential voltage values from the acquired measurement series, a compensation calculation is performed to determine the polynomial coefficients of the overdetermined equation G _V (θ _S ) = g _{2 S} ² + g _{1 S} + g ₀ and used for a later infrared absorption correction method real measurement conditions in the memory unit of the signal processing unit 7 of the infrared camera system 11 stored in an FPGA, for example.

The sensor temperature dependence of the sensor element offset can be described by a third degree polynomial: O _V (θ _S ) = o ₃ · θ _S ³ + o ₂ · θ _S ² + o ₁ · θ _S + o ₀ , but this change is not can be measured directly, since both the sensor temperature θ _S and the camera interior temperature θ _K change at variable ambient temperature. The influence of the changing internal camera radiation can be described as a function of θ _K with a second degree polynomial: O _K (θ _K ) = k ₂ · θ _K ² + k ₁ · θ _K + k ₀ . To determine the sensor temperature dependence of the offset, both influences on the sensor signal voltages, namely the influence of the sensor element offset which changes as a function of the sensor temperature and the influence of the camera internal radiation, which changes as a function of the camera temperature, must be taken into account.

This is the infrared camera system 11 to a homogeneous radiation source 2 directed with constant object temperature. The ambient temperatures in the temperature cabinet 1 are changed according to a given but arbitrary temperature regime. The raw data sets at sensor element voltages U _{p, ij} and the sensor temperature θ _S and the camera temperature θ _{K, m} are at least two different locations m in the interior of the housing 4 of the infrared camera system 11 measured over time t. The coefficients of the linear superposition of all influences O (θ _S , θ _{K, m} ) = O _V (θ _S ) + ΣO _{K, m} (θ _{K, m} ) are determined with a compensation calculation. Due to the large number of nodes, the equation is overdetermined. The coefficients with which the influence of the sensor temperature can be calculated by using the equation O _K (θ _K ) = k ₂ · θ _K ² + k ₁ · θ _K + k ₀ will be used for a later correction method of infrared taking real Measurement conditions in the memory unit of the signal processing unit 7 of the infrared camera system 11 stored in an FPGA, for example.

As a result of these calibration steps, it is possible to correct the influence of the sensor temperature on the sensitivity and the offset of the sensor elements of the infrared camera system, so that the corrected signal voltages behave as if the sensor is temperature-stabilized. The following steps for calibration correspond to those of an infrared camera system with a temperature-stabilized sensor, see the comments on the first exemplary embodiment of a calibration method of an infrared camera system with temperature-stabilized sensor.

In contrast to an infrared camera system with temperature-stabilized sensor, the second characteristic curve calibration of an infrared camera system with temperature-stabilized sensor also measures the sensor temperature θ _{S0 in} addition to the camera temperatures θ _{K0, m} . Together, these temperatures are used as reference temperatures for the correction process.

An exemplary embodiment of the correction method using a temperature-stabilized sensor will be given below. 3 shows the schematic sequence of the correction method according to the invention. The infrared camera system records an object scene during a temperature measurement. The uncorrected sensor element voltages of all sensor elements of the sensor element array are _transmitted to the signal processing unit as so-called raw data U _{p, ij} 7 of the infrared camera system 11 passed. For each sensor element, with the coefficient matrices a _{2, ij} a _{, ij} , a _{0, ij} corresponding to a three-point correction, a corrected sensor element voltage U ** _{p, corr, ij} for non-uniformity correction with respect to a changing sensor element voltage U ** _p Camera temperature calculated. In the calibration method, since the correction coefficient matrices a _{n, ij} where n = 0, 1, 2 were calculated by using at least three nodes, the correction of the nonuniformity in the camera internal radiation is performed as the characteristic adjustment of the degree n = 2.

The influence of the camera interior temperature, ie the internal camera radiation, is calculated by means of the correction function U _K (θ _K ) and the difference to the correction value U _K (θ _K0 ) is subtracted from the corrected sensor voltage U ** _{p, cor} at the reference camera temperature θ _K0 preset by the calibration , with U _K (θ _K ) = U _{K, 1} (θ _{K, 1} ) + U _{K, 2} (θ _{K, 2} ) + U _{K, 3} (θ _{K, 3} ) + ... + U _{K, m} (θ _{K, m} ), where for determining U _K (θ _K ) the camera temperature at at least one location within the housing 4 of the infrared camera system, preferably at two or more locations with m = 1, 2, 3,... measured and the coefficients α _{K, m} , β _{K, m} , γ _{K, m} stored from the calibration method are used in accordance with the second-order polynomial ,

Subsequently, the nonuniformity correction takes place with respect to different object temperatures, ie object radiation, wherein a corrected sensor element voltage U _{p, korr, ij is} calculated for each sensor element with the correction coefficient matrices b _{n, ij} stored from the calibration method in accordance with a two-point correction. As a result, with constant object radiation, a homogeneous, ambient temperature-independent infrared image results.

The signal voltages of the sensor elements determined during the calibration, whose output voltage signals lie outside defined limits, are ignored and replaced by the corrected signal voltages of adjacent sensor elements.

The corrected signal voltages of the sensor elements are finally converted into temperatures using the conversion rule of the calibration performed during the calibration.

4 shows a second embodiment of an infrared camera system 11 with temperature-stabilized sensor. It takes the infrared camera system 11 during a temperature measurement on an object scene. The uncorrected sensor element voltages of all sensor elements of the sensor element array are _transmitted to the signal processing unit as so-called raw data U _{p, ij} 7 of the infrared camera system 11 passed. In the first correction step, the sensor temperature-dependent sensitivity change of the sensor elements on the basis of the currently measured sensor temperature θ _S to U _{g, korr, ij} with a second degree polynomial using the determined in the calibration process coefficients g ₂ , g ₁ , g ₀ , which by the correction value in the in the calibration measured reference sensor temperature θ _S0 is corrected. Subsequently, the sensor temperature-dependent offset change is corrected by U _{o, corr, ij} = U _{g, corr, ij} + o ₃ · θ _S ³ + o ₂ · θ _S ² + o ₁ · θ _S + o ₀ - O _V (θ _S0 ), wherein the determined in the calibration method correction coefficients o ₃ , o ₂ , o ₁ , o ₀ and the reference sensor temperature θ _{S0 are} used. After this correction step, the sensor elements of the sensor matrix of the infrared camera system behave as if an infrared camera system with temperature-stabilized sensor were present. The remaining correction steps are carried out analogously to the first exemplary embodiment of the correction method using a temperature-stabilized sensor.

LIST OF REFERENCE NUMBERS

1

temperature cabinet

2

Panel radiator with object temperature

3

cover

4

casing

5

infrared-transparent lens system

6

Sensor matrix with sensor elements

7

Signal processing unit

8a, b

 Temperature measurement means

10

Viewing range of the infrared camera system

11

Infrared camera system

θ _S

sensor temperature

θ _K

Camera interior temperature

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited non-patent literature

• Budzier, H., "Radiometric Calibration of Uncooled Infrared Cameras", TUDpress, Dresden, 2014, ISBN 978-3-944331-45-4 [0010]
• Olbrycht, R. et al., "Thermal drift compensation method for microbolometer thermal cameras", Applied Optics, Vol. 51, no. 11, p. 1788ff., 2012 [0011]
• Bieszczad, G. et al., "Method of detectors offset correction in a thermovision camera with an uncooled microbolometer focal plane array", SPIE Proceedings, Vol. 7481, 74810O-1-8, 2009 [0013]

Claims (16)

Calibration method for a shutterless infrared camera system ( 11 ) comprising a housing ( 4 ), an infrared-transmissive optic ( 5 ), a sensor matrix ( 6 ) in the following sensor, consisting of several sensor elements, the one in the housing ( 4 ) detect incident radiation and, in response thereto, respectively send a signal to a signal processing unit ( 7 ) and each having an offset and a sensitivity, wherein for each sensor element an offset correction as a function of a temperature in a ( 1 ) controllable ambient temperature and of controllable radiation sources ( 2 ), characterized in that under defined conditions, a measuring uncertainty representing a noise component influenced by the ambient temperature of the supplied signal is determined as a function of a sensor temperature and / or at least one camera temperature, the influences of the ambient temperature being determined by correction functions as a function of the sensor temperature and or the camera temperature are estimated and their camera system-specific correction coefficients in a memory unit of the signal processing unit ( 7 ) are deposited for later correction of an infrared image. Calibration method according to claim 1, characterized in that the calibration method is used for an infrared camera system with a temperature-stabilized sensor. Calibration method according to claim 1, characterized in that the calibration method is used for an infrared camera system with a temperature-stabilized sensor. Calibration method according to claims 1 and 2, characterized in that - for the sensor matrix ( 6 an operating point is set, wherein an object temperature range and an ambient temperature range are specified, that a camera-temperature-specific nonuniformity of all sensor elements is determined by means of a first characteristic adjustment of degree n, where n ≥ 2 and correction coefficient matrices a _{n, ij} with n = 0, 1, 2, ... n in the memory unit of the signal processing unit ( 7 ), wherein the nonuniformity correction is carried out at n + 1 different ambient temperatures and a constant, homogeneous object temperature, - that a correction curve for the consideration of a camera interior radiation is determined, the camera temperature at m locations within the housing ( 4 ) of the infrared camera system ( 11 ), with m ≥ 1, and the dependence of the signal voltage of the sensor ( 6 ) is determined from the camera temperature by means of a compensation calculation, in particular a polynomial approximation of at least second order, the polynomial coefficients α _{K, m} , β _{K, m} , γ _{K, m} in the memory unit of the signal processing unit ( 7 ), that an object temperature-specific nonuniformity of the infrared camera system ( 11 ) is determined by means of a second characteristic adjustment of degree n, where n ≥ 1 and correction coefficient matrices b _{n, ij} with n = 0, 1, 2,... n in the memory unit of the signal processing unit ( 7 ), wherein the nonuniformity correction is performed at n + 1 different object temperatures and a constant ambient temperature. Calibration method according to claims 1 and 3, characterized in that - for the sensor matrix ( 6 an operating point is set, wherein an object temperature range and an ambient temperature range are specified, that a sensor temperature-dependent sensitivity change is approximated with a second-order polynomial and correction coefficients g ₂ , g ₁ , g ₀ in the memory unit of the signal processing unit 7 ), wherein a signal voltage difference change of at least two different constant object temperatures depending on the measured sensor temperature at different ambient temperatures is measured, - that a sensor temperature dependent offset change is approximated with a third order polynomial and correction coefficients o ₃ , o ₂ , o ₁ , o ₀ in the memory unit of the signal processing unit ( 7 ), wherein a signal voltage change is measured at a constant object temperature as a function of the sensor temperature at different ambient temperatures, - that a camera temperature specific nonuniformity of all sensor elements is determined by means of a first characteristic adjustment of degree n, where n ≥ 2 and correction coefficient matrices a _{n , ij} with n = 0, 1, 2, ... n in the memory unit of the signal processing unit ( 7 ), wherein the nonuniformity correction is carried out at n + 1 different ambient temperatures and a constant, homogeneous object temperature, - that a correction curve for the consideration of a camera interior radiation is determined, the camera temperature at m locations within the housing ( 4 ) of the infrared camera system, with m ≥ 1, and the dependence of the signal voltage of the sensor on the camera temperature is determined by means of a compensation calculation, in particular a Polynomapproximation at least second order, wherein certain polynomial coefficients α _{K, m} , β _{K, m} , γ _{K, m} in the memory unit of the signal processing unit ( 7 ), that an object temperature-specific nonuniformity of the infrared camera system is determined by means of a second characteristic adjustment of degree n, where n ≥ 1 and correction coefficient matrices b _{n, ij} with n = 0, 1, 2,... n in the memory unit of FIG Signal processing unit ( 7 ), wherein the nonuniformity correction is performed at n + 1 different object temperatures and a constant ambient temperature. Calibration method according to one of the preceding claims, characterized in that the first characteristic adjustment represents a three-point correction. Correction procedure for a shutterless infrared camera system ( 11 ) comprising a housing ( 4 ), an infrared-transmissive optic ( 5 ), a sensor matrix ( 6 ) in the following sensor, consisting of several sensor elements, the one in the housing ( 4 ) detect incident radiation and, in response thereto, respectively send a signal to a signal processing unit ( 7 ), wherein for each sensor element an offset correction by a correction calculation with camera system-specific correction parameters is carried out, characterized in that - a sensor temperature-dependent sensitivity change is corrected with a second degree polynomial in a first correction step, - in a second correction step a sensor temperature-dependent offset change with a third polynomial - corrected in a third correction step, a camera temperature-specific nonuniformity by means of a curve adjustment of degree n, with n ≥ 2, - is corrected in a fourth correction step, the influence of a camera interior radiation according to a correction curve, - in a fifth correction step, an object temperature specific nonuniformity of Infrared camera system is corrected by means of a curve comparison of the degree n, with n ≥ 1, - wherein infrared camera system-specific Ko correction coefficients are used for the correction. Correction method according to claim 7, characterized in that the first correction step in dependence on a measured sensor temperature θ _s uses a sensor temperature-dependent sensitivity change correcting correction coefficients g ₂ , g ₁ , g ₀ , which consists of a measured signal voltage difference change between at least two different constant object temperatures depending on a measured sensor temperature can be determined at different ambient temperatures. Correction method according to Claim 7, characterized in that the second correction step uses, as a function of a measured sensor temperature θ _s, a correction coefficient o ₃ , o ₂ , o ₁ , o ₀ correcting sensor temperature-dependent offset change, which changes from a signal voltage change at a constant object temperature depending on the Sensor temperature can be determined at different ambient temperatures. Correction method according to claim 7, characterized in that the third correction step uses correction coefficient matrices a _{n, ij} with n = 0, 1, 2, ... n, which are determined at n + 1 different ambient temperatures and a constant, homogeneous object temperature. Correction method according to claim 7, characterized in that the fourth correction step polynomial coefficients α _{K, m} , β _{K, m} , γ _{K, m} a correction curve and a measured camera temperature θ _{K, m} at m different locations within the housing ( 4 ) of the infrared camera system ( 11 ), where m ≥ 1, wherein the polynomial coefficients of the correction curve are determined from the dependence of the signal voltage of the sensor on the camera temperature θ _{K, m} at different locations m, with m ≥ 1. Correction method according to claim 7, characterized in that the fifth correction step uses correction coefficient matrices b _{n, ij} with n = 0, 1, 2, ... n, which are determined at n + 1 different object temperatures and a constant ambient temperature. Correction method according to claim 7, characterized in that for an infrared camera system ( 11 ) with a temperature-unstabilized sensor, the correction steps one to five are performed. Correction method according to claim 7, characterized in that for an infrared camera system ( 11 ) with a temperature-stabilized sensor only the correction steps three to five are performed. Infrared camera system comprising a housing ( 4 ), an infrared-transmissive optic ( 5 ), a sensor matrix consisting of several sensor elements ( 6 ) and a digital signal processing unit ( 7 ) using the calibration method and the correction method of the preceding claims, characterized in that inside the housing ( 4 ) of the infrared camera system ( 11 ) at least two camera temperature measuring Temperature measuring means ( 8a . 8b ) are arranged. Infrared camera system according to claim 15, characterized in that the temperature measuring means ( 8a . 8b ) is formed as a thermocouple and / or a thermal resistor.

Images