Thermal imaging member detection

Abstract

The invention relates to a device and a method for thermal imaging component detection, comprising an excitation source for generating a transient heat flow in a detection object, an infrared detector array for detecting thermal radiation emitted from a surface of the detection object, a surface scanner, a control device and an evaluation device, the apparatus includes an inertial measurement unit for detecting motion of the apparatus.

Description

Technical Field

[0001] The invention relates to a device and a method for non-destructive component detection. Background Art

[0002] Active thermal imaging is known from the state of the art as a non-destructive and imaging test method for characterizing materials and components. This is based on thermal excitation of the test body by means of absorption of optical radiation, induced eddy currents, coupled mechanical waves or other forms of energy, which leads to a time-dependent temperature change in the test body. The thermal radiation of the test body can be acquired contactlessly by means of infrared sensors (implemented as point, line or surface detectors). Based on the measured temperature field at the component surface, the time-variable heat diffusion processes are analyzed, and component features can be detected and identified from this. Known technical implementations of active thermal imaging are, for example, laboratory structures on fixed tripods or mobile, manual or robot-guided detection systems. Summary of the invention

[0003] The object of the present invention is to create a device and a method for non-destructive component testing, with which deep component features or defects can be assigned to the three-dimensional component geometry with greater reliability. In addition, the object of the present invention is to create a device with which non-destructive component testing can be used in a wider range of application areas and with improved operator friendliness.

[0004] This object is achieved by the device and the method according to the claims.

[0005] The device for thermal imaging component detection according to the present invention comprises: an excitation source for generating transient heat flow in the detection object, an infrared detector array for detecting thermal radiation emitted from the surface of the detection object, a surface scanner, a control device and an evaluation device, wherein the device comprises an inertial measurement unit for detecting the movement of the device. In addition, the device also comprises a physical unit for user identity verification.

[0006] A development of the device is particularly advantageous in which the evaluation device comprises a geometric acquisition system for calculating the spatial coordinates of the surface of the test object.

[0007] According to an advantageous embodiment of the device, it is provided that the geometry acquisition system is designed to calculate the relative spatial position between the device and the test object.

[0008] A further development of the device provides that the infrared detector array of the surface scanner and the inertial measurement unit are arranged in the device at a defined distance and in a defined direction relative to one another. This has the advantage that a flexible and operator-friendly applicability of the device can be achieved thereby.

[0009] A design of the device is also advantageous, according to which the evaluation device is designed for program-controlled reconstruction of defects in the test object or material differences or material properties of the test object, wherein the evaluation device calculates the time- and position-dependent surface temperature signal T from the infrared image data from the infrared detector array. mess , and where the surface temperature signal T is mess Convert to specular source representation T Sq This has the advantage that reconstructed component features can thereby be assigned to real component features (defects, interfaces, etc.) with greater reliability.

[0010] The object of the present invention is also solved independently by a method for thermal imaging component detection of a test object using a device, the device comprising an excitation source, an infrared detector array, a surface scanner, a physical unit for user authentication, a control device and an evaluation device, the method having the following method steps: authenticating the user, acquiring the spatial position and orientation of the device relative to the test object and acquiring the surface of the test object using the surface scanner; generating a transient heat flow in the test object by the excitation source; recording an infrared image of the surface of the test object by the infrared detector array during a preselected measurement duration; reconstructing the spatial position of a defect by means of the evaluation device from the data of the acquired infrared image, wherein, in order to reconstruct the defect by means of the evaluation device, a surface temperature signal T is calculated from the data of the infrared image. mess , and where the surface temperature signal T is mess Convert to specular source representation T Sq .

[0011] It has also proven to be advantageous if in the method a Green's function based on the heat conduction equation is used in the regularization method.

[0012] A method principle is advantageous according to which the spatial position of the defect is reconstructed using additional information from the group consisting of heat flow dimension, number of boundary layers in the test object, position of boundary layers in the test object, thermophysical material properties or boundary conditions.

[0013] According to an advantageous development of the method, it is provided that the device performs component recognition, thereby enabling a priori information about the detection object to be used as additional information.

[0014] According to an advantageous development of the method, it is provided that the user authentication is performed by the device. Thus, the additional information is dependent on the user of the appliance.

[0015] For component identification and user authentication, technologies such as contactless transmitter-receiver systems (RFID technology), optoelectronically accessible fonts, barcodes or multi-dimensional codes (with the help of a camera or scanner), biometric authentication (including through face, fingerprint, iris or voice recognition), manual authentication with the help of a user interface or card reader, or digital authentication based on cryptographic handshake methods can be used.

[0016] According to an advantageous development of the method, it is provided that an inertial measurement unit is arranged in the device, wherein the infrared detector array of the surface scanner and the inertial measurement unit are arranged in the device at a defined distance and in a defined direction relative to one another.

[0017] Also advantageous is the method principle that the spatial position and orientation of the device relative to the test object is measured by an inertial measurement unit. This enables the image sequence to be corrected and stabilized by the inertial measurement unit.

[0018] According to an alternative method principle, it is provided that external data sources are used to determine the spatial position and orientation of the device and to determine the surface of the test object.

[0019] An advantageous embodiment of the method provides that additional information about the component geometry and material properties of the test object is used for an optimized discretization of the reconstruction space.

[0020] By repeating the method steps one or more times according to the method principle, executing from different spatial positions and orientations of the device relative to the test object, and superimposing the results of the reconstruction of the spatial position of the defect, the advantages of higher accuracy and reliability in assigning defects or component features to the three-dimensional geometry of the test object are achieved.

[0021] A development of the method according to which temperature- and position-corrected image data are calculated from the infrared image has also proven to be advantageous.

[0022] According to an equally advantageous method principle, it is provided that the extracted defect features can be used for data compression. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to better understand the present invention, the present invention is explained in more detail according to the following drawings.

[0024] Here are highly simplified diagrams:

[0025] Figure 1 A device for performing non-destructive component testing on a test object is shown;

[0026] Figure 2 Shown according to Figure 1 a view of the device corresponding to a viewing direction toward its sensor or its infrared detector array;

[0027] Figure 3 A flow chart showing a method for thermal imaging component detection;

[0028] Figure 4 Shown according to Figure 3 Details of the regularization method in the method step “Regularization and Reconstruction”;

[0029] Figure 5 The apparatus is shown in the process of performing thermal imaging inspection on a test sample;

[0030] Figure 6 A diagram showing an evaluation step performed using a regularization method of a measured surface temperature signal according to a temporal and spatial temperature distribution;

[0031] Figure 7 The evaluation steps of the regularization method for another application case are shown by means of time and space temperature curves;

[0032] Figure 8 Another embodiment of an apparatus for non-destructive component testing is shown. DETAILED DESCRIPTION

[0033] By way of introduction, it should be understood that in the various described embodiments, identical parts are provided with identical reference symbols or identical component names, wherein the disclosure contained in the entire description may be meaningfully transferred to identical parts having identical reference symbols or identical component names. Likewise, selected positional descriptions in the description, such as top, bottom, side, etc., refer to the directly described and displayed figures, and when the position changes, these positional descriptions shall be meaningfully transferred to the new position.

[0034] Figure 1 A device 1 for non-destructive component testing of a component or test object 2 is shown. The device 1 is suitable for carrying out a method for component testing with the use of active thermal imaging and comprises an excitation source 3 and an infrared detector array 4 for this purpose. In addition to a power supply unit 5, a control device 6 or a processor unit is provided for controlling the inspection sequence and controlling or regulating the components of the device 1 required for this purpose. A part of the control device 6 is formed by an evaluation device 7 for program-controlled processing of the measurement signals detected by the infrared detector array 4 or other sensors. In addition, the device 1 comprises a surface scanner 8 for acquiring geometric data or geometric features of the test object 2. Finally, the device 1 is also configured with an inertial measurement unit 9 for motion detection of the device 1. In addition, the device 1 comprises an operator terminal 10, which also comprises at least one digital bidirectional communication interface for providing data or retrieving additional information from an external data source. Optionally, the device 1 can also be equipped with an optical camera or an optical imaging system 11 for the visible wavelength range (so-called RGB camera).

[0035] Figure 2 Shown according to Figure 1 1 , which corresponds to a viewing direction towards the infrared detector array 4 or to a viewing direction of the optical axis 12 of the infrared detector array 4 .

[0036] like Figure 1 As shown in FIG. 1 , the test object 2 has inhomogeneities, referred to in the following for short as defects 14 , in its interior below a surface 13 .

[0037] Figure 3 According to the flowchart, the detection object 2 ( Figure 1 ) is a process for thermal imaging component detection. This is a multi-stage signal processing process. In the initial measurement process, the surface 13 of the detection object 2 is first optically acquired, and the initial posture (position and direction) of the device 1 relative to the detection object 2 or its surface 13 is determined. Before the detection object 2 is thermally excited by the excitation source 3, one or more static (passive) infrared images of the detection object 2 are acquired by the infrared detector array 4.

[0038] By determining the transformation parameters, one or more coordinate points p can be assigned to an image point I (u, v) ( Figure 5 ). Since no movement of the device 1 occurs between the determination of the position of the device 1, the thermal excitation by means of the excitation source 3 and the recording of the thermal response by the infrared detector array 4, this assignment also applies to the temporally dynamic part of the infrared radiation emitted by the test object 2 and therefore does not necessarily have to be performed during the measurement of the thermal excitation and the response. The coordinate transformation can be determined by a previously performed calibration of the infrared detector array 4 and the surface scanner 8 or by an image registration method based on characteristic image features (e.g. edges or emissivity differences).

[0039] In order to reduce artifacts in the reconstruction and registration steps, the scanner data are merged (if necessary, outlier removal, noise filtering, image size reduction, local resolution reduction). In addition, the normal vector n is determined for all relevant coordinate points of the surface 13 of the test object 2. p Here, the normal vector n p Always oriented along the direction inside the component ( Figure 5 ).

[0040] After thermal excitation by the excitation source 3, the thermal response of the test object 2 is acquired sufficiently long by recording an infrared image by means of the infrared detector array 4. Here, the measurement duration and the form of excitation determine the desired penetration depth into the test object 2. The frame rate when recording the infrared image by means of the infrared detector array 4 determines the resolution of the device 1 in the axial direction (i.e., in the depth direction of the test object 2).

[0041] The thermal response for each image point is inversely transformed into a specular source representation by means of a so-called regularization by means of the evaluation device 7. In this representation, subsurface structures, such as defects 14, are represented as pulse-shaped features in the discrete equidistant signals I(u, v, w).

[0042] Figure 3 The measures to be performed in the thermal imaging detection method and their chronological sequence are summarized. After the device 1 is put into use, the spatial position and orientation of the device 1 are determined by means of an inertial measurement unit (IMU). On the other hand, the external geometry of the test object 2 is measured by means of a surface scanner 8. The measurement data obtained here provide information about the geometry and spatial relative arrangement of the device 1 and the test object 2 as a measure 100. The measurement according to measure 100 is preferably performed permanently, that is, continuously during the entire time of the component detection of the test object 2. This is particularly advantageous for the case where the device 1 is configured as a handheld device. Section 101 includes measurements or detections performed before the test object 2 is thermally excited. In this way, one or more static (passive) infrared images of the test object 1 are recorded using the thermal excitation of the infrared detector array 4, and the conversion parameters for assigning the coordinate point p to the corresponding image point I (u, v) of the infrared image are determined. In the subsequent section 102, thermal excitation is performed by means of an excitation source 3, and a series of infrared images are recorded by the infrared detector array 4 to obtain the thermal response. The image recorded here by the infrared detector array 4 represents the surface temperature data T of a point p on the surface 13 of the detection object 2. mess (p) and the time curve of the surface temperature. In the final section 103, the surface temperature signal thus obtained is subjected to an evaluation procedure using the evaluation device 7. For this purpose, the surface temperature signal T mess This enables the defect 14 in the test object 2 to be reconstructed by a so-called regularization method.

[0043] Figure 4 Shows Figure 3 The details corresponding to the method step “regularization and reconstruction” are further explained in section 103. After the thermal response from the surface 13 of the test object 2 has been recorded by means of the infrared detector array 4, a conversion to the aforementioned specular source representation takes place. In order to reconstruct a defect 14 located at or below the surface 13 or to determine material parameters, the measured surface temperature signal Convert to the corresponding specular source representation (IR, the set of real numbers). Here, N u and N v Describes the number of pixels in the u and v directions of the surface detector of the infrared detector array 4. The number of time steps for the measurement is expressed by N t Indicated by, and the number of coordinate points in the depth direction is represented by Nw Here, in the simplest case, the transformation matrix can be obtained by 33 The Green's function is expressed in the following form.

[0044] Equation (1)

[0045] This function corresponds to the basic solution of the heat conduction equation. It should also be mentioned here that other Green's functions can also be used based on the existing component characteristics and detection environment characteristics. The thermal diffusivity α in the depth direction is used 33 , and using discrete number steps and , and in using the run variable and In the case of , the elements of the transformation matrix K can be written as follows using the Green's function described above:

[0046] Equation (2)

[0047] Here, the item describes a dimensionless number where is a dimensionless scaling parameter used to optimize the quality of the inverse solution, and represents the temporal resolution of the surface temperature signal, and Indicates the depth resolution represented for the specular source.

[0048] By mathematically convolving the transformation kernel or transformation matrix K with respect to time and location, possible temporal and spatial excitation patterns can be considered. Since the diffusion process and thus the heat conduction equation are considered at the macroscopic level, the linear matrix equation T is obtained in discrete form mess =KT sq , and thus the linear inverse problem is obtained. Here, it can also be determined that equation (2) is independent of the horizontal coordinates u and v. Therefore, solving the matrix equation T mess =KT sq represents a local inverse problem. This means that the measured temperature signal can be converted locally (pixel by pixel) to the corresponding specular source representation.

[0049] For this conversion, a very ill-posed inverse problem has to be solved due to the entropy generation during thermal diffusion. For this purpose, different regularization methods can be used. It should also be noted that the inverse problem can also be solved using machine learning methods.

[0050] In the present case, the inverse problem is solved using additional information in the evaluation process, such as positivity or sparsity (solution matrix sparsity). The calculated specular source distribution as a function of depth discloses the characteristics of the measured surface temperature signal in the form of sources (positive amplitudes) and sinks (negative amplitudes) relative to structurally identical interfaces (surface 13) and defect interfaces. Sources and sinks can occur here, for example, via defects or prevalent boundary conditions, wherein a clear distinction can be made between characteristics due to defects and characteristics due to boundary conditions. Boundary conditions or inspection environment characteristics can be described here by thermal conduction, thermal radiation or convection or a mixture of these effects. Furthermore, adiabatic boundary conditions provide only positive specular source amplitudes. Thermal radiation and convection can provide negative and positive specular source amplitudes ( Figure 6 , 7 ).

[0051] Based on the defect characteristics, the depth position of defect 14 can only be roughly estimated. In addition, due to boundary conditions and the observation surface, the amplitude does not provide any relevant information about the position of defect 14 and the back wall. Therefore, the amplitude due to defect 14 and the back wall is extracted. The noise-free defect temperature signal is then calculated from the resulting signal. By eliminating the noise signal, the defect depth can now be determined more accurately by evaluating the maximum defect temperature signal.

[0052] Based on the calculated defect depth, a replacement mirror source signal can be determined for each pixel of the infrared detector array 4 and each defect signal , for multidimensional defect representation (see Figure 6 f and Figure 6 h).

[0053] Furthermore, if the thermal diffusivity of the base material e 2 Greater than the thermal diffusivity e of the defect material 1 , then on the depth scale of the specular source distribution, positive amplitudes (sources) are visible before negative amplitudes (sinks) ( Figure 6 c, 6d). If e 1 Greater than e 2 , then in depth a sink is first obtained, and then a source is obtained ( Figure 7 ).

[0054] The depicted conversion of the measured temperature signal into a corresponding specular source representation does not take into account lateral diffusion processes and geometric information about the possibly anisotropic and complexly shaped test object 2 .

[0055] In a further evaluation step, a local filtering can be performed with the aid of the geometric information of the test object 2 recorded by the surface scanner 8 in combination with the known thermal conductivity tensor α. This allows a more accurate reconstruction of the real interface dimensions. This local filtering can also be performed by solving an inverse problem. This additional and optional evaluation step is particularly recommended for defects 14 located deep below the surface 13 of the test object 2 and when the ratio of the thermal diffusivity in the lateral direction to that in the depth direction is large.

[0056] according to Figure 4 In the diagram in FIG. 1 , the surface temperature signal undergoes a series of multiple steps for its evaluation, regularization and defect reconstruction. Here, the first step 111 corresponds to the surface temperature signal T mess In the subsequent step 112, the measured surface temperature signal T mess Corresponding to the linear matrix equation T mess =KT sq is converted into a corresponding specular source distribution. In a subsequent step 113, defect features are extracted. In step 114, a defect temperature signal is calculated based on the extracted defect features. The next step 115 corresponds to determining the true defect depth based on the calculated defect temperature signal. The true defect depth is used in a subsequent step 116 to calculate a substitute specular source signal. In a final step 117, local filtering is performed based on the measured geometric information of the test object 2 and information about the known thermal diffusivity tensor. Based on the results obtained here, the size determination of the defect 14 in the test object 2 can be improved or optimized.

[0057] Figure 5 The device 1 is shown in the process of performing a thermal imaging inspection of an inspection object 2 .

[0058] By means of the excitation source 3, the test object 2 has been thermally excited, and the thermal response of the component is detected by the infrared detector array 4 of the device 1 within a preset measurement time, and geometric information is acquired by means of the surface scanner 8 of the device 1 ( Figure 1 , 2 ). That is, the infrared detector array 4 records a series of thermal images. Based on at least one spatial position k of the device 1, a 3D depth signal I is reconstructed based on this. k (u, v, w). As a prerequisite, here for the surface point p i and the subordinate normalized normal vector n pi There must be at least one image location (u p* , v p* ). From the depth signal I k (u p* , v p* , w) are projected into the global coordinate system.

[0059] In the known material parameters (for example in the form of a thermal conductivity sensor) and in the normal vector implicitly known from the surface 13 or manually defined, the relevant surrogate feature P in the global coordinate system consists of its position and its intensity.

[0060]

[0061] according to Figure 5 , projecting the surrogate feature locations of one or more data sets I(u, v, w) into the global coordinate system results in the surrogate features being clustered at or near the location of the physical component features, as long as the surrogate features belong to the same component feature (defect 14). For example, for adjacent surface points from a single measurement, or for multiple measurements from device 1 at congruent or different spatial locations (i.e., performing steps 101, 102, and 103 multiple times), according to Figure 3 ), this is the case for congruent or adjacent surface points. When the component or object 2 to be inspected is thin enough or the measurement of the temperature response is long enough, multiple measurements of relatively spaced positions from the device 1 (i.e., multiple executions of steps 101, 102, 103) are performed according to Figure 3 ) surface points, this aggregation of substitute features is also the case. In this case, the opposite, invisible component surface becomes visible as a substitute feature in each measurement.

[0062] The thermographic measurements to be superimposed may differ from one another in all common measurement parameters, since these parameters serve as prior information in the reconstruction and localization of the substitute features. These measurement parameters may be: the temporal and spatial excitation form by the excitation source 3; the measurement frequency (corresponding to the recording of the thermal image by the infrared detector array 4); the spatial resolution and the distance to the surface 13 of the object 2 or the spatial position of the device 1.

[0063] In the case of the converted data being stored in the form of a point cloud (i.e. the coordinates are explicitly present), the depth resolution can be defined independently of the resolution of the geometric detection by the surface scanner 8. In addition to alternative feature positions, alternative feature intensities can also be assigned to the data points. This enables an improved visualization of the features of the inspection object 2, for example by texturing the point cloud or rendered surface of the features or defects 14. Furthermore, if the result data are to be present in a three-dimensional Cartesian grid with implicit coordinates (volume pixels, or “voxels” for short), this enables the determination of intensity values. In this case, the alternative feature positions of the relevant data points are transferred to the grid by interpolation.

[0064] By systematically positioning the surrogate features in the global coordinate system, further reconstruction and segmentation methods can be used in subsequent steps, which enables a more precise positioning and visualization of the component features. The introduction of geometric (surface points, normal directions, prior information about the features of the inspection object 2, such as orientation, size and shape) and material-specific additional information also makes it possible to apply known methods based on superposition, triangulation, lateral measurement, regression and machine learning.

[0065] The following will be based on Figure 6 and Figure 7 The method for the thermographic reconstruction of the defect 14 is explained in more detail.

[0066] Figure 6 The evaluation procedure for reconstructing a defect 14 and for determining material parameters of a test object 2 is shown. FIG. a) shows by way of example a defective test object 2 in which the heat outflow rate e of the base material is 2 Greater than the heat outflow rate e of the defective material 1 . Figure b) shows the characterization temperature signal measured in reflection mode for the defect area. Figure d) shows the specular source distribution calculated therefrom. Figure c) shows the extracted defect and component features, Figure e) shows the defect temperature signal calculated therefrom, and Figure f) shows the corrected specular source depth distribution. Figures g) and h) show 2D visualizations of the extracted defect and component features and the corrected specular source depth distribution.

[0067] The method steps correspond to Figure 4 The characterization of the component of the test object 2 depends on the thermal impedance of the base material and the defect material, wherein the thermal impedance is formed by the respective outflow rate, which in turn depends on the thermal conductivity k, the specific heat capacity c p and material density ρ ( Figure 6 a). The greater the difference in thermal impedance, the more intensely the characteristic component features of the test object 2 are displayed. Since the positions of the image points relative to each other in space are known, the individual feature intensities and positions can be used for more precise feature localization and quantification by appropriate superposition. For example, a component feature below the surface 13 of the test object 2 imaged through a defect interface can appear in multiple adjacent image points due to its spatial extent, with a distortion of the spatial extent occurring simultaneously due to heat diffusion. Figure 6 b exemplarily shows the surface temperature signal measured for a reflection (pulse-echo) configuration based on very short excitation pulses (excitation source 3 and infrared detector array 4 on the same side) as a function of time in the form of a Dirac delta distribution with respect to time.

[0068] However, the described procedure for thermographic defect reconstruction can be applied to all temporal and spatial thermal excitation functions, as well as in the case of a transmission configuration (excitation source 3 and infrared detector array 4 are located on opposite sides).

[0069] In order to ensure that all relevant component properties are captured, test measurements can be performed. Using test measurements, the thermal diffusion time t d and the measurement time t mess .

[0070] The temperature signal measured after thermal excitation can be converted locally (i.e. pixel by pixel) to the corresponding specular source representation. Figure 6 The graph shown in d illustrates the calculated specular source distribution as a function of depth and discloses the characteristics of the measured surface temperature signal in the form of sources (identifiable by positive amplitudes) and sinks (identifiable by negative amplitudes) relative to the component interface and the defect interface. By extracting the amplitudes due to defect 14 and the back wall, we finally obtain Figure 6 The diagram shown in c.

[0071] The noise-free defect temperature signal is then calculated from the signal thus generated. By evaluating the maximum defect temperature signal of the filtered defect temperature signal, the defect depth can then be determined more accurately. This is done by Figure 6 e and 6g are explained, where Figure 6 The diagram in g shows a cross section through the surface 13 of the test object 2 together with the corresponding specular source distribution.

[0072] Based on the calculated defect depth, a replacement specular source signal can then be determined for each camera pixel and defect signal , for multi-dimensional defect representation, such as by Figure 6 f and 6h. Figure 6 In the example, the heat outflow rate of the base material is based on the prerequisite e 2 Greater than the heat outflow rate e of the defective material 1 .

[0073] Figure 7 The evaluation procedure for reconstructing the defect 14 and for determining the material parameters is shown in FIG. Figure 6 The heat release rate of the base material is different depending on the treatment situation. 2 Less than the heat outflow rate e of the defective material 1 The temperature signal was recorded in reflection mode. The corresponding specular source distribution in the defect area is shown in a). The extracted defect and component features are shown in b), and the defect temperature signal calculated therefrom is shown in d). The corrected specular source depth distribution is shown in c). Figure 7The outflow rate e for the defective material is thus shown 1 Greater than the outflow rate of the base material e 2 An example of thermal imaging defect reconstruction is shown in the case of . In this case, at the depth scale, a sink is first derived and then a source is derived ( Figure 7 a and Figure 7 b).

[0074] In order to consistently record new thermal imaging data and reconstruct the component features of the test object 2 from these data, the geometric features acquired by the surface scanner 8 (e.g. by a TOF camera, a contour scanner, by superimposing image data from different postures according to the device itself) and the position estimate from the inertial measurement unit are used. The position estimate of the sensor in the original coordinate system is continuously performed.

[0075] A preferred embodiment of the device 1 is a handheld device, on which all necessary components are positioned at a defined relative distance from each other.

[0076] In an alternative embodiment variant of the device 1 , the spatial position relative to the component surface or the surface 13 of the test object 2 is provided by an external data source, for example by a single-axis or multi-axis robot which moves the device 1 .

[0077] In another alternative embodiment of the device 1, the surface 13 of the test object 2 and the inertial position of the device 1 relative to the surface 13 are provided by an external data source, for example by a CAD model of the test object 2. In an equally alternative embodiment, the device 1 remains stationary and the test object 2 is moved by a suitable device. In this case, the relative position of the device 1 is determined by a manipulation device of the test object 2 and provided to the device 1 for evaluation.

[0078] Below, we refer again to Figure 1 In the case of the illustration of the device 1 of FIG. 1 , another possible application possibility or construction mode of the device for thermal component detection is described. Here, before the method steps already described, the detection object 2 is identified at the beginning. That is, the control device 6 or the evaluation device 7 is configured to acquire or automatically read the identifier 15 coded in a certain form installed on the surface 13 of the detection object 2. This makes it possible to clearly assign the results of the performed thermal imaging component detection to the corresponding detection object 2 later. As a basis for this, an image preferably taken by an infrared camera or an infrared detector array 4 of the identifier 15 is used. Alternatively, it is also possible to record the image required for component identification with the help of an optical camera 11. The control device 6 or the evaluation device 7 is configured to program-control the decoding of the image information data and identify the identifier 15. In this embodiment variant of the device 1, the identifier 15 can be formed by writing or a one-dimensional or multi-dimensional code (such as a barcode or a QR code) that can be acquired photoelectrically.

[0079] In a preferred construction of the device 1, a program-controlled user check 16 is provided in the control device 6. Therefore, in this construction variant of the device 1, its debugging can be limited to a circle of people who are authorized for this. Thus, for example, user identity verification can be performed by checking a password entered by the user via the operating terminal 10. Alternatively to this, user identity verification can also be based on technologies such as facial recognition or eye iris recognition. Alternatively, an infrared camera or infrared detector array 4 or an optical camera 11 of the device 1 can be used to record the corresponding image. Alternatively, the program-controlled user check 16 of the device 1 can also be performed based on other technologies, such as reading RFID elements and identifying fingerprints using corresponding readers or scanners. Corresponding to the technology used respectively, the user check 16 of the device 1 is constructed for image recognition or for decoding the corresponding detection signal of the corresponding reading device. Another embodiment of user identification can also provide a two-factor or multi-factor identity verification.

[0080] According to the following Figure 8 Another embodiment variant of the device 1 for detecting thermal imaging components is described. Figure 8 A device 1 for non-destructive component testing of a test object 2 is shown, which also comprises two excitation sources 3 for generating heat flows in the test object 2 and a control device 6 with an evaluation device 7. In this example of the device 1, two infrared detector arrays 4 are provided for detecting or acquiring thermal radiation emitted from a surface 13 of the test object 2. This allows a combined evaluation of surface temperature signals obtained from the corresponding infrared images of the two infrared detector arrays 4, thereby enabling further greater spatial or temporal resolution to be achieved.

[0081] On the other hand, the quality of the component inspection result determined using the method can also be improved by the type of thermal excitation of the surface 13 of the inspection object 2 caused by the excitation sources 3. One possibility for this is to arrange filters 17 in the respective radiation paths of the two excitation sources 3. The two filters 17 select a narrow spectral range from the infrared radiation generated by the two excitation sources 3.

[0082] Another embodiment variant provides that a focusing device or focusing lens 18 is respectively arranged in the radiation path of the respective excitation source 3. The focusing lens 18 allows the method for component testing to be carried out in such a way that thermal excitation with infrared radiation can be carried out in a locally concentrated manner on the surface 13 of the test object 2. By correspondingly constructing or adjusting the focusing lens 18, the heat input can be concentrated, for example, onto a surface area that is as point-shaped or line-shaped or segment-shaped as possible. This is advantageous for the applicable evaluation method, because the initial conditions of the heat flow generated in the test object 2 can be determined or preset more accurately.

[0083] The embodiments illustrate possible implementation variants, wherein it should be noted here that the present invention is not limited to the implementation variants specifically shown, but on the contrary various combinations of the individual implementation variants with one another are possible and due to the teaching of the technical actions of the objective invention, such variation possibilities are within the skill range of a person skilled in the art.

[0084] The scope of protection is determined by the claims. However, the claims should be interpreted using the description and the drawings. Individual features or combinations of features in the different embodiments shown and described may represent independent inventive solutions in themselves. The problems on which the independent inventive solutions are based can be gathered from the description.

[0085] All descriptions of value ranges in the specific description should be understood to include any range and all sub-ranges thereof, for example, a description of 1 to 10 should be understood to include all sub-ranges starting from a lower limit of 1 to an upper limit of 10, that is, all sub-ranges start with a lower limit of 1 or greater and end with an upper limit of 10 or less, for example, 1 to 1.7, or 3.2 to 8.1, or 5.5 to 10.

[0086] Finally, for the sake of neatness, it should be noted that elements are partially not shown to scale and / or are shown enlarged and / or reduced in size for a better understanding of the structure.

[0087] Reference numerals list

[0088] 1 Device

[0089] 2 Detection Objects

[0090] 3 Excitation source

[0091] 4 Infrared detector array

[0092] 5 Power supply unit

[0093] 6 Control device

[0094] 7 Evaluation device

[0095] 8 Surface Scanner

[0096] 9 Inertial Measurement Unit

[0097] 10 Operation Terminal

[0098] 11. Camera

[0099] 12 Optical axis

[0100] 13 Surface

[0101] 14 Defects

[0102] 15 Identifiers

[0103] 16 User Check

[0104] 17 Filter

[0105] 18 Focusing lens

[0106] 100 measures

[0107] 101 Segment

[0108] 102 Segment

[0109] 103 Segment

[0110] 111 Steps

[0111] 112 Steps

[0112] 113 Steps

[0113] 114 Steps

[0114] 115 Steps

[0115] 116 Steps

[0116] 117 Steps

Claims

1. A device (1) for thermal imaging component detection, comprising: - an excitation source (3) for generating a transient heat flow in the test object (2), - an infrared detector array (4) for detecting thermal radiation emitted by the surface (13) of the test object (2), - Surface scanner (8), - control device (6) and evaluation device (7), Characterized in that the device (1) comprises an inertial measurement unit (9) for detecting the movement of the device (1).

2. The device (1) according to claim 1, characterized in that The construction geometry acquisition system is used to calculate the spatial coordinates of the surface (13) of the detection object (2).

3. The device (1) according to claim 2, characterized in that The geometry acquisition system is designed to calculate the relative spatial position between the device (1) and the detection object (2).

4. The device (1) according to any one of the preceding claims, characterized in that The infrared detector array (4), the surface scanner (8) and the inertial measurement unit (9) are arranged in the device (1) at a defined distance and in a defined orientation relative to one another.

5. Device (1) according to any one of the preceding claims, characterized in that Two infrared detector arrays (4) are designed to detect thermal radiation emitted by a surface (13) of the test object (2).

6. Device (1) according to any of the preceding claims, characterized in that Construct two excitation sources (3).

7. Device (1) according to any of the preceding claims, characterized in that The excitation source (3) comprises a filter (17) for selecting a part of the spectral range of the infrared radiation.

8. Device (1) according to any of the preceding claims, characterized in that The excitation source (3) comprises a focusing lens (18).

9. Device (1) according to any of the preceding claims, characterized in that The control device (6) and / or the evaluation device (7) are designed for program-controlled identification of the test object (2) by decoding an identifier (15) applied to a surface (13) of the test object (2).

10. The device (1) according to any one of the preceding claims, characterized in that A user check (16) is included and is designed for program-controlled authentication of the user, wherein a method for authenticating the user selected from the group consisting of entering a password, performing facial recognition or eye iris recognition, detecting a fingerprint, reading an RFID element or reading a user card provided with a magnetic stripe is used.

11. Device (1) according to any of the preceding claims, characterized in that The evaluation device (7) is designed for program-controlled reconstruction of a defect (14) in the test object (2), wherein the evaluation device (7) calculates a surface temperature signal T based on infrared image data from the infrared detector array (4). mess , and wherein the surface temperature signal T is regularized using a regularization method mess Convert to specular source representation T sq .

12. A method for performing thermal imaging component detection on a detection object (2) using a device (1), the device comprising an excitation source (3), an infrared detector array (4), a surface scanner (8), a control device (6) and an evaluation device (7), the method comprising the following method steps: - obtaining the spatial position and orientation of the device (1) relative to the detection object (2), and obtaining the surface (13) of the detection object (2) by using the surface scanner (8); - generating a transient heat flow in the detection object (2) by the excitation source (3); - acquiring an infrared image of the surface (13) of the test object (2) using the infrared detector array (4) during a preselected measurement duration; - reconstructing the spatial position of the defect (14) by means of the evaluation device (7) from the data of the acquired infrared image, It is characterized in that In order to reconstruct the defect (14), the evaluation device (7) calculates the surface temperature signal T from the data of the infrared image. mess ,and - The surface temperature signal T mess is converted to a specular source representation T sq , where T mess =KT sq , and where K is the transformation matrix.

13. The method according to claim 12, characterized in that According to the specular source representation T sq Calculate alternative specular sources and / or alternative features.

14. The method according to claim 13, characterized in that A regularization method or a machine learning method is applied to calculate the substitute specular source and / or the substitute feature.

15. The method according to claim 14, characterized in that A Green's function based on the heat conduction equation is applied in the regularization method.

16. The method according to any one of claims 12 to 15, characterized in that When reconstructing the spatial position of the defect (14), additional information from the group consisting of the dimensionality of the heat flow, the number of boundary layers in the test object (2), the position of the boundary layers in the test object (2), thermophysical material properties or boundary conditions is used.

17. The method according to any one of claims 12 to 16, characterized in that An inertial measurement unit (9) is arranged in the device (1), wherein the infrared detector array (4), the surface scanner (8) and the inertial measurement unit (9) are arranged at a defined distance and in a defined direction relative to one another in the device (1).

18. The method according to any one of claims 12 to 17, characterized in that The spatial position and the orientation of the device (1) and the detection object (2) relative to each other are measured by the inertial measurement unit (9).

19. The method according to any one of claims 12 to 18, characterized in that External data sources are taken into account to obtain the spatial position and orientation of the device (1) and to obtain the surface (13) of the detection object (2).

20. The method according to any one of claims 12 to 19, characterized in that In the method step of recording an infrared image of the surface (13) of the detection object (2), the infrared image is recorded by two infrared detector array cameras (4).

21. The method according to any one of claims 12 to 20, characterized in that In the method step of generating a heat flow in the test object (2) by means of the excitation source (3), the infrared radiation emitted by the excitation source (3) is selected by means of a filter (17) to form a reduced spectral range of the infrared radiation.

22. The method according to any one of claims 12 to 21, characterized in that In the method step of generating a heat flow in the detection object (2) by means of the excitation source (3), the infrared radiation emitted by the excitation source (3) is applied to the surface (13) of the detection object (2) in a point-shaped and / or line-shaped or segment-shaped manner that is as concentrated as possible by means of a focusing device or a focusing lens (18).

23. The method according to any one of claims 12 to 22, characterized in that Additional information about the component geometry and material properties of the test object (2) is used for an optimized discretization of the reconstruction space.

24. The method according to any one of claims 12 to 23, characterized in that The method steps are repeated one or more times, performed from different spatial positions and orientations of the device (1) relative to the test object (2), and the reconstruction results of the spatial position of the defect (14) are superimposed.

25. The method according to any one of claims 12 to 24, characterized in that Calculate temperature and position calibrated image data from infrared images.

26. The method according to any one of claims 12 to 25, characterized in that The extracted features of the defect (14) or boundary layer are used for data compression.

27. The method according to any one of claims 12 to 26, characterized in that Prior to the method step of acquiring the spatial position and orientation of the device (1) relative to the detection object (2), component identification is performed by acquiring an identifier (15) attached to the detection object (2).

28. The method according to any one of claims 12 to 27, characterized in that Prior to the method step of acquiring the spatial position and orientation of the device (1) relative to the detection object (2), user authentication is performed, wherein the user authentication comprises a method selected from the group consisting of entering a password, performing facial recognition or eye iris recognition, detecting a fingerprint, reading an RFID element, or reading a user card provided with a magnetic stripe.

Images