Method and apparatus for inspecting painted surfaces containing effect pigments

Abstract

1. A method for inspecting a lacquered surface (10), preferably having one or more layers with absorbing and / or effect pigments, characterized in that radiation is irradiated by a first radiation device (2) at a first predetermined angle of incidence (a1) onto the surface (10) to be inspected, and a color image recording device (4) records a spatially resolved image of the surface illuminated by the incident direction at a first observation angle (b), the image recording device (4) having a first predetermined sensitivity (F(l)) that depends on the wavelength of the radiation incident on the image recording device, and an image evaluation device performs a patch-by-pattern, preferably pixel-by-pixel, evaluation of the image recorded by the image recording device, and the color image recording device (4) is also used to evaluate and / or assess the absorption and / or effect pigments and / or the influence of the effect pigments on the image recording and / or integral color measurement is taken into account and / or excluded as part of the image evaluation.

Description

[Technical Field]

[0001] The present invention relates to a method and an apparatus for inspecting lacquered surfaces, in particular surfaces having a lacquer mixture preferably containing an absorption pigment and an effect pigment. Such lacquer layers have been known in the prior art for a long time. Various methods and apparatus for inspecting and / or analyzing such surfaces are also known in the prior art.

[0002] It is known to record the spectral characteristics of the illuminated measuring spot with a spectroscopic element (such as a grating, prism or filter) and compare this, for example, with a standard. It is also known that the measurement results of such measuring methods often deviate significantly from one another, on the one hand because differences between the image recording characteristics of a camera and the image recording characteristics of the human eye are not fully taken into account, and on the other hand because optical filter devices also differ significantly. Summary of the Invention

[0003] Therefore, there is a need for a procedure that allows the evaluation of images of such surfaces that are as uniform or characteristic as possible. According to the present invention, this is achieved by the subject matter of the independent patent claims. Advantageous embodiments and further developments are the subject matter of the dependent claims.

[0004] In a method according to the invention for inspecting lacquered surfaces, preferably having one or more layers with absorbing and / or effect pigments, radiation is irradiated onto the surface to be inspected at a first predetermined angle of incidence by at least one first radiation device, and a color image recording device records spatially resolved images of the surface illuminated by the irradiation device at different observation angles, the image recording device having a first predetermined sensitivity that depends on the wavelength of the radiation impinging on the image recording device.

[0005] According to the invention, the image evaluation device performs a section-by-section, preferably pixel-by-pixel, evaluation of the image recorded by the image recording device, and the colour image recording device is also used to evaluate and / or assess absorption and / or effect pigments and / or the influence of the effect and / or absorption pigments on the image recording and / or integral colour measurement is taken into account and / or excluded as part of the image evaluation.

[0006] Effect pigments give coatings a strongly angle-dependent (flop effect) color impression. In contrast to classical pigments (absorption pigments), effect pigments often have a platelet-like geometry and their mode of operation is based on optical effects such as directional reflection or interference.

[0007] The present invention is therefore based on the idea that such effects or absorbing pigments affect the recorded color image in different ways, and therefore provides a new approach to separating color from absorbing pigments in such coatings.

[0008] It has now been proposed to use color imaging (RGB) cameras to quantitatively evaluate these data, particularly by observing only the image itself.

[0009] As mentioned above, the first approach is to use equalizer filters to fit the RGB channel sensitivity curves to the tristimulus curves of the human eye.

[0010] In the context of the present invention, an additional approach is proposed to distinguish high precision color measurements of absorption pigments from measurements of effect pigments.

[0011] In a preferred method, the classification of absorption and / or effect pigments is carried out with respect to at least one characteristic property of these absorption and / or effect pigments, which means that a distinction is made with respect to a specific property as to whether the absorption or effect pigment is observed.

[0012] In a further preferred method, clustering is performed. This can be performed according to different criteria. Clustering according to a predetermined characteristic characteristic is particularly preferred. Preferably, this characteristic characteristic is selected from a group of characteristics containing pigment size, a specific a* and / or b*. For example, the pigments seen in the image can be divided into different groups of pigment sizes, such that a first number n1 of effect pigments with a first specific pixel size and a second number n2 of effect pigments with a second pixel size are seen. The a* and b* values ​​can also be classified accordingly.

[0013] A preferred method distinguishes between absorption and effect pigments during image evaluation, e.g. the above mentioned values ​​L*a*b* are advantageously determined only for absorption pigments.

[0014] In a further preferred method, the different coatings are classified. Preferably, the spectrum is measured here at a high angle relative to the specular reflection (for example, an angle of 110°). The reason for using this angle is that effect pigments or flakes are expected to have the least influence on the overall measured spectrum.

[0015] Preferably, the coating is selected from the group of coatings comprising a silver metal coating, a chromatic metal coating, a Xylar or MICA coating or an interference coating.

[0016] Silver metal coatings have only aluminum pigments. These may contain small amounts of TiO2 or may be slightly colored. A gray-neutral spectrum is expected.

[0017] The chrome metal coating has a low percentage of aluminum and only a small amount of flakes are visible at 110°.

[0018] Xylar or MICA coatings have a very low percentage of effect pigments at 110° (<1%), with only a small amount of pigment or flakes visible.

[0019] In a further preferred method, the influence of at least some of the absorption and / or effect pigments is eliminated, especially as part of the image evaluation. This means that for the evaluation of the image, especially if the absorption or effect pigment reaches an intensity above (or below) a certain threshold, these reflections are eliminated. Thus, it is particularly preferred to create a threshold that eliminates, for example, reflections or shadows from the effect pigments. This is explained below with reference to the drawings.

[0020] In a further preferred method, interference features on the surface are identified, and surfaces with such interference features are preferably excluded from the evaluation. For example, areas of the surface with scratches or fingerprints can be excluded from the evaluation. This is based on the fact that in these cases, the visual impression is also significantly affected by these interference effects, such as scratches or fingerprints, and these are not due to the surface itself, but rather to the interference effects.

[0021] Image assessment is proposed to recognize features such as fingerprints or scratches and to exclude these areas from the overall assessment of the image, thus achieving a more accurate measurement.

[0022] Advantageously, the image evaluation device performs a section-by-section, preferably pixel-by-pixel, evaluation of the image recorded by the image recording device.

[0023] Preferably, the surface is an exterior surface of an automobile, in particular a lacquered exterior surface of an automobile, in particular a passenger car, however, other surfaces can also be investigated, such as furniture surfaces.

[0024] Preferably, the results of this evaluation are used and / or taken into account in (future) measurements by the equipment used for the evaluation. Preferably, a "filter device", in particular a software filter device, is determined and / or generated by the evaluation and taken into account and / or used in (future) measurements.

[0025] For example, as described above, the evaluation can be performed pixel by pixel with respect to the image recording device. This evaluation allows at least one calibration value to be assigned to each pixel or each region of pixels for future measurements. This calibration value is preferably determined as part of the evaluation for each individual pixel. The calibration value determined as part of the evaluation (particularly for each pixel) can also be taken into account when outputting measurement results for each individual pixel of the image recording device for future measurements using the device.

[0026] The applicant's prior art is known to arrange optical filter devices between a surface and an image recording device. This is also intended to compensate for the different evaluation characteristics of the human eye, on the one hand, and the camera, on the other. However, it has been shown that such filter devices themselves exhibit a high degree of scattering (in terms of their characteristics) and therefore result in different evaluations. In addition, corresponding lighting devices, such as LEDs, also suffer from strong scattering. This means that even two LEDs from the same manufacturer, which should be identical, differ from each other in terms of their beam characteristics. Furthermore, there is a high degree of variability in RGB filters from camera to camera and even within a single camera.

[0027] This requires specially adapted filters that take into account changes in camera or light source characteristics (e.g., when the camera chip or LED needs to be replaced). Additionally, the state of the art allows for only one standard illuminant. By introducing specially adapted filters, different standard illuminants can be mathematically taken into account.

[0028] The present invention therefore proposes a section-by-section, in particular pixel-by-pixel, evaluation of the image, in particular as a function of wavelength, in order to be able to adapt to the respective conditions, i.e. the specific radiation characteristics of the illumination device, and also the image recording device or its characteristics. The evaluation can be repeated, for example, at predetermined times. The results and / or measurements of the evaluation are preferably stored.

[0029] As is known from the state of the art, the image recording device has an image recording element with a plurality of image pixels, each suitable for detecting radiation incident thereon. For example, the image recording device can have a CCD chip. The evaluation is carried out for at least some of the pixels mentioned, preferably for at least 30%, preferably at least 50%, preferably at least 60%, particularly preferably at least 70% of the pixels mentioned. The evaluation can be carried out for each individual pixel, but it is also conceivable to combine several pixels for evaluation, which will result in a certain reduction in the resolution of the evaluation.

[0030] For example, this type of image analysis can be performed at predetermined time intervals.

[0031] In a preferred method, the measurement signals of the individual pixels are weighted to take into account the pixel-by-pixel evaluation. In this way, software-based filter devices can be used or generated, which in particular also influence the image evaluation of subsequent images.

[0032] In a preferred method, the evaluation is carried out as a function of the wavelength of the radiation impinging on the image recording device, which means that a wavelength-dependent evaluation of the sensitivity of the image recording device, in particular the sensitivity of each individual pixel, is recorded as a function of wavelength.

[0033] It is therefore particularly preferred that the evaluation is carried out as a function of the wavelength-dependent (especially pixel-by-pixel) sensitivity of the image-recording device. Preferably, an individual evaluation is carried out for each individual image-recording device. Preferably, this evaluation is also carried out pixel-by-pixel.

[0034] Preferably, the wavelength-dependent sensitivity of the image recording device is determined, in particular the (especially wavelength-dependent) sensitivity can be determined pixel by pixel or the wavelength-dependent sensitivity can be determined for each individual pixel.

[0035] However, it is also possible to carry out the evaluation over several pixels, for example by averaging several pixels of the same intensity.

[0036] In a further preferred method, an image recording device, in particular a color image camera, is also used to evaluate and / or assess the effect pigments.

[0037] In a further preferred method, the influence of effect pigments on image recording and / or integral color measurement is taken into account and / or eliminated, especially during image evaluation.

[0038] The current state of the art has the problem that integral color measurements are subject to errors because they cannot distinguish whether the resulting measurement results are due to the color of the flakes, the effect pigments, or another cause. The preferred proposed method makes it possible to make such a distinction. More precisely, spatially resolved color measurements are used for this purpose.

[0039] Integral color measurements can introduce errors, especially when the effect pigments themselves produce coloring effects that are different in color than those caused by absorption effects.

[0040] For example, a solid color with only one absorption pigment (e.g., solid red) and the same absorption pigment (e.g., red) with the addition of a colored effect pigment will result in different color values ​​XYZ being measured in an integral color measurement.

[0041] Multi-angle colorimeters, known from the state of the art, allow integrated, averaged, non-position-resolved color measurements over an entire measurement spot illuminated at several angles. In the history of device development, the first devices were without a camera, and later camera devices were added. The camera measurements are only used to evaluate "glossy" directional light (direct sunlight) and "granular" (diffuse illumination, cloudy weather), providing additional information unrelated to the color measurement for the characterization of effect pigment lacquers.

[0042] The use of a camera in a multi-angle colorimeter when measuring solid color lacquers (containing only absorbing pigments) is unnecessary and only makes sense for effect lacquers containing a mixture of absorbing pigments and one (or more) types of effect pigments.

[0043] The method proposed here eliminates the influence of the effect pigment measurement on the absorption pigment measurement, so that in the above example both measurements give the same value.

[0044] In a preferred method, the influence of effect pigment measurements on absorption pigment measurements is reduced and / or eliminated.

[0045] In a preferred embodiment, it is also proposed in the context of the present invention to evaluate and / or assess effect pigments, in particular their color properties, using an image recording device, in particular a color image camera.

[0046] However, in addition, it is preferable to obtain information about the colorimetry and / or color distribution of areas of the image that are due to and / or contain effect pigments, which means that the advantages achieved by using a color imaging camera can still be retained.

[0047] Preferably, the wavelength-dependent sensitivity is determined by a spectrometer and / or a monochromator and / or the evaluation of the image recorded by the image recording device is carried out by a spectrometer and / or a monochromator.Several procedures are conceivable for determining the spectral sensitivity of the image recording device, in particular of each individual pixel.

[0048] For example, it would be possible to obtain the characteristic curves of the individual channels of an RGB-CMOS / CCD camera chip, preferably with a Bayer pattern, as a sum over multiple wavelengths using the following equation:

number

[0049] where p is the measurement (red, green, blue). i )=s i represents the spectral sensitivity of the pixel / filter combination. E i,j is the wavelength l i 1 shows a calibration tile with known reflectance spectrum number j at

[0050] Additionally, it is also possible to perform multiple linear regression, which can be performed using the following formula:

number

[0051] In the context of the present invention, it is proposed to determine the spectral sensitivity of each pixel by means of a monochromator and / or a (particularly absolutely calibrated) spectrometer. Based on these recorded spectral sensitivity, deviations can in each case be determined, which can be taken into account in the subsequent image evaluation in order to record and / or output colorimetrically correct images of the individual pixels.

[0052] In a preferred method, to determine the wavelength-dependent sensitivity of an image recording device, radiation is applied to a surface at a predetermined angle on a series of reference surfaces with known reflectances, and the image recording device records an image of the surface, preferably at an angle greater than 20°, preferably greater than 30°, preferably greater than 40°, preferably greater than 50°, and particularly preferably greater than 60°, relative to the normal.

[0053] Additionally or alternatively, the surface can be illuminated with monochromatic light from further, particularly external, auxiliary light sources. These auxiliary light sources can be, for example, monochromatic LEDs or white light filtered using multiple bandpass filters. Again, the illumination angle is preferably greater than 20° relative to the vertical, preferably greater than 30°, preferably greater than 40°, preferably greater than 50°, and particularly preferably greater than 60°.

[0054] The reason for recording images with very large angles to the orientation of the observed surface or with very flat illumination is that these surfaces behave in the least distorted way under these illuminations, as defined by the effect pigments used in the coating. For example, a silver metal coating may consist only of aluminum flakes or partly of flakes with a certain percentage of TiO2.

[0055] In this case, a gray-neutral light directional spectrum can be expected under flat incident angles or illumination. Further surfaces have chromatic metal coatings containing aluminum, which usually contain only a few effect pigments. In this case, only a few of these flakes are visible under flat illumination angles. So-called Xylal or MICA coatings have even fewer effect pigments, i.e., in this case, none of these flakes are visible at the mentioned angles. Preferably, image evaluation is performed separately and / or independently for the absorbing pigments and the effect pigments (flakes).

[0056] Preferably, in the case of absorption pigment blends, pixel numbers are recorded and / or stored together with the intensity values ​​assigned or determined thereto (output by the associated pixel). In a further step, histograms can be recorded and the maximum values ​​of the respective frequencies can be determined. In a further step, mean values ​​XYZ are preferably recorded for statistically defined pixel numbers.

[0057] For the evaluation of effect pigments, flakes are preferably selected that are separate from one another and reach or cover all three filters (i.e., their emission characteristics or emission maxima are within the respective wavelength ranges of the associated filter devices of the image recording device), and product XYZ is determined only for these flakes. Preferably, no demosaicing is used in this case. Preferably, at least two images are recorded with a specific exposure time.

[0058] Preferably, the color effect of the absorption pigment is assessed in an image taken at a first predetermined angle, in particular an angle away from the gloss, so that distortion of the color measurement due to the effect pigment can be neglected to a good approximation. An angle away from the gloss is understood to be an angle that deviates from the direction of reflection by at least 30°.

[0059] The sparkle caused by the effect pigments is preferably discerned in a camera image taken at a second angle, in particular at a near-gloss angle, which is understood to be an angle that deviates from the direction of reflection by no more than 25°, preferably no more than 20°, preferably no more than 15°.

[0060] Due to the angle close to the gloss, effect pigments can be identified in the camera image as areas of high intensity (above a certain threshold), i.e. it is known with pixel accuracy whether it is an area on the sample with absorbing pigment or effect pigment.

[0061] In a further advantageous method, the sensitivity of the human eye, which depends on the wavelength of the radiation impinging on the human eye, is taken into account during evaluation and / or measurement with the device.

[0062] In a further preferred method, the data determined as part of the evaluation are taken into account to generate a filter device, in particular a software filter device, for subsequent measurements by the apparatus also performing the evaluation, which calibrates the measurements recorded or determined by the image recording device. Preferably, the recorded image is calibrated pixel by pixel and / or the measurements output by individual pixels are calibrated individually.

[0063] This difference between the first sensitivity (of the image recording device) and the second sensitivity (of the human eye) can be at least partially compensated for by one, in particular the filter device.

[0064] In a further preferred method, radiation is applied to the surface by a second radiation device and a second predetermined angle of incidence, and the image recording device records an image of the surface irradiated by the second radiation device. Alternatively, a second observation device can be used. In addition, a third radiation device is preferably provided for applying radiation to the surface to be investigated.

[0065] Illumination at different angles is particularly preferred. In a further preferred method, at least one radiation device emits directional or diffuse radiation onto the surface.

[0066] In a further preferred method, the data recorded during the evaluation are reduced, and this reduction is preferably different for absorbing pigments and effect pigments. The reduction can be performed, for example, when evaluating the wavelength-dependent sensitivity of a pixel or when evaluating impinging radiation, so that only wavelength ranges in which specific intensities (especially resulting from the spectral profile), such as (local) intensity maxima, occur are investigated. In this way, intensity limits can be determined that allow for the detection of flake-free areas. In this way, areas of the image that contain flake images or areas that do not contain flakes are preferably identified.

[0067] When observing surfaces, the problem arises that commercially available image recording devices, such as RGB cameras, have specific wavelength-dependent sensitivities that differ from those of the human eye. The goal is therefore to enable the most realistic possible image recording of the illuminated surface (or the most realistic possible evaluation of this image recording).

[0068] The invention therefore proposes to achieve at least partial adaptation of the image recording device to the human eye by means of a filter device (in particular a software filter device, in particular such a filter device that takes into account the data recorded during the evaluation).

[0069] The CIE standard valence system, or CIE color system, is a color system defined by the International Commission on Illumination (CIE) to establish the relationship between human color perception (color) and the physical causes of color stimuli (color valence). It captures the entire range of perceptible colors. Using color space coordinates, the term Yxy color space or CIE-Yxy is also commonly used, and primarily in English-speaking countries, tristimulus color space is also used.

[0070] In English-speaking countries, the three base values ​​X, Y, and Z are called tristimulus. In this sense, they are the three parts of a standardized set of basic colors defined for this purpose. Each color can be identified by a triplet of such numbers. Therefore, the term tristimulus system is commonly used for the CIE standard system. This curve is also called a tristimulus curve.

[0071] Thus, in one embodiment, the image is recorded and the individual pixels are evaluated, particularly with respect to color, and a wavelength-dependent evaluation and / or weighting is performed.

[0072] In a preferred method, the evaluation is performed in such a way that a wavelength-dependent difference between the first sensitivity (of the image recording device) and the second sensitivity (of the human eye) is at least temporarily compensated.

[0073] When selecting a filter device, the values ​​and / or curve characteristics of the emission spectrum L(λ) of the radiation device, the intensity curve I(λ) of the standard light, in particular at least one tristimulus function X(λ) of the human eye, and / or the filter characteristic F(λ) of the image recording device are particularly preferably taken into account.

[0074] Preferably, the wavelength dependent transmittance T(λ) is obtained from the filter device as follows: t(λ)=x(λ) / (i(λ)·l(λ)·f(λ))

[0075] I(λ) denotes the wavelength-dependent characteristics of the illuminant, e.g., D65, L(λ) denotes the wavelength-dependent characteristics of the light source, F(λ) denotes the wavelength-dependent characteristics of the viewing device (especially RGB filters and especially filters thereof), and X(λ) denotes the wavelength-dependent light reception rate (tristimulus function) of the eye.

[0076] Preferably, the wavelength-dependent properties of the viewing device and the wavelength-dependent light sensitivity of the eye have different functions over at least two, preferably three, predetermined wavelength ranges.

[0077] Preferably, the first wavelength range is 300 nm to 600 nm, preferably 350 nm to 550 nm, preferably 400 nm to 500 nm. Furthermore, the second wavelength range is preferably 400 nm to 700 nm, preferably 450 nm to 650 nm, preferably 500 nm to 600 nm. Furthermore, the third wavelength range is preferably 400 nm to 800 nm, preferably 500 nm to 700 nm, preferably 550 nm to 650 nm.

[0078] Preferably, the full range of human eye perception is covered.

[0079] The wavelength-dependent characteristics of the observation device can also be divided into the wavelength ranges described above.

[0080] In a further preferred method, radiation is emitted onto the surface at a second predetermined angle of incidence by a second radiation device, and the image recording device records an image of the surface illuminated by the second radiation device.

[0081] Preferably, the first and second radiation devices irradiate the surface at different times or periods. Alternatively or additionally, it is also conceivable that the second image recording device observes the surface at a second observation angle.

[0082] It is also possible to detect effects resulting from different arrangements of effect pigments by illuminating them using two or more illumination devices.

[0083] In a further preferred method, a third radiation device is also provided, preferably for emitting radiation onto the surface at a third angle of incidence.

[0084] In a further preferred method, the observation angle relative to the surface normal is less than 10°, preferably less than 5°, preferably less than 3°.

[0085] In a more preferred method, the first angle of incidence relative to the direction perpendicular to the surface is 70° to 20°, preferably 60° to 30°, and more preferably 50° to 40°.

[0086] Preferably, the second angle of incidence of the second radiation device relative to the direction normal to the surface is between 85° and 50°, preferably between 85° and 60°, preferably between 85° and 70°.

[0087] Preferably, the at least one radiation device directs directional or diffuse radiation onto the surface, with diffuse radiation being used to simulate solar radiation when the sky is overcast, and directional radiation being used to simulate solar radiation when the sky is clear.

[0088] Preferably, at least one further radiation device, preferably all radiation devices, direct diffuse or in particular directional radiation onto the surface.

[0089] The present invention is further directed to an apparatus for inspecting lacquered surfaces having a mixture of an absorbing pigment and at least one further effect pigment, comprising a first radiation device for irradiating radiation onto the surface to be inspected at a first predetermined angle of incidence, and a color image recording device for recording a spatially resolved image of the surface illuminated by the incident direction at a first observation angle, the image recording device having a first predetermined sensitivity that depends on the wavelength of the radiation incident on the image recording device.

[0090] According to the invention, the apparatus comprises an image evaluation device which performs a section-by-section, preferably pixel-by-pixel, evaluation of the image recorded by the image recording device.

[0091] Furthermore, the color image recording device can also be used to evaluate and / or assess absorption and / or effect pigments and / or the image evaluation is suitable and intended to take into account and / or exclude the influence of effect pigments on the image recording and / or integral color measurement.

[0092] In a preferred embodiment, the apparatus comprises a memory device in which the measurement values ​​determined by the evaluation device are stored. Preferably, the memory device allows these measurement values ​​to be stored pixel by pixel.

[0093] In a further preferred embodiment, the apparatus has a filter device, in particular a software filter device, which calibrates the further images recorded by the image recording device, in particular calibrates them taking into account the values ​​determined by the evaluation device, and / or is suitable and intended for this purpose.

[0094] Preferably, the filter device (and / or the processor device implementing this filter device) is suitable and intended to calibrate the recorded image pixel by pixel.

[0095] Preferably, the filter device is modifiable, i.e. in particular the way in which it influences the image output by the image recording device is modifiable, which means that by modifying the (software) filter device the image output by the image recording device and / or also the measurements output by the entire apparatus can be modified.

[0096] Preferably, the apparatus is operable in a calibration mode in which images recorded by the image recording device are evaluated and software filter devices are determined and / or modified. Preferably, the apparatus is also operable in an operation mode in which the software filter devices determined in particular in the calibration mode are used.

[0097] Preferably, the apparatus comprises a calibration device suitable and intended for calibrating individual pixels.

[0098] Particularly preferably, this is a calibration in the XYZ color space of an RGB camera.

[0099] Preferably, the apparatus is a multi-angle measuring device and is therefore suitable and intended to inspect surfaces at several (illumination and / or illumination) angles.

[0100] However, the invention can also be used in unilaquers (without effect pigments) for automobiles (or other surfaces).

[0101] Preferably, the radiation device and the observation device, and optionally also the filter device, are arranged in a common housing.Preferably, the inner wall of this housing is designed to be light-absorbing.In a more preferred embodiment, the housing essentially has only one opening through which the surface is observed.In a more preferred embodiment, the device is portable.

[0102] In a further preferred embodiment, the image recording device has a filter, in particular an RGB filter. Preferably, the radiation device emits standardized light, in particular D65 standardized light. Standardized light refers to the standardized spectral radiation distribution curve of a characteristic emitter. Standardized light type D65 is a radiation distribution with a color temperature of 6504 Kelvin (approximately corresponding to a gray, overcast sky).

[0103] In a preferred embodiment, the distance between the surface and the radiation device is between 3 cm and 30 cm, preferably between 4 cm and 20 cm, preferably between 4 cm and 10 cm.

[0104] In a preferred embodiment, the radiation device is suitable and intended to emit radiation of different wavelengths. A filter device can be provided, such as a filter wheel with different filters that allow only light of specific wavelengths to pass.

[0105] In a further preferred embodiment, the first radiation device comprises a light emitting diode (LED), in particular a triphosphor LED. Preferably, as described above, the apparatus also comprises further radiation devices. These also preferably comprise light emitting diodes, in particular triphosphor LEDs.

[0106] In a further preferred embodiment, the apparatus comprises at least one second radiation device and / or a second sensor device, which can be designed as an image recording device, but which can also be a sensor device for determining the intensity of the radiation impinging on it.

[0107] In a further preferred embodiment, the apparatus comprises at least three radiation devices (or illumination devices) which preferably illuminate the surface from at least three different angles.

[0108] In a further preferred embodiment, the filter device performs a pixel-by-pixel calibration of the values ​​or signals output by the individual pixels of the image recording device.

[0109] Further advantages and embodiments are shown in the accompanying drawings. [Brief explanation of the drawings]

[0110] [Figure 1] 1 is a schematic diagram of an apparatus according to the present invention; [Figure 2]FIG. 1 is a diagram showing the spectral characteristics of RGB filters of a digital camera. [Figure 3] FIG. 1 shows the sensitivity curves of three color receptors X (red), Y (green) and Z (blue). [Figure 4] FIG. 1 is a diagram showing the radiant flux of standardized illuminant D65. [Figure 5] FIG. 1 is a diagram showing the emission spectrum of an LED. [Figure 6] FIG. 1 shows the transmission behavior of a filter device. [Figure 7a] FIG. 10 shows a comparison of the obtained sensitivities. [Figure 7b] FIG. 10 shows a comparison of the obtained sensitivities. [Figure 7c] FIG. 10 shows a comparison of the obtained sensitivities. [Figure 8] FIG. 10 shows a comparison of theoretical and actual strength curves. [Figure 9] FIG. 1 is a diagram illustrating the deviation between the theoretical course and the actual course. [Figure 10] Image of silver metal coating at 15°. [Figure 11] Image of silver metal coating at 80°. [Figure 12] Image of ultra red metallic coating at 15°. [Figure 13] Image of ultra red metallic coating at 80°. [Figure 14] RGB camera image of a crystal silver xylaric coating with a black base color at a 15° observation angle. [Figure 15] RGB camera image of a crystal silver xylaric coating with a white base color at a 15° observation angle. [Figure 16] RGB camera image of a crystal silver xylaric coating with a black base color at an 80° observation angle. [Figure 17] RGB camera image of a crystal silver xylaric coating with a white base color at an 80° observation angle. [Figure 18] A more detailed image of the ultra-red metallic coating at 80°, where the effect pigments are visible as bright reflections (orientations) or dark dots. [Figure 19] A more detailed image of the ultra-red metallic coating at 80°, where digital image processing has been used to remove effect pigment areas oriented in the reflective direction of the camera. [Figure 20] In addition to Figure 19, here is a more detailed image of the ultra-red metallic coating at 80°, where the misoriented effect pigment areas (visible as dark dots) have also been removed using digital image processing. [Figure 21] FIG. 10 is an illustration for calibrating individual pixels. [Figure 22] FIG. 10 is an illustration for calibrating individual pixels. [Figure 23] FIG. 10 is an illustration for calibrating individual pixels. [Figure 24] FIG. 10 is an illustration for calibrating individual pixels. [Figure 25] FIG. 1 is a diagram showing the Hyperrot test procedure. [Figure 26] FIG. 1 is an explanatory diagram of a test method for Mamba Green. [Figure 27] FIG. 1 shows a comparison of an Ultra Red surface without effect pigments and a surface with effect pigments. [Figure 28] This is the RGB histogram for ultra red. [Figure 29] FIG. 1 shows a comparison of a Mamba Green surface without effect pigments and a surface with effect pigments. [Figure 30] This is the RGB histogram of Mamba Green. [Figure 31] 1 is an image of a super red surface with a fingerprint. [Figure 32] Photograph of the scratched surface. [Figure 33] FIG. 1 shows a comparison of a scratch-free surface and a scratched surface. [Figure 34] FIG. 1 depicts a camera-based measurement of background color. [Figure 35] FIG. 1 is an illustration of a measurement arrangement. [Figure 36] FIG. 1 is a diagram illustrating a measurement arrangement. [Figure 37] FIG. 2 is a further illustration of the measurement arrangement. DETAILED DESCRIPTION OF THE INVENTION

[0111] 1 shows a schematic diagram of an apparatus 1 for inspecting a surface 10. The apparatus comprises a first radiation or illumination device 2 which projects light onto the surface 10 in a beam S2.

[0112] Reference number 4 denotes an image recording device which records at least one spatially resolved image of the surface illuminated by the first radiation device (beam path S4). Reference number O denotes an opening in the housing 12 through which the surface 10 is illuminated and through which the image recording device 4 observes the surface. The image recording device records images at an observation angle of 0°, i.e. is positioned vertically above the surface 10.

[0113] Reference number 12 denotes an optionally available filter device arranged in the beam path S4 between the surface 10 and the image recording device, with which the image recording device records an image of the surface 10.

[0114] Reference numeral 14 denotes an optionally available lens device which functions to collimate light reflected and / or scattered by surface 10 so that it impinges on the filter device in collimated form, preferably perpendicular to the filter device.

[0115] Reference number 20 denotes an evaluation device for evaluating the image recorded by the image recording device 4. The evaluation device is preferably capable of outputting data that are characteristic of the physical properties of the surface.

[0116] Reference number 22 identifies a processor device which calibrates and / or corrects the images recorded by the image recording device in the operating mode of the apparatus, in particular calibrating and / or correcting them pixel by pixel and / or taking into account the data determined by the evaluation device, and which therefore preferably determines the above-mentioned software filter device.

[0117] Reference number 6 denotes a second radiation device which emits radiation, in particular light, onto the surface (at a different angle of incidence or along beam path S2), which can in particular be used to evaluate the recorded image.

[0118] Reference number 8 denotes a third radiation device which emits radiation, in particular light, along a beam path S3 onto a surface 10.

[0119] Preferably, a control device (not shown) is provided which activates the radiation devices 2, 6 and 8 with a time delay.

[0120] Figure 2 shows the characteristics of an image recording device as a function of the wavelength of the incident radiation, or more precisely, the sensitivity of the RGB filters of this image recording device or camera.

[0121] Three curves are shown, R, G, B, which refer to the "red", "green" and "blue" components. The quantum efficiency (%) is plotted on the axis, with the wavelength of the incident light plotted on the vertical axis.

[0122] It can be seen that the quantum efficiency of the camera as a whole first increases in the wavelength range of 400 nm to 800 nm, then decreases again. Thus, the image recording device has a unique characteristic of image reproduction or image recording.

[0123] Figure 3 shows the tristimulus function of the human eye. Here again, three curves are shown: x(λ), y(λ) and z(λ), with wavelength (nm) on the vertical axis and tristimulus values ​​on the coordinates.

[0124] A comparison of the illustrations shown in Figures 2 and 3 shows that the wavelength-dependent sensitivity curves of image recording devices and the human eye are quite different. These differences are at least partially compensated for by the present invention.

[0125] Figure 4 shows a diagram of the intensity curve of a D65 standardized light source in the range of 300 nm to 800 nm. This type of light is approximated by the curves of daylight and cloudy skies. The second curve, A, shows the curve of a conventional incandescent lamp.

[0126] Standardized illuminant type D represents the daylight spectrum and is therefore particularly important for many industrial areas. The D65 illuminant derives its name from its color temperature of 6,504 Kelvin (K). D65 is used in the chemical and pharmaceutical industries, paint manufacturing, and the ceramic, textile, paper, and automotive industries.

[0127] The standardized illuminant type D65 has a high blue content, allowing the perception of fluorescent colors.

[0128] The D65 light source is used as the evaluation light source. The spectral distribution of the D65 light source is defined in DIN 5033 and lies between 300 nm and 780 nm, i.e., between ultraviolet and red.

[0129] Figure 5 shows the emission spectrum of a light source preferably used in the context of the present invention, namely a triphosphor high CRI LED. It can be seen that this light source emits essentially between 400 and 800 nm. The color temperature here is 5600 K. This emission characteristic is also preferably taken into account in the design of the filter device.

[0130] The abbreviation CRI stands for Color Rendering Index. Color rendering index is a quantitative measure of a light source and its ability to reproduce the colors of an object compared to an ideal or natural light source. The term CRI is often used in commercial lighting products. A precise definition should be called Ra - general color rendering index - or Ri - specific color rendering index - depending on the test color sample being evaluated.

[0131] CRI is calculated by comparing the color rendering of a test light source to that of a defined light source. For test light sources below 5000K, a blackbody spotlight is used as the defined comparison source. Daylight (D illuminant) is used for comparison of test light sources above 5000K. The calculation of Ri and Ra is explained in detail in the CIE 13.3-1995 Technical Report. The test method uses a set of 8 Ra or 14 Ri CIE-1974 color samples from an early version of the Munsell color system. The first 8 samples are moderately saturated, contain a color tone circle, and have approximately equal luminance. The remaining 6 samples provide additional information about the color rendering of the light source.

[0132] Figure 6 shows a transmission curve for a filter device according to the present invention, calculated from the above data and the formula shown above. Based on this data, a filter device is preferably manufactured that exhibits approximately the transmission behavior shown in Figure 6. In manufacturing a filter device, there are several ways to achieve the desired transmission curve, as described above. Additionally, a software filter can be created based on these values, or the above adjustments can be made on the software side. Such adjustments can also be made on a pixel-by-pixel basis.

[0133] Figures 7a to 7c show three representations of the progression (plotted in arbitrary coordinate units). Figure 7b shows again the progression of the human eye, also shown in Figure 3. Figure 7c shows the curve obtained from the image recording device proposed according to the present invention without a filter device. Figure 7a shows the sensitivity or curve obtained when a filter device is used. It can be seen that the curve shown in Figure 7a is much closer to the "natural" curve shown in Figure 7b than the curve shown in Figure 7c.

[0134] Figure 8 shows a diagram illustrating the method according to the invention. It shows a comparison between the curves shown in Figures 7b and 7c in more detail. It can be seen that these curves are close to each other in some wavelength ranges, but quite different in others. The tristimulus curves X, Y, Z and the resulting sensitivity curves B, G, R are shown on the other hand.

[0135] Figure 9 shows a diagram showing the percentage deviations diff x, diff y, diff z of the curves from each other, where we can also see that some regions have large deviations and other regions only small deviations.

[0136] As mentioned above, in this case the influence of individual flakes is very small, so the measured spectrum is recorded at a very flat angle of incidence.

[0137] The values ​​of X, Y and Z can be determined using the following formulas:

number

[0138] The following applies to k:

number

[0139] I(l) denotes the wavelength-dependent relative intensity of the normalized illuminant type. L(l) denotes the wavelength dependent intensity of the radiation device.

[0140] In a further process step, the area of ​​this maximum is selected for evaluation of the absorbing pigment, and the values ​​L*a*b are calculated and averaged over a sufficient number of pixels within this area. As mentioned above, this procedure can be used to identify areas of the image that reproduce flakes and areas that do not.

[0141] To evaluate a flake or a layer containing flakes, as described above, a separate flake is preferably selected, for example, a particular region of a pixel can be assigned to the flake.

[0142] FIG. 9 shows the area of ​​the histogram representing the absorption pigments and the area representing the effect pigments.

[0143] Figures 10-13 show different images of the surface. Figure 10 shows a silver metal surface recorded at a 15° angle, while Figure 11 shows the same surface recorded at an 80° angle. Significant differences can be seen between the two images. Figure 12 shows the ultra-red metal layer recorded at 15°, while Figure 13 shows the ultra-red metal layer at 80°. Again, considerable differences can be seen in the image recordings. In particular, it can be seen that the effect pigment's influence is much less noticeable in the 80° angle than in the image taken at 15°.

[0144] Figures 14-17 show RGB camera images of a crystalline silver xylaric coating at 15° and 80°, showing black as the absorbing pigment-based color (Figures 15 and 17) and white as the absorbing pigment-based color (Figures 16 and 18). Here, it can be seen that the effect pigment is clearly visible at 15°, but has no or virtually no effect in the image captured at 80°. As mentioned above, one aspect of the present invention is to match the image recording device to the characteristics of the human eye. However, this creates a number of problems. First, there are multiple filter curves that vary relative to each other. LED manufacturing also introduces significant differences in image recording characteristics.

[0145] A further problem is the temperature-dependent changes of light-emitting diodes and their radiation activity.

[0146] The RGB filters also vary from camera to camera.

[0147] Therefore, if the camera chip or LED curve changes, a new filter design is required. Care should be taken to have only standard illumination and standard observation devices. In addition, compatibility between different observation devices or equipment becomes an issue.

[0148] Figure 18 shows a more detailed view of the ultra-red metallic coating recorded at 80°, where some prominent effect pigments can still be discerned.

[0149] Figure 19 shows a representation where the RBG values ​​above the maximum value of the histogram assigned to correctly oriented effect pigments have been set to their maximum values. In this representation, it can be seen that the correctly oriented effect pigments no longer appear or have been "removed." Only the incorrectly oriented effect pigments can still be recognized.

[0150] Figure 20 shows further processing of the image data based on the data of Figure 20, whereby RBG values ​​below the maximum value of the histogram assigned to the misoriented effect pigment were set to their maximum value. In this representation, it can be seen that the misoriented effect pigment no longer appears or has been "removed."

[0151] Figure 21 shows the ultra-red histogram for the three channels (RGB). Individual curves are labeled.

[0152] The calibration of individual pixels is described below. In the xyz color space, the following relationships are obtained for x, y, and z:

number

[0153] The following applies:

number

[0154] The following relationships apply to RGB camera values:

number

[0155] JPEG2025541991000009.jpg49170

number

[0156] k G and k B The corresponding relationship is obtained for

[0157] where L*, a* and b* can be calculated for each individual pixel (using hyperspectral imaging):

number

[0158] The following applies:

number

[0159] This gives us L*, a* and b*:

[0160] FIG. 22 again shows the TRI stimulation values ​​plotted against wavelength.

[0161] Figure 23 shows the z B , y G and x R The values ​​are shown.

[0162] FIG. 24 shows the sensitivity of the spectrometer, which is also considered.

[0163] To determine the system-wide parameters of each individual pixel, it is proposed to illuminate a diffusing sphere with monochromatic light with a half-maximum wavelength of 10 nm from 400 nm to 700 nm in 10 nm increments.

[0164] The sphere has a measurement input and an external spectrometer. In addition, multiple images with a specific resolution are recorded at different wavelengths to cover the visible spectral range, for example, 25-35 images each with a resolution of 5 MP.

[0165] The spectral sensitivities of all 5 million pixels are then preferably calculated.

[0166] For temperature calibration, the spectral distribution of the light-emitting diode is recorded at different temperatures between 10 °C and 40 °C, preferably in a climatic chamber.

[0167] This results in the diagram shown in Figure 24, which shows the measured sensitivity curves of the individual RGB channels of the camera sensor.

[0168] The result is less pixel-to-pixel variation in the spectral sensitivity curves, as indicated by the individual bars. This results in a total of three average sensitivity curves for each RGB channel, which fully represent or evaluate a particular RGB camera chip.

[0169] There are many advantages if the actual physical parameters of the LEDs, filters and camera are known.

[0170] First, it can compensate for variations in LEDs and SPDs from different manufacturers. It can also compensate for filter variations. Furthermore, it can compensate for variations related to the camera, which makes it possible to change cameras in a measurement setup.

[0171] Furthermore, a real calibration of the entire system is possible.

[0172] Different standard illuminations can also be calculated.

[0173] Additionally, this adjustment between different measuring devices is improved. Finally, the temperature deviation of the LEDs is also taken into account.

[0174] A general description of this method is provided below.

[0175] In this method, a sufficient amount of a known effect solution that does not contain effect pigments is mixed. This amount is then divided into two (particularly equal) parts or two (particularly equal) containers. In a further step, the effect pigments are added to one of the two containers, especially in the correct proportions.

[0176] Here, there are two color mixtures that are identical except for the addition of an effect pigment to one of the two parts.

[0177] These facts are illustrated in FIGS.

[0178] Furthermore, two mixtures with the same application parameters are preferably sprayed onto two test surfaces. Measurement of these two test surfaces with an integrated, non-spatially resolving spectrophotometer reveals a clear deviation in color tone, expressed as dE*3.78, which can also be observed directly with the eye. The addition of effect pigments not only produces the desired sparkle effect, but also changes the overall color impression of the sample.

[0179] In a further method, the absorption pigment is measured equally in the formulation with and without the effect pigment using an RGB camera. Figure 27 shows the corresponding curves, with measurements taken at 80° with and without the effect pigment. In each case, both curves are labeled.

[0180] FIG. 28 shows the corresponding RGB histogram for ultra-red.

[0181] In the table below, the values ​​L*, a* and b* were calculated from integral color measurements based on spectral data at an angle of 110°:

[0182] [Table 1]

[0183] In the table below, the values ​​L*, a* and b* were calculated from RGB camera image data when illuminated with an 80° LED.

[0184] [Table 2]

[0185] In this way, the influence of the effect pigment can be eliminated, which serves to measure the color of the absorbing pigment essentially independently of the concentration of the effect pigment. This is demonstrated by the fact that the deviation in measurements between the two samples "Ultra Red with effect" and "Ultra Red without effect" is only dE*=0.44, instead of dE* of 3.78.

[0186] More precisely, the influence of the effect pigment can be screened out by comparison in order to determine as far as possible the color of the absorbing pigment independent of the concentration of the effect pigment. The same calculation can be performed when the color is Mamba Green, as shown in Figures 29 and 30.

[0187] Figure 31 shows an image of an ultra-red surface with a fingerprint, and Figure 32 shows an image of an ultra-red surface with a scratch. In the following, we propose to discard these as well.

[0188] As a result, the following values ​​are obtained for an unscratched surface, scratches and fingerprints:

[0189] Integral color measurement at 110°:

[0190] [Table 3]

[0191] Camera-based measurement of background color at 80°:

[0192] [Table 4]

[0193] Figure 33 shows the corresponding images of these three surfaces, where the differences can be seen especially in the range from 640 nm onwards.

[0194] Corresponding measurements are also determined for camera-based measurements of background color, again allowing the effects of fingerprints and scratches to be recognized.

[0195] FIG. 35 shows the corresponding RGB histograms for the three cases spectrally depicted in FIG. 34: intact sample, scratched and fingerprinted.

[0196] In this way, it is possible to clearly identify blemish-causing data (such as scratches, fingerprints, etc.) in the RGB camera histogram and reject them accordingly.

[0197] Figures 35 to 37 show three different measurement arrangements. In the measurement arrangement shown in Figure 35, the illumination starts at +65°, i.e., an angle of 65° (counterclockwise) from the normal to the surface.

[0198] Therefore, the reflection occurs at an angle of -65°. The color is captured at an angle of +45°, and therefore at an angle of 110° to the reflected radiation.

[0199] In the measurement arrangement shown in Figure 36, illumination is again at +65°, but observation by the RGB imaging camera is at +15°, and therefore at an angle of 80° to the reflected radiation.

[0200] In the measurement arrangement shown in Figure 37, illumination can be performed at three angles, namely +65°, but also at 0° and -30°. Here, color detection is again performed at +45°. The dashed lines at +30°, 0°, and -65° indicate the reflected beam paths of the individual illuminations in each case. It can be seen that measurements can be made in this way at several different angles, depending on which illuminations are activated.

[0201] The applicant reserves the right to claim all features disclosed in the application as essential to the invention, provided that they, individually or in combination, are new compared to the prior art. It should also be noted that each figure describes features that may be advantageous in their own right. Those skilled in the art will readily recognize that a particular feature described in a figure may be advantageous without adopting additional features from that figure. Furthermore, those skilled in the art will recognize that advantages may result from a combination of several features shown in each figure or in different figures.

Claims

1. A method for inspecting a lacquered surface (10), preferably having one or more layers with absorbing and / or effect pigments, in which radiation is irradiated by a first radiation device (2) at a first predetermined angle of incidence (a1) onto the surface (10) to be inspected, and a color image recording device (4) records a spatially resolved image of said surface illuminated by the incident direction at a first observation angle (b), said image recording device (4) having a first predetermined sensitivity (F(l)) depending on the wavelength of said radiation impinging on said image recording device, 1. A method according to claim 1, wherein an image evaluation device performs a section-by-section, preferably pixel-by-pixel, evaluation of the image recorded by the image recording device, and wherein the color image recording device (4) is also used to evaluate and / or assess the absorption and / or effect pigments and / or the influence of effect pigments on the image recording and / or integral color measurement is taken into account and / or excluded as part of the image evaluation.

2. characterised in that the classification of the absorption and / or effect pigments is carried out with respect to at least one characteristic property of these absorption and / or effect pigments, The method of claim 1.

3. Cluster formation is performed, The method of claim 2.

4. wherein the characteristic property is selected from the group of properties including pigment size, a* value and b* value; The method of claim 3.

5. characterised in that a distinction is made between absorbent pigments and effect pigments in the evaluation The method according to at least one of claims 1 to 4.

6. characterised in that the influence of at least a part of the absorption and / or effect pigments is eliminated, The method according to at least one of claims 1 to 5.

7. Interference features located on the surface are identified, and preferably surface areas having such interference features are ignored in the evaluation. The method according to at least one of claims 1 to 6.

8. characterised in that the wavelength-dependent sensitivity of the image recording device is determined by a spectrometer and / or a monochromator and / or the evaluation of the image recorded by the image recording device is carried out by a spectrometer and / or a monochromator. The method according to at least one of claims 1 to 7.

9. In the evaluation, the sensitivity of the human eye (X(λ)) depending on the wavelength of the radiation incident on the human eye is taken into account. The method according to at least one of claims 1 to 8.

10. radiation is irradiated onto the surface at a second predetermined angle of incidence (a2) by a second radiation device (14), and the image recording device records an image of the surface irradiated by the second radiation device (14), The method according to at least one of claims 1 to 9.

11. characterized in that the observation angle (b) relative to the direction normal to the surface (10) is less than 10°, preferably less than 5°, preferably less than 3°, and / or the first angle of incidence relative to the direction normal to the surface is between 70° and 20°, preferably between 60° and 30°, preferably between 50° and 40°, The method according to at least one of claims 1 to 10.

12. An apparatus (1) for inspecting a lacquered surface (10), preferably having one or more layers with absorbing and / or effect pigments, comprising a first radiation device (2) for irradiating radiation onto the surface (10) to be inspected at a first predetermined angle of incidence (a1), and a color image recording device (4) for recording a spatially resolved image of the surface illuminated by the incident direction at a first observation angle (b), said image recording device (4) having a first predetermined sensitivity (F(l)) depending on the wavelength of the radiation incident on said image recording device, 1. Apparatus (1), characterized in that said apparatus comprises an image evaluation device (20) which performs a section-by-section, preferably pixel-by-pixel, evaluation of said image recorded by said image recording device, said color image recording device (4) being also used to evaluate and / or assess said absorption and / or effect pigments and / or said image evaluation being suitable and intended to take into account and / or eliminate the influence of effect pigments on image recording and / or integral color measurement.

13. characterised in that the device (1) is suitable for and comprises a calibration device intended for calibrating each individual pixel, 13. The device (1) according to claim 12.

14. a filter device for performing pixel-by-pixel calibration of the values ​​or signals output by the individual pixels of the color image recording device; 14. Device (1) according to claim 12 or 13.

Images