Method and device for scanning an object using an imaging technique

Abstract

Method for scanning an object (3) using an imaging technique, wherein the method comprises the steps: - Simultaneous illumination of an object (3) with at least two different light sources (17, 26, 30), wherein the light sources (17, 26, 30) have different light spectra (33, 34, 35), - Scanning the object (3) using at least two sensors which have different spectral sensitivities (36, 37, 38), and - Separating the intensities of the light spectra (33, 34, 35) of the at least two light sources (17, 26, 30) in the sensor signals detected by the sensors using specific sensitivities of the respective sensors for the different light spectra.

Description

[0001] The present invention relates to a method and a device for scanning an object using an imaging method, wherein in particular at least two different light sources are used to illuminate the object to be scanned.

[0002] When inspecting products using imaging techniques, the illumination and, in particular, the angle of incidence at which the light shines onto an object to be scanned is of central importance for the detection of defects.

[0003] WO 2009 / 018 849 A1 describes an arrangement for capturing images of an object using an image capture device and an illumination device with at least one light source for illuminating the object. The image capture device has multiple channels for separately capturing different image color components (R, G, B). Several light sources emit colored light corresponding to the image color components (R, G, B). According to the method, the object is illuminated with light of the different spectral colors (R, G, B), and simultaneously, separate images corresponding to the spectral colors are captured by the image capture device.

[0004] US Patent 2006 / 0007458A1 describes a method for performing chromatic matching. This method predicts the final color data observed under a final light source from the initial color data observed under an initial light source. To do this, the initial and final light spectra are first received, and the spectral sensitivity functions of the colors between which chromatic matching is to be performed are determined. A linear chromatic matching transformation is then determined by minimizing the residual error between the color data predicted by the chromatic matching transformation and the final color data. Different spectral bands of the illumination spectra and the spectral sensitivity functions are weighted based on the contribution of each spectral band to a predefined function.The process then applies the linear chromatic adjustment transformation to the original color data.

[0005] German patent DE 10 2014 115 540 A1 relates to a camera for capturing objects within a detection area. The camera comprises a light receiver with a multitude of light-receiving pixels, a multispectral illumination device for illuminating the detection area in different spectral bands, and an evaluation unit. The detection area is illuminated by the illumination device in different spectral bands during various illumination time windows. Image data from the detection area is captured by the light receiver within these time windows. The light receiver is designed as a double-line receiver with a first and a second line of light-receiving pixels, with the acquisition time windows for the first and second lines being different.

[0006] German patent DE 102 39 225 A1 describes a method for the optical inspection of banknotes, in which several spectrally distinct light sources are used and their outputs are superimposed to produce homogeneous light. First, the outputs of light-emitting diodes are superimposed, and the light is directed onto a banknote via lenses. Each lens has a different spectral characteristic. For each specific inspection, i.e., for each type of banknote to be examined and for each type of optical sensor used, illumination with a suitable spectral composition can be provided. The corresponding device for the optical inspection of banknotes comprises at least one optical sensor and the illumination, which is formed from more than three light sources with the differently limited spectra.

[0007] Basically, there are three types of lighting: 1. Darkfield illumination 2. Brightfield illumination 3. Transmitted light lighting

[0008] In dark-field illumination, the light from the light source is deflected by means of beam-directing devices (reflector, lens) in such a way that direct (specular) reflection of the light is avoided. In bright-field illumination, the light emitted by the light source is

[0009] The object is directed at an angle so that it is reflected towards a camera.

[0010] Transmitted light illumination is used for scanning transparent objects, with the light source located on one side of the object and the sensors for detecting the light passing through the transparent object located on the other side of the object.

[0011] For materials that exhibit both diffuse and glossy reflective properties and are also transparent, different types of defects can be detected simultaneously using the three illumination methods. For example, bright-field illumination can reveal defects in a paint structure, dark-field illumination can detect defects in the paint application, and transmitted light illumination can reveal inclusions in a transparent area.

[0012] These types of lighting and variations thereof are known to be used to automatically detect a wide variety of product defects.

[0013] It is often advantageous to provide two different types of lighting simultaneously. Since the light from one type of lighting overlaps with the light from the other when illuminated simultaneously, it is not easily possible to use two different types of lighting at the same time. Therefore, the individual lighting types are often used in time-division multiplexing, meaning that the individual lighting types are switched on sequentially. This prevents the individual lighting types from interfering with each other. However, the possible throughput when scanning objects is reduced, as the object cannot be scanned simultaneously with all lighting types.

[0014] It is also known to design the individual lighting types with spectrally different light sources and to provide several cameras equipped with special filters, so that each camera can only receive the light of a specific lighting type. Such a device has the advantage that an object to be monitored can be scanned simultaneously with several lighting types. The disadvantage is that multiple cameras are required and the filters must be of high quality to precisely separate the light from the different light sources.

[0015] In principle, it would also be conceivable to use a color camera, which typically has different sensors, each sensitive to a specific color or spectral range. If light sources were used, each detected by a different sensor type within the color camera, then the individual lighting types could be spectrally separated without additional filters. However, this is not possible with conventional, commercially available color cameras because the individual spectral sensitivity ranges of different sensor types overlap, meaning that each sensor type detects light from light sources with different spectra. This makes it impossible to unambiguously assign the sensor signals to a specific light source. This effect is referred to as spectral coupling.

[0016] The invention is based on the objective of creating a method and a device for scanning an object using an imaging method, in which the object can be simultaneously irradiated with different illumination cycles and the light of the respective illumination types can be reliably detected using simple and cost-effective means.

[0017] The problem is solved by the items specified in the independent claims. Advantageous embodiments are specified in the respective dependent claims.

[0018] An inventive method for scanning an object using an imaging method comprises the following steps: - Simultaneous illumination of an object with at least two different light sources, where the light sources have different light spectra, - Scanning the object using at least two sensors that have different spectral sensitivities, and - Separating the light spectra of at least two light sources in the sensor signals detected by sensors using specific sensitivities of the respective sensors for the different light spectra.

[0019] The inventor of the present invention recognized that the spectral coupling described above at the sensors leads to output signals which, while not directly usable, allow the light intensities caused by these light spectra in the sensor signals to be separated based on the specific sensitivities of the respective sensors to the different light spectra of the light sources. Each sensor type exhibits a specific sensitivity for the respective light spectrum emitted by a light source. With two light sources and two sensors, this results in four specific sensitivities, which can be represented in a 2 x 2 matrix. These specific sensitivities can be easily measured beforehand by independently directing light from each light source onto the sensors and recording the intensity of the sensor signals.Multiplying the inverse matrix of specific sensitivities, referred to below as the separation matrix, by the vector of sensor signals yields a vector describing the intensities of the light from each light source at the sensors. This provides an intensity value for each light source, representing the light that has reached the sensors. The intensities caused by each light source at the sensors are thus separated.

[0020] This eliminates the need for multiple cameras and high-quality spectral filters. Light can be detected using a conventional camera equipped with several different sensors exhibiting varying spectral sensitivities. Such cameras are readily available, mass-produced items at low cost. The sensor signals simply need to be separated into their individual light spectra using a separation matrix.

[0021] Such matrix multiplication is often integrated into cameras to correct color reproduction. Instead of a conventional correction matrix for color reproduction, the invention uses a separation matrix to separate the individual light spectra of the different light sources. In a simple embodiment, existing hardware can thus be used, although the use of the separation matrix achieves a completely different effect.

[0022] The different light sources preferably have beam-directing devices with different geometries. Beam-directing devices include reflectors and lenses. Bright-field illumination preferably has a reflector and / or a lens that directs the light in a beam that is as straight as possible onto the object being scanned. In dark-field illumination, one or more reflectors are provided that direct the light onto the object being scanned from as many different directions as possible. Accordingly, the geometries of the reflectors and lenses in bright-field and dark-field illumination differ.

[0023] Preferably, the object is scanned with a color camera that has sensors with at least three spectrally different light sensitivities. The sensors are typically sensitive to red, green, or blue light. Such color cameras are available on the market in a wide variety of configurations, including line scan cameras and area scan cameras. They are manufactured in large quantities, which is why they are inexpensive. These cameras have a multitude of sensors with the same spectral sensitivity. If such a camera extends in one or two dimensions, the individual sensors can be illuminated with different light intensities due to the geometry of the light beam guides. This usually results in illumination inhomogeneities, where the light intensity is lower at the edges than in the center. The sensor signals can be corrected spatially using appropriate correction factors.Such a correction is called shading correction.

[0024] Furthermore, the individual sensors of a camera can exhibit significant variations in their respective sensitivities. These differences can be as high as 10%. This effect is often referred to as PRNU (photo response non-uniformity). This effect can also be corrected with appropriate correction values. These must be determined separately for each sensor element. Since this PRNU correction value can differ for each sensor, it is sensor-dependent and therefore also location-dependent.

[0025] Since the deviations due to the PRNU effect are caused by the sensors themselves, the measured sensor values ​​(s r , s g , s b ) are first corrected with the PRNU correction coefficients, then with the separation matrix (M SEPThe light signals must be separated and then corrected using correction factors for illumination inhomogeneities to obtain the intensities of the light arriving at the sensors from each light source. Illumination inhomogeneities cause a different distribution of light intensity at the individual sensors. Since spectral crosstalk depends on light intensity, this correction can only be performed on the already separated intensity values. Mathematically speaking, this means that the correction for illumination inhomogeneities, the separation using the separation matrix, and the PRNU correction are not commutative. Therefore, the order of these operations cannot be reversed.

[0026] Conventional cameras often correct for the effects of lighting (shading effect) and sensitivity differences (PRNU effect) using a single set of common correction coefficients. This is not possible here without considering the separation matrix.

[0027] However, it is possible to specify a location-dependent separation matrix for each sensor of a camera, which is already corrected according to the PRNU effect and / or illumination inhomogeneities.

[0028] The invention is explained in more detail below by way of example with reference to the accompanying drawing. The drawing shows: Fig. 1 Schematic representation of a device for scanning objects which are moved along a conveying direction in a side view, Fig. 2 the device from Fig. 1 in a front view, Fig. 3 the device from Fig. 1 and Fig. 2 in a sectional view along line AA in Fig. 2, Fig. 4 a diagram showing the light spectra of different light sources and the spectral sensitivities of different sensors, and Fig. 5 schematically a device for scanning objects with multiple lighting devices in a side view.

[0029] An embodiment of a device 1 according to the invention for scanning objects is arranged adjacent to a conveyor belt 2, on which objects 3 to be scanned rest and are transported in a conveying direction 4 ( Fig. 1, Fig. 2) The conveyor belt 2 consists of two endless transport belts 5, each guided around rollers 6 at its end, so that the transport belts 5 each form an upper run 7 and a lower run 8. The transport belts 5 are arranged parallel to each other and spaced apart. The objects 3 to be scanned rest on the upper run 7 of the transport belts 5.

[0030] In the present embodiment, the scanning device 1 comprises an upper scanning unit 9 and a lower scanning unit 10.

[0031] A lighting device very similar to the upper scanning unit 9 is described in DE 10 2015 101 252 A1, which is why reference is made to this patent application.

[0032] This upper scanning unit 9 has two reflector arrangements 11, 12 ( Fig. 3) Each of the reflector arrangements 11, 12 is formed from a trough-shaped reflector 13 and a flat reflector 14. The trough-shaped reflector 13 has a mirrored reflector surface 15 with an elliptical cross-section.

[0033] A light source holder 16, projecting from the side of the reflector surface 15, is arranged at each edge of the trough-shaped reflector 13. Light-emitting diodes 17 are provided as light sources on the light source holder 16. The light-emitting diodes 17 are arranged on the light source holder 16 such that their emission direction is directed towards the reflector surface 15. The flat reflector 14 adjoins the edge of the trough-shaped reflector 13 opposite the light source holder 16. The flat reflector 14 has a mirrored reflector surface 18, which is arranged on the same side as the reflector surface 15 of the trough-shaped reflector 13.

[0034] The two reflector arrangements 11, 12 are arranged opposite each other with their reflector surfaces 15, 18. Furthermore, the two reflector arrangements 11, 12 are arranged symmetrically with respect to a plane of symmetry 19 of the scanning device 1.

[0035] The upper scanning unit 9 has a housing 20 with two longitudinal side walls 21, two diametrically opposed end walls 22, and a bottom wall 23. The flat reflectors 14 are arranged perpendicular to the bottom wall 23. The bottom wall 23 is recessed in the area between the two reflector arrangements 11, 12, forming a passage opening 24 through which light can escape.

[0036] Adjacent to the plane of symmetry 19 is a mounting rail 25, which extends between the opposing end walls 22. Light-emitting diodes 26 are arranged in a row on the mounting rail 25. The light-emitting diodes 26 are directed towards the passage opening 24 and form a bright-field illumination.

[0037] The light emitted by the LEDs 17 is first reflected by the adjacent trough-shaped reflector 13. From the trough-shaped reflector 13, a large portion of the light is directed onto the flat reflector 14 of the opposite reflector arrangement 11, 12. A small portion of the light emitted by the LEDs 17 is also directed directly from the trough-shaped reflector 13 to the opening 24 and exits to the outside.

[0038] The light reflected by the flat reflector 14 is mostly directed through the opening 24. A small portion is directed back to the opposite flat reflector 14 and reflected again before exiting through the opening 24. These singly, doubly, and triply reflected light rays form a diffuse light with homogeneous illuminance. A predetermined, strip-shaped area adjacent to the opening 24 is uniformly illuminated. This strip-shaped area is referred to below as the illumination area 27.

[0039] The light-emitting diodes 17 together with the reflector arrangements 11, 12 form a dark-field illumination.

[0040] This dark-field illumination is, on the one hand, very compact and, on the other hand, produces a very homogeneous illumination in the illumination area 27. Within the scope of the invention, other reflector arrangements can of course be provided which also produce diffuse illumination to form a dark-field illumination.

[0041] The scanning device 1 is arranged with its plane of symmetry 19 transverse to the conveying direction 4.

[0042] The objects 3 to be scanned rest on the two upper sections 7 of the transport belts 5. The upper sections 7 thus define a transport plane 28.

[0043] Below the transport level 28, a line-shaped light source arrangement 29 is arranged in the lowest scanning unit 10, extending over the entire length of the illumination area 27, i.e., in the area between the opposing end walls 22 of the upper scanning unit 9. The light source arrangement 29 comprises LEDs 30, which are arranged in a line and oriented with their emission direction towards the passage opening 24. The LEDs 30 of the lower scanning unit 10 thus emit light upwards through the passage opening 24.

[0044] The upper scanning unit 9 includes a mounting 31 for a camera 32. The camera 32 is preferably a line scan camera, extending in particular over the entire length of the scanning device 1, i.e., between the opposing end walls 22, in order to scan the entire illumination area 27. However, the camera 32 can also be an area scan camera.

[0045] Camera 32 can be equipped with a zoom lens. Camera 32 is generally oriented towards the opening 24.

[0046] An object 3 located below the passage opening 24 is illuminated with dark-field light by the LEDs 17, with bright-field light by the LEDs 26 and with transmitted light by the LEDs 30.

[0047] The light sources for dark-field illumination (LEDs 17), bright-field illumination (LEDs 26), and transmitted-light illumination (LEDs 30) are each designed to emit light with a different light spectrum. In the present embodiment, the LEDs 17 for dark-field illumination emit blue light, the LEDs 26 for bright-field illumination emit green light, and the LEDs 30 for transmitted-light illumination emit red light. Fig. Figure 4 shows the blue light spectrum 33, the green light spectrum 34 and the red light spectrum 35 of the different light-emitting diodes.

[0048] Camera 32 is a color camera which has sensors with different spectral sensitivities. In the present embodiment, camera 32 is an RGB camera which has sensors with blue, green, and red spectral sensitivities. Fig. Figure 4 shows the blue spectral sensitivity 36, the green spectral sensitivity 37, and the red spectral sensitivity 38. As can be seen from the diagram... Fig. As can be seen from Figure 4, the spectral sensitivities 36, 37, and 38 overlap considerably. In particular, the green spectral sensitivity 37 strongly overlaps with the blue spectral sensitivity 36 and the red spectral sensitivity 38, and extends into the regions of the blue light spectrum 33 and the red light spectrum 35. However, the blue spectral sensitivity 36 and the red spectral sensitivity 38 also extend into the regions of other colored light spectra. This causes the spectral coupling described at the beginning.

[0049] Spectral coupling can be expressed using the following formula: (srsgsb)=[k00k01k02k10k11k12k20k21k22]×(aragab) where a vector s with coefficients s r , s g , s b the sensor signals for red, green and blue for a pixel, the vector a with the coefficients a r , a g , a bthe response of the object to be scanned 3 to the illumination with the three light spectra red, blue and green and the matrix M K with the coefficients k00 to k22 representing the coupling of the different light spectra.

[0050] The answers a r , a g , ab physically represent the light intensities of the light reflected or transmitted by the scanned object from the three spectrally different light sources at the respective pixel of the sensor.

[0051] The coefficients k00 to k22 of the coupling matrix M K The specific sensitivities of the red sensor (k00, k01, k02) for the red, blue, and green light spectrum, the green sensor (k10, k11, k12) for the red, blue, and green light spectrum, and the blue sensor (k20, k21, k22) for the red, blue, and green light spectrum are physically represented. Formula (1) can therefore also be written as follows: s=MK×a

[0052] A separation matrix M SEP is considered the inverse matrix of the coupling matrix M K defined: MSEP=MK−1

[0053] This means: a=MSEP×s where the signal intensities (s r , s g , s b ) into the responses or light intensities (a r , a g , a b ) are converted or separated at the respective pixel of the camera.

[0054] The coupling coefficients (k00-k22) can be determined using a calibration procedure. In this procedure, bright-field illumination, dark-field illumination, and transmitted-light illumination are calibrated independently of each other.

[0055] To calibrate the brightfield illumination, a flat, mirrored object 3 is used, the LEDs 26 of the brightfield illumination are switched on, and the signals from the red, green, and blue sensors are recorded. These signal values ​​form the corresponding coupling coefficients.

[0056] To calibrate the dark-field illumination, an object 3 with a flat, white surface is used. This object 3 is illuminated with light from the LEDs 17, and the corresponding signals for the colors red, green, and blue are generated by the camera. These signal values ​​correspond to the respective coupling coefficients for the dark-field illumination.

[0057] During calibration for transmitted light illumination, no object is placed in the illumination area 27, allowing the light from the lower scanning unit 10 with the LEDs 30 to reach the camera 32 unimpeded. The signals for the colors red, green, and blue, detected by the camera, form the coupling coefficients for the transmitted light illumination.

[0058] During calibration, a signal vector s is generated for each pixel. Preferably, during calibration, all signals for brightfield illumination, darkfield illumination, and transmitted light illumination are averaged separately, and the average values ​​are used as coupling coefficients. The different efficiencies of the illumination types are taken into account by selecting the illuminance or luminance so that they generate approximately the same signals at the camera. Alternatively, it may be advantageous to normalize the coupling coefficients for brightfield illumination, darkfield illumination, and transmitted light illumination with specific normalization coefficients for each, which account for the different, inherent light intensities during the three calibration processes.

[0059] Are the coupling coefficients, and thus the coupling matrix M, K If the separation matrix M is known, then the inverse matrix can be calculated to determine the separation matrix M. SEP be determined.

[0060] Cameras often already have a module for multiplying the signal vector s with a matrix to correct color reproduction. In the invention, the separation matrix M is used as the matrix in such a module. SEP used, thereby the sensor signals (s r , s g , s b ) into the responses or light intensities (a r , a g , a b ) the individual spectrally different light sources are separated.

[0061] Optionally, light intensity correction can be performed. A fundamental distinction is made between correcting for illumination inhomogeneity and correcting for the different sensitivities of the sensors (PRNU).

[0062] The illumination inhomogeneities arise from the geometry of the light sources and the light beam deflection devices, which generate a light beam that is generally not perfectly homogeneous. This results in the different pixels of camera 32 being illuminated with varying light intensities. To describe the light intensities due to the illumination inhomogeneities, an attenuation vector sb is used. x with the coefficients sbr x , sbg x and SBB xThese attenuation coefficients differ for the various light spectra of different light sources and are location-dependent. The following description applies to a one-dimensional line scan camera, in which several pixels are arranged next to each other in the X direction. In an area scan camera, the pixels are distributed in the X and Y directions, so that the attenuation coefficients depend on both the X and Y directions.

[0063] A method using the location-dependent damping factors sbr x , sbg x , sbb x Response vector a corrected for illumination inhomogeneities bx This results as follows: abx=(arbxagbxabbx)=(ar×sbrxag×sbgxab×sbbx)

[0064] Since the response vector a is calculated from the signal vector s in the procedure for sampling the objects, the components of the response vector a calculated with formula (4) must be multiplied by the reciprocals of the damping factors to obtain the appropriately corrected responses from the signals.

[0065] This is described by the following formula: (aragab)=[1sbrx0001sbgx0001sbbx]×MSEP×(srsgsb)

[0066] The individual sensors of a camera can have different sensitivities. These sensitivities can differ by up to 10%. This is known as PRNU (photo response non-uniformity). The sensor signals are consequently distorted, which is compensated for by corresponding damping factors. x , seg x and Seb x can be expressed. The actually measured sensor signals s excan be calculated from the actual target sensor signals s, taking into account the location-dependent and light-spectra-specific attenuation coefficients, as follows: sex=(srexsgexsbex)=(sr×serxsg×segxsb×sebx)

[0067] The calculation of the answers a from the actually measured signals s ex This is done taking both corrections into account according to the following formula: (aragab)=[1sbrx0001sbgx0001sbbx]×MSEP×[1serx0001segx0001sebx]×(srexsgexsbex)

[0068] Since the correction of illumination inhomogeneities affects the light directed to the sensors, and the correction of differential sensitivities (PRNU) affects the detected sensor signals, the response calculation must first correct the sensor signals using the inverses of the attenuation values ​​for the differential sensitivities, then separate them into the corresponding responses using the separation matrix, and finally correct them using the inverses of the attenuation values ​​for the illumination inhomogeneities. These operations are not commutative, meaning they cannot be interchanged. Within the scope of the invention, it may also be advantageous to perform no correction or only one correction, either the correction of illumination inhomogeneities or the correction of differential sensitivities. Both corrections are location-dependent. The separation matrix M SEP is not fundamentally location-dependent.

[0069] Determining the answers (a r , a g , a b The response vector can be implemented in hardware using multiple multiplication stages, where the damping factors are location-dependent, i.e., stored separately for each sensor in a lookup table and read out to correct the sensor signals of the respective sensor. Alternatively, the determination of the response vector can be implemented using a software module that performs the corresponding matrix and vector multiplications.

[0070] The separation matrix M SEP The matrix has the coefficients m00, m01,... m22. Using the matrix in terms of its coefficients, the formula (8) given above can be expressed and expanded as follows: (aragab)=[m00sbrx×serxm01sbrx×segxm02sbrx×sebxm10sbgx×serxm11sbgx×segxm12sbgx×sebxm20sbbx×serxm21sbbx×segxm22sbbx×sebx]×(srexsgexsbex)

[0071] Applying this formula minimizes the number of calculations required. A pre-corrected separation matrix is ​​stored in a lookup table for each sensor. These separation matrices are themselves location-dependent due to the position-dependent attenuation factors; therefore, a separate separation matrix is ​​stored in the lookup table for each sensor.

[0072] As can be seen from formula (9), the attenuation values ​​are in the denominator of the coefficients, which is why they act as amplification factors and amplify the actually measured signals accordingly.

[0073] The methods described above allow the use of light sources with different light spectra and the separation of the sensor signals such that a separate intensity signal is determined at each sensor, reflecting the intensity caused by each of the light sources. This also allows the use of sensors exhibiting spectral coupling. No additional color filters are required. Other scanning methods can also be implemented using a device suitable for carrying out the method according to the invention, particularly when using additional light sources emitting white light. In such cases, a color correction matrix is ​​used instead of a separation matrix to produce approximately color-accurate images.

[0074] The embodiment described above uses a special lighting device for illuminating a strip-shaped area across a surface. The invention can also be implemented with other lighting devices. Fig. Figure 5 schematically shows in a side view another embodiment which has four lighting devices 39-42 to illuminate a common line-shaped area of ​​an object plane 43.

[0075] The lighting devices 39-42 are designed in a linear configuration and each has light sources arranged in a linear configuration, such as light-emitting diodes, and corresponding light-directing means, which may be mirrored reflectors and / or lenses or objectives.

[0076] This scanning device 1 further comprises a camera 32, which in the present embodiment is a line scan camera with which the object plane 43 is scanned line by line. The line scan camera 32 is oriented with its optical plane 44 perpendicular to the object plane 43. It is also possible, in principle, to arrange the line scan camera 32 such that the optical plane 44 forms an angle with the object plane 43 that deviates from a right angle. A semi-transparent mirror 45 is arranged in the region of the optical plane 44 between the object plane 43 and the line scan camera 32. A first of the illumination devices 39 is arranged such that the light emitted by it is directed onto the semi-transparent mirror 45 and is directed by the mirror to the line-shaped illumination area 46 in the object plane 43.The corresponding beam of light from the semi-transparent mirror 45 to the linear illumination area 46 runs approximately parallel to the object plane 43.

[0077] The illumination devices 40, 41 are arranged such that they illuminate the line-shaped illumination area 46 with the emitted light beam, the light beams being directed at the object plane 43 at an angle that does not form an angle of incidence or reflection with the optical plane 44 of the camera 32, so that the light emitted by the illumination devices 40, 41 at the object plane 43 is not directed directly into the camera 32. The illumination devices 40, 41 thus produce dark-field light, with illumination device 40 emitting the light at a shallower angle than illumination device 41 onto the object plane 43.

[0078] The lighting device 42 is arranged below the object plane 43 and is aligned such that the light emitted by the lighting device 42 illuminates the linear lighting area 46.

[0079] The lighting devices 39-42 are designed such that at least two of them emit light with different light spectra. Preferably, each lighting device 39-42 produces light with a different light spectrum. By using the different light spectra, the illumination area 46 can be illuminated simultaneously by the various lighting devices 39-42, and the intensities of the light spectra of the different lighting devices 39-42 can be separated, as explained in detail above with reference to the first embodiment.

[0080] According to the invention, two different lighting devices or light sources 39-42 are sufficient to simultaneously perform two different optical examinations on an object to be scanned. For this purpose, the object to be scanned is arranged in the object plane 43 within the lighting area 46. Reference symbol list: 1 Device 2 Conveyor belt 3 objects 4 Direction of conveyance 5 transport belts 6 rolls 7 Tower 8 Trum 9 upper scanning unit 10 lower scanning unit 11 Reflector arrangement 12 Reflector arrangement 13 trough-shaped reflector 14 flat reflector 15 mirrored reflector surfaces 16 light source holders 17 light-emitting diodes 18 mirrored reflector surfaces 19 Plane of symmetry 20 cases 21 Longitudinal side wall 22 Front wall 23 Floor wall 24 Through opening 25 retaining rail 26 light-emitting diodes 27 Lighting area 28 Transport level 29 Light source arrangement 30 LEDs 31 bracket 32 camera 33 blue light spectrum 34 green light spectrum 35 red light spectrum 36 blue spectral sensitivity 37 green spectral sensitivity 38 red spectral sensitivity 39 Lighting equipment 40 Lighting equipment 41 Lighting equipment 42 Lighting equipment 43 Object level 44 optical plane 45 semi-transparent mirror 46 Lighting area

Claims

[1] Method for scanning an object (3) using an imaging technique, wherein the method comprises the steps: - Simultaneous illumination of an object (3) with at least two different light sources (17, 26, 30), wherein the light sources (17, 26, 30) have different light spectra (33, 34, 35), - Scanning the object (3) using at least two sensors which have different spectral sensitivities (36, 37, 38), and - Separating the intensities of the light spectra (33, 34, 35) of the at least two light sources (17, 26, 30) in the sensor signals detected by the sensors using specific sensitivities of the respective sensors for the different light spectra. [2] Method according to claim 1, characterized by, that the separation is carried out by multiplying the simultaneously acquired sensor signals of the at least two sensors using a separation matrix, wherein the separation matrix is ​​the inverse matrix of a matrix describing the specific sensitivities. [3] Method according to claim 1 or 2, characterized by , that the light sources (17, 26, 30) have light beam guiding means (13, 14) with different geometries. [4] Method according to any one of claims 1 to 3, characterized by , that the object is scanned with a color camera (32) which has sensors with at least three spectrally different sensitivities (36, 37, 38) which are particularly sensitive to the colors red, green and blue. [5] Method according to any one of claims 1 to 4, characterized by , that the spectral sensitivities (36, 37, 38) of the respective sensors partially overlap spectrally. [6] Method according to any one of claims 1 to 5, characterized by , that the spectral sensitivities (36, 37, 38) of the respective sensors each comprise one of the light spectra (33, 34, 35) of the light sources (17, 26, 30). [7] Method according to any one of claims 1 to 6, characterized by , that several sensors of the same spectral sensitivity are used, whereby the signals determined with the individual sensors are corrected in a location- or sensor-specific manner. [8] Method according to any one of claims 1 to 7, characterized by , that the signals determined with the individual sensors are corrected with respect to the different sensitivities of the individual sensors (PRNU). [9] Method according to any one of claims 1 to 8, characterized by , that the separated intensities of the light spectra are corrected with respect to the illumination inhomogeneities. [10] Device for scanning an object (3) using an imaging technique comprising: - at least two different light sources (17, 26, 30) for simultaneously illuminating an object (3) with different light spectra (33, 34, 35), - at least two sensors for scanning the object (3) which have different spectral sensitivities (36, 37, 38), and - Means for separating the light spectra (33, 34, 35) of the at least two light sources (17, 26, 30) in the sensor signals by means of specific sensitivities of the respective sensors for the different light spectra (33, 34, 35). [11] Device according to claim 10, characterized by , that the device (1) is designed to carry out a method according to one of claims 1 to 7. [12] Device according to claim 10 or 11, characterized by , that the light sources include light-emitting diodes (17, 26, 30). [13] Device according to any one of claims 10 to 12, characterized by that the at least two light sources are designed as at least two of the following three light sources: dark-field light source, bright-field light source, transmission light source

Images