Apparatus, apparatus and method for calibrating digital camera of image processing system

Abstract

The invention provides a calibration device to improve calibration of a digital camera of an image processing system. The information memory contains a predetermined arrangement characteristic describing an arrangement of the calibration features on the calibration device, and contains a reference target position of a reference calibration feature of the calibration features on the calibration device. When the calibration device is used, a reference target position of a predetermined arrangement characteristic and a reference calibration characteristic can be read from the information memory by the digital camera. When the calibration device is used, a target position of a further reference calibration feature of the calibration features on the calibration device can be ascertained by the evaluation unit from the predefined arrangement characteristic and the reference target position, and the information memory contains a correction parameter for each of the reference calibration feature and / or the further calibration feature, a relative deviation between the known actual position of the corresponding calibration feature and the target position is described, from which the calibration parameters can be read by the digital camera when the calibration device is used.

Description

Field of the Invention

[0001] The present invention relates to a calibration device, in particular a calibration plate, for determining at least one calibration parameter for calibrating a digital camera of an image processing system, wherein calibration features and at least one information memory are arranged on the calibration device, and the at least one information memory is inseparably connected to the calibration device, and the at least one information memory contains pre-given arrangement characteristics that describe the arrangement of the calibration features on the calibration device. The present invention also relates to a device and a method for using the calibration device. Background Art

[0002] Different image processing systems are used in industrial image processing. In particular, these image processing systems are applied in the field of automation, such as in automated equipment (pick and place), virtual reality (VR), augmented reality (AR), measurement technology (2D or 3D measurement), robotics, etc. Also included are so-called vision system processes, such as inspection tasks, image or symbol coding, measurement of object orientation, measurement of object size, or many other tasks.

[0003] An image vision system generally consists of a shooting unit, such as a digital camera having a lens and an image sensor, an additional lighting device, a control unit for controlling image shooting, and an analysis and evaluation unit for processing the captured image data for further use. The analysis and evaluation unit can also be part of the digital camera, or the digital camera can also perform certain preprocessing on the captured image data. (Additional) subsequent processing of the captured images may be required, such as for filtering, preparing, or correcting the image data. The image data that may be post-processed can also be analyzed and evaluated, such as for detecting measurement values, determining the position, location, or orientation of an object, etc.

[0004] For accurate and reliable image processing applications, the digital camera is calibrated by the image processing system because otherwise it will lead to deviations in measurement results and loss of accuracy. Optical imaging errors (so-called aberrations) such as distortion, curvature, etc. can also be determined during calibration. Calibration data or calibration parameters for the digital camera of the image processing system are determined during calibration, which are used to correct the captured image data (for example, to compensate for curvature in the image). The calibration parameters are stored in the image processing system.

[0005] Calibration devices, in particular calibration plates, are used to calibrate the digital cameras of image processing systems. The basic structure of such a calibration device consists of a body (e.g., a plate) of a material (such as metal, plastic, ceramic, glass or composite material) and its spatial extent, as well as one or more calibration features that are captured by the digital camera of the image processing system when the calibration device is in use. These calibration features can be, for example, patterns composed of different geometric and colored graphics, usually circular (for very precise calibration) or, for example, rings or squares. The calibration features are provided on the calibration device or machined into the calibration device with a defined spatial extent (e.g., diameter) and arrangement relative to each other (e.g., a two-dimensional arrangement in x and y positions). A possible implementation is, for example, a checkerboard arrangement of alternating white and black squares as calibration features.

[0006] For calibration, the calibration device is arranged, for example, on a machine in a measurement plane and captured by the digital camera of the image processing system in order to determine calibration parameters from the image data, with the aid of which the relationship (e.g., a mathematical correlation) between the coordinates in the image, such as pixels, and the spatial coordinates in the space where the calibration device is located, also called world coordinates, is generated.

[0007] The calibration parameters basically consist of intrinsic parameters and extrinsic parameters. The extrinsic parameters describe the position and orientation of the digital camera in space and are used to convert the position of an object in space from three-dimensional world coordinates (e.g., x, y, z) to the three-dimensional coordinates of the digital camera of the image processing system in space. The intrinsic parameters describe the characteristics of the optical components of the digital camera of the image processing system and the position of the image sensor. For this purpose, the intrinsic parameters convert the position of an object from the three-dimensional coordinates of the digital camera of the image processing system in space to the two-dimensional coordinates (e.g., x, y) of the image sensor and the image plane, and thus describe the internal geometry of the digital camera of the image processing system. This also includes correct imaging, e.g., without image distortion or warping. Thus, as described above, the correlation between the world coordinates and the coordinates in the image is determined.

[0008] To improve the accuracy of such calibration, in addition to the calibration device, information (e.g., in the form of data) that can be used to compensate for manufacturing inaccuracies of the calibration device is usually also provided. However, problems can arise here if the data is incorrectly assigned to a specific calibration device, e.g., due to human error during application. In these cases, incorrect calibration occurs and thus incorrect results in image processing. In addition, more information, such as the spatial extent of the calibration device or the focal length of the optics of the digital camera, must be manually entered or provided as a basis for calibration.

[0009] In addition to the risk of incorrect calibration and other error-proneness, the effort for calibration also increases because this additional information must be stored in a manner that can be safely distributed for further use with a specific calibration device. This is the case, for example, if a malfunctioning digital camera is replaced at a later point in time and the new digital camera is calibrated.

[0010] Another possibility lies in directly providing the information just described on the calibration device in the form of a two-dimensional code (e.g., a QR code). These two-dimensional codes contain, for example, information about the calibration device, such as the product name, serial number, or characteristics of the calibration features (e.g., the number of geometric figures), which are captured by the digital camera and read by the analysis and evaluation unit from the image processing system or the digital camera itself. In US 9,230,326 B1, the position and / or size of the two-dimensional code are stored in the two-dimensional code on the calibration device, where this position serves as a reference position for calibration.

[0011] Furthermore, DE 11 2018 002 048 T5 discloses a calibration device with calibration features, on which a two-dimensional code is also provided. Here, the re-measured position of the calibration features on the calibration device is stored in the two-dimensional code. These re-measured positions are measured during or after the manufacture of the calibration device and saved as absolute measurement values in the two-dimensional code. Thus, manufacturing inaccuracies of the calibration features and / or the calibration device can be taken into account. The disadvantage of absolute measurement data is the high storage requirement in the two-dimensional code, especially when there are multiple calibration features on the calibration device, because the re-measured position of each calibration feature must be given.

[0012] To more precisely calibrate the digital camera of an image processing system, specific optical characteristics of the digital camera are required, such as the exact focal length of the optics. This information can be obtained in a so-called factory calibration (as disclosed, for example, in EP 3 557 523 B1). Here, a correction model is obtained during the manufacture of the image processing system for correcting imaging errors. Imaging errors belong to the internal parameters and are, for example, deviations of the image principal point (describing the optical axis from the image sensor to the optics) or image distortion caused by the optics or mechanical design of the digital camera. If these detailed information are missing, the calibration device must be captured and measured by the digital camera in different orientations (tilted) during calibration to ensure sufficient accuracy of the calibration. These additional steps during calibration result in slowness and increased complexity and error-proneness. Summary of the Invention

[0013] Therefore, the object of the present invention is to improve the calibration of the digital camera of an image processing system, in particular to simplify and optimize the calibration.

[0014] According to the present invention, the task of a calibration device for calibrating a digital camera of an image processing system is solved by a reference target position of a reference calibration feature including calibration features in at least one information memory on the calibration device, wherein when using the calibration device, the pre-given arrangement characteristics and the reference target position of the reference calibration feature can be read by the digital camera of the image processing system from the at least one information memory, so that when using the calibration device, the target position of at least one additional calibration feature of the calibration feature can be obtained by the analysis and evaluation unit of the image processing system from the pre-given arrangement characteristics and the reference target position of the reference calibration feature, and the at least one information memory respectively includes correction parameters for the reference calibration feature and / or at least one additional calibration feature, and the correction parameters describe the relative deviation between the known actual position and the target position of the corresponding calibration feature, wherein when using the calibration device, the correction parameters can be read by the digital camera of the image processing system from the at least one information memory. Since the correction parameters contain information about the calibration device, such as measurement data from measuring the geometry of the calibration device and / or calibration features after manufacturing (such as deviations from manufacturing errors, manufacturing defects, etc.), the calibration accuracy of the digital camera of the image processing system is improved. By including the correction parameters as relative deviations in the information memory and the target position of at least one additional calibration feature is obtained by the analysis and evaluation unit of the image processing system, compared with storing absolute positions, the storage requirement of the information memory on the calibration device can be significantly reduced.

[0015] According to the present invention, the correction parameters additionally describe the relative deviations of the size and / or shape and / or orientation of the reference calibration feature and / or at least one additional calibration feature.

[0016] In a preferred embodiment of a calibration device for calibrating a digital camera of an image processing system, at least one information memory can be optically read by means of the digital camera. Since the digital camera needs to take at least one image of the calibration device for performing calibration, it is advantageous that the digital camera also reads at least one information memory here. Alternatively, at least one information memory can also be implemented as an RFID that can be read by the digital camera.

[0017] In a preferred embodiment of a calibration device for calibrating a digital camera of an image processing system, the pre-given arrangement characteristics, the reference target position of the reference calibration feature and the correction parameters are stored in the at least one information memory as two-dimensional codes, preferably QR codes. This type of two-dimensional code provides more storage space especially compared with one-dimensional codes and can be reliably or quickly read by the digital camera. In addition, QR codes can be read very robustly and can be read by the digital camera even in case of damage.

[0018] In another embodiment, at least one information memory additionally contains the identification code of the calibration device. After calibrating the digital camera of the image processing system, at least one calibration parameter can be unambiguously assigned to the identification code of the used calibration device. Thus, at a later point in time, for example when re-determining at least one calibration parameter, the used calibration device can be unambiguously identified. In addition, by means of the identification code, it is also possible to ensure that the pre-given arrangement characteristics in the information memory and the reference target positions of the reference calibration features are consistent with the used calibration device.

[0019] In another embodiment, at least one information memory additionally contains the position of at least one information memory on the calibration device. The known position of at least one information memory can be used, for example, to define a reference point for the orientation of the calibration device and / or to determine the orientation of the calibration device, for example in an image of the calibration device taken by the digital camera.

[0020] In the device according to the invention for calibrating the calibration device of the digital camera of an image processing system, the digital camera of the image processing system is set up to read the pre-given arrangement characteristics, the reference target positions of the reference calibration features and the correction parameters from at least one information memory of the calibration device, and to transmit at least one image of the calibration device, the pre-given arrangement characteristics, the reference target positions of the reference calibration features and the correction parameters to the analysis and evaluation unit of the image processing system, wherein the analysis and evaluation unit is configured to determine the target positions of at least one further calibration feature from the pre-given arrangement characteristics and the reference target positions of the reference calibration features and to analyze and evaluate at least one image of the calibration device and the reference target positions of the reference calibration features, the target positions of at least one further calibration feature and the correction parameters from at least one information memory. The correction parameters improve the accuracy of the calibration because, for example, there are measurement data from measuring the geometry of the calibration device and / or the calibration features after manufacture (such as deviations due to manufacturing inaccuracies, manufacturing defects, etc.).

[0021] In an advantageous embodiment of the analysis and evaluation unit, the analysis and evaluation unit is set up to receive at least one camera parameter from a factory calibration and to analyze and evaluate it to determine at least one calibration parameter. By means of at least one camera parameter, imaging errors in the images taken by the digital camera during calibration can be corrected without tilting the calibration device, for example image distortion caused by the optics and / or mechanical design of the digital camera. Thereby, the accuracy of the calibration is improved.

[0022] Here, at least one camera parameter from factory calibration is preferably the pre-given focal length of the digital camera of the image processing system. Manual input of the focal length of the digital camera is no longer required for calibration. Thus, calibration can be performed more quickly and the error-proneness due to incorrect input of the focal length of the digital camera can be avoided. Additionally, the tolerance of the pre-given focal length from factory calibration is less than the tolerance during manual input, thereby improving the accuracy of calibration.

[0023] The digital camera and / or the analysis and evaluation unit advantageously store and use the determined at least one calibration parameter to correct the images taken by means of the digital camera. Precise and reliable image processing applications by means of the image processing system can be achieved with the corrected images and deviations or loss of accuracy of the measurement results during image processing applications can be avoided.

[0024] In a method according to the invention for image correction of images taken by means of a digital camera of an image processing system or for correcting at least one characteristic of an object in an object image taken by means of a digital camera of an image processing system, the digital camera reads a pre-given arrangement characteristic, a reference target position of a reference calibration feature, and a correction parameter from at least one information memory and transmits them to the analysis and evaluation unit of the image processing system, wherein the analysis and evaluation unit determines the target position of at least one additional calibration feature from the pre-given arrangement characteristic and the reference target position of the reference calibration feature, and analyzes and evaluates at least one image of the calibration device and the reference target position of the reference calibration feature, the target position of at least one additional calibration feature, and the correction parameter from at least one information memory to determine at least one calibration parameter, and the digital camera and / or the analysis and evaluation unit store and use the determined at least one calibration parameter to correct the images taken by means of the digital camera, or the digital camera and / or the analysis and evaluation unit determines at least one characteristic of the object from the object image taken by means of the digital camera and stores and uses the determined at least one calibration parameter to correct at least one characteristic of the object. The accuracy of image correction is improved by the correction parameter, for example because measurement data from the geometry of the measurement calibration device and / or calibration features after manufacturing (such as deviations due to manufacturing inaccuracies, manufacturing defects, etc.) are used.

[0025] Advantageously, the analysis and evaluation unit receives at least one camera parameter from factory calibration, and determines at least one calibration parameter for calibrating the digital camera of the image processing system from the at least one camera parameter, the at least one image of the calibration device, and the reference target positions of the reference calibration features, the target positions of at least one additional calibration feature, and the correction parameters from at least one information memory. Imaging errors in the images captured by the digital camera, such as image distortion caused by the optics and / or mechanical design of the digital camera, can be corrected with the at least one camera parameter. Thus, during calibration, it is not necessary to photograph and measure the calibration device at different orientations (tilts) of the digital camera to ensure sufficient accuracy of calibration. By eliminating these additional steps during calibration, the duration for calibration can be shortened and the complexity and error-proneness of calibration can be reduced.

[0026] Preferably, the digital camera and / or the analysis and evaluation unit stores and uses the determined at least one calibration parameter to correct the object image captured by the digital camera. With the corrected object image, accurate and reliable image processing applications with the image processing system can be achieved, and deviations or loss of accuracy in the measurement results during image processing applications can be avoided.

[0027] Preferably, at least one calibration parameter is determined for each of the plurality of pre-given calibration distances between the digital camera of the image processing system and the calibration device, and the at least one calibration parameter for each pre-given calibration distance is stored in the digital camera and / or the analysis and evaluation unit, wherein the object image captured by the digital camera is corrected with the stored at least one calibration parameter, and the calibration distance of the calibration parameter is closest to the distance between the digital camera and the object when photographing the object. Thus, for example, when using the image processing system on a conveyor belt, a suitable at least one calibration parameter can be selected according to the size of the object (thereby changing the distance between the object and the digital camera) and the object image captured by the digital camera can be corrected. Alternatively, interpolation can also be performed on the stored calibration parameters to determine a suitable at least one calibration parameter for the distance between the digital camera and the object during object photography. The flexibility of the digital camera can be improved by the stored calibration parameters, because suitable calibration parameters can be selected or determined for multiple different distances between the digital camera and the object. Description of the Drawings

[0028] In the following, the present invention will be explained in more detail with reference to Figures 1 to 3 which show, by way of example, schematically and non-limitingly, preferred configurations of the present invention. Shown herein are:

[0029] Figure 1 : The basic structure of the calibration device according to the present invention and the arrangement with the digital camera of the image processing system,

[0030] Figure 2a: Arrangement of the calibration device according to the invention, digital cameras with an image processing system at different calibration distances from the calibration device,

[0031] Figure 2b : Arrangement of a digital camera of an image processing system for image processing applications, and

[0032] Figure 3 : Exemplary flow of a method of using a calibration device according to the invention. Detailed Description

[0033] In Figure 1 the basic structure of the calibration device 1 is shown. Here, for example, a calibration plate is shown as the calibration device 1, since this calibration plate constitutes a common embodiment in the prior art. Of course, the shape of the calibration device 1 is not limited to this embodiment. The calibration device 1 is preferably made of a light but firm material with as low a coefficient of thermal expansion as possible, such as glass, ceramic or similar materials, so that the calibration device 1 is easy to operate during use and maintains its shape (e.g., does not distort or otherwise deform) during temperature fluctuations.

[0034] On the calibration device 1, calibration features 4, usually a plurality of calibration features 4, are arranged, which are arranged distributively on at least a part of the surface of the calibration device 1. The calibration features 4 are basically recognizable features with a defined geometry on the calibration device 1. The arrangement of the calibration features 4 on the calibration device 1 is known, for example, due to the manufacture of the calibration device 1. The arrangement of the calibration features 4 on the calibration device 1 is described herein by arrangement characteristics. The arrangement characteristics describe, for example, the pattern type (e.g., rectangular), the number of calibration features 4 (e.g., per column and per row), the distance between the calibration features 4, the numbering of the calibration features 4, etc. of the arrangement of the calibration features 4 on the calibration device 1.

[0035] In Figure 1 for example, a plurality of circles of the same size are provided on the calibration device 1, spaced apart from each other at equal distances (e.g., measured between the centers) and distributed on the calibration device 1, as the calibration features 4. Here, the calibration features 4 are arranged distributively in columns and rows on the calibration device 1, where of course the calibration features 4 can also be arranged, for example, circularly, annularly, spirally or hexagonally or in any other way on the calibration device 1. The important thing is only that the arrangement of the calibration features 4 is known or pre-given, and this arrangement is described by pre-given arrangement characteristics.

[0036] In Figure 1For example, a Cartesian coordinate system having x, y, and z directions is shown. The origin of the coordinate system may be located at a reference point on the calibration device 1 (e.g., at a corner or in the middle of the calibration device 1). For example, the position of the calibration feature of the calibration element 4 on the calibration device 1 (in the x, y, and z directions of the coordinate system) can be derived from the reference point (e.g., to the center of the calibration feature). The arrangement of the calibration features 4 on the calibration device 1 can be determined from a reference position on the calibration device 1 in terms of a pattern type, the number of calibration features 4, and a pre-given distance between the calibration features 4, without having to specify the position for each calibration feature. In another embodiment, the calibration features 4 are arranged on concentric circles, where a determined number of calibration features 4 are arranged on each circle. Thus, it is sufficient to describe the center of the concentric circles, the diameter of the concentric circles, and the number of calibration features 4 on each circle if the pre-given arrangement characteristics are described.

[0037] Therefore, any arrangement of the calibration features 4 on the calibration device 1 can be described with very little information in the form of arrangement characteristics.

[0038] Instead of being circular, the calibration features 4 on the calibration device 1 can also be configured in other geometric shapes, such as squares, rectangles, rings, etc.

[0039] The calibration features 4 can be machined into the calibration device 1 (e.g., drilled, stamped, or engraved) or printed, glued, etc., where the calibration features 4 should not be detached from the calibration device 1. Here, the calibration features 4 can be implemented in different colors, where high contrast, such as Figure 1 dark circles on a bright, as little reflective as possible surface of the calibration device 1, is advantageous in order to be able to clearly identify the calibration features 4 in the images taken by the digital camera 3 of the image processing system 2.

[0040] The target position of the calibration features 4 on the calibration device 1 describes the position where the calibration features 4 should be present on the calibration device 1. The target position is derived from the construction of the calibration device 1 used. Compared with the target position of the calibration features 4, the actual position of the calibration features 4 describes the actual position of the calibration features 4 that actually exist on the calibration device 1. The actual position of the calibration features 4 is related to the manufacturing accuracy, manufacturing defects, etc. of the calibration features 4 and / or the calibration device 1 and is determined, for example, by measuring the calibration features 4 and / or the calibration device 1 during or after the manufacture of the calibration device 1.

[0041] At least one information memory 5 is arranged on the calibration device 1, and at least one information memory 5 is inseparably connected to the calibration device 1. Inseparable here means that at least one information memory 5 remains connected to the calibration device 1 during the service life of the calibration device 1 and cannot be detached from the calibration device 1 without damage when using the calibration device 1 according to the present invention. Thus, it can be ensured that the data stored in the information memory 5 is correctly assigned to a specific calibration device 1. Incorrect calibrations (e.g., due to human errors during application) and thus incorrect results in image processing are avoided. At least one information memory 5 can be fixed to the calibration device 1, for example, by gluing, printing, pressing (e.g., stamping, engraving, etc.) or in other suitable ways. Here, as Figure 1 shown, the information memory can be arranged in the middle of the calibration device 1. Depending on the application, at least one information memory 5 can also be arranged in the corner, on the side or on the back side of the calibration device 1.

[0042] Preferably, at least one information memory 5 can be optically read from the digital camera 3 of the image processing system 2. As a preferred embodiment of at least one information memory 5, a two-dimensional code (especially a QR code or a data matrix code as Figure 1 shown) is provided on the calibration device 1. However, in principle, all conceivable embodiments of the information memory 5 can be used, which is constructed to store data for calibrating the digital camera 3 of the image processing system 2 and to be preferably quickly and reliably read from the digital camera 3 of the image processing system 2 for calibration. For example, other optically readable characters, such as one-dimensional codes (e.g., barcodes), and non-optically readable information memories 5, such as RFID transponders, can also be used.

[0043] At least one information memory 5 can also be divided into a plurality of information storage units. For example, the optically readable information memory 5 can contain a plurality of two-dimensional codes as information storage units. Depending on the calibration device 1 and its spatial extent, a plurality of information memories 5 can also be provided on the calibration device, and each information memory 5 can in turn contain a plurality of information storage units. Thus, it can be ensured that when using the calibration device for calibration, the digital camera 3 can read at least one information memory 5. Here, the information memories 5 can be arranged on the calibration device 1 in different embodiments (e.g., types, sizes, etc.) and at different positions. Thus, the data for calibration can also be assigned to different information memory units of the information memory 5 or to different information memories 5. However, in a preferred configuration, only a single unique information memory 5 is arranged on the calibration device 1.

[0044] At least one information memory 5 contains the reference target positions of the reference calibration features of the calibration features 4 on the calibration device 1. Thus, the reference calibration features are among the calibration features 4 arranged on the calibration device 1. Preferably, the reference target positions from the only reference calibration feature among the calibration features 4 are sufficient. However, it can be advantageous, for example, if only a part of the calibration device 1 is photographed by means of the digital camera 3 during calibration, to use a plurality of the calibration features 4 as reference calibration features. However, the number of reference calibration features is in any case less than the number of calibration features 4 arranged on the calibration device 1, in particular less than the number of calibration features 4 arranged on the calibration device 1 that can be detected by the digital camera 3 in an image. For example, the reference target positions of the reference calibration features can be derived from a reference point (e.g., a corner or the middle of the calibration device 1) on the calibration device 1 in the x, y, and z directions of the coordinate system as target positions on the calibration device 1.

[0045] Furthermore, at least one information memory 5 contains pre-given arrangement characteristics that describe the arrangement of the calibration features 4 on the calibration device 1. During use of the calibration device 1, the pre-given arrangement characteristics and the reference target positions of the reference calibration features can be read from the at least one information memory 5 by the digital camera 3 of the image processing system 2. During use of the calibration device 1, the analysis and evaluation unit 6 of the image processing system 2 is configured to determine the target positions of at least one (different from the reference calibration feature) further calibration feature of the calibration features 4 on the calibration device 1 from the pre-given arrangement characteristics and the reference target positions of the reference calibration features. For this purpose, a suitable algorithm can be stored in the analysis and evaluation unit 6 of the image processing system 2, which algorithm is configured to determine the target positions of at least one further calibration feature based on the arrangement characteristics and the reference target positions of the reference calibration features.

[0046] Preferably, the type of the pre-given pattern, the number and size of the calibration features 4, and the distances between the calibration features 4 are used as arrangement characteristics. These arrangement characteristics must be pre-given in any case so that the analysis and evaluation unit 6 of the image processing system 2 can determine the target positions of at least one further calibration feature. The fewer the arrangement characteristics are pre-given, the lower the storage requirement of the at least one information memory 5.

[0047] The target positions of all other calibration features 4 can be determined from the reference target positions of the only reference calibration feature and the arrangement characteristics. Thus, the target positions of the calibration features 4 can be stored in the at least one information memory 5 very space-savingly.

[0048] At least one information memory 5 also contains correction parameters for each calibration feature 4 available for calibration on the calibration device 1 and thus also contains correction parameters for the reference calibration feature if the reference calibration feature is set to be used during calibration. The correction parameters describe the relative deviation between the known actual position and the target position of the reference calibration feature and / or the relative deviation between the known actual position and the target position of at least one further calibration feature (determined by the reference calibration feature and the arrangement characteristics). For example, during or after the manufacture of the calibration device 1, the known actual position is determined, for example, by measuring the calibration feature 4 and / or the calibration device 1. Suitable and known measuring methods can be used for this purpose, and thus this need not be discussed further here.

[0049] When the calibration device 1 is in use, the correction parameters can be read from at least one information memory 5 by the digital camera 3 of the image processing system 2.

[0050] The correction parameters preferably additionally describe the relative deviation of the dimensions and / or shape and / or orientation of the reference calibration feature and / or at least one further calibration feature. Here, for example, compared to the position, the correction parameters can describe the relative deviation between the known actual dimensions and the target dimensions of the reference calibration feature (for example included in the pre-given arrangement characteristics) and / or at least one further calibration feature.

[0051] By including the correction parameters as relative deviations in at least one information memory 5 and the target position of at least one further calibration feature being determined by the analysis and evaluation unit 6 of the image processing system 2, the storage requirements of the information memory 5 on the calibration device 1 can be significantly reduced. For example, compared to storing the correction parameters as relative deviations (for example -0.01 mm), storing the absolute actual position and the absolute target position of the calibration feature (for example 999.99 mm and 1000.00 mm as coordinates in a coordinate system) requires more storage in at least one information memory 5. Additionally, the relative deviations can be advantageously encoded and expressed, for example, as a power of 10 (for example -10 -5 m, so that for a deviation of 0.01 mm only the value "1" needs to be stored more), thereby reducing the storage requirements of at least one information memory 5.

[0052] For the calibration features available for calibration on the calibration device 1, as a minimum prerequisite, their target positions on the calibration device 1 and the associated correction parameters must be given. The target position (reference target position) of the reference calibration feature is pre-given. The target position of at least one further calibration feature can be determined as described by the reference target position of the reference calibration feature and the arrangement characteristics.

[0053] Preferably, the position of at least one information memory 5 on the calibration device 1 is included in at least one information memory 5 (for example, as inFigure 1 the middle of the middle calibration device 1). This position can be read by the digital camera 3 from at least one information memory 5 and used, for example, as a reference point for calibration and / or for determining the orientation of the calibration device 1 in the image taken by the digital camera 3. The position of the at least one information memory 5 on the calibration device 1 can be given, for example, from the corners or the middle of the calibration device 1 in the x, y, and z directions in the coordinate system of Figure 1 the coordinate system.

[0054] In Figure 1 FIG. shows an exemplary arrangement of the digital camera 3 of the image processing system 2 and the calibration device 1. The digital camera 3 includes, for example, optical components, such as a lens with a focal length and an aperture, and a control unit for controlling image capture. The digital camera 3 can be integrated into the image processing system 2 here or can also be external. For calibrating the digital camera 3 of the image processing system 2, the calibration device 1 is arranged in the capture area of the digital camera 3. The calibration device 1 is spaced apart from the digital camera 3 of the image processing system 2 by a pre-given calibration distance A (for example, in Figure 1 the z direction in FIG.). Here, the calibration device 1 is preferably oriented at a right angle to the main axis of the digital camera 3. The calibration device 1 can also be tilted, for example, positioned at an angle other than a right angle to the main axis. The digital camera 3 of the image processing system 2 is set up to capture at least one image of the calibration device 1 and read at least one information memory 5. The digital camera 3 is positioned such that the calibration device 1 is at least partially located in the capture area of the digital camera 3, so that at least one image of the calibration device 1 can be captured by the digital camera 3 and at least one information memory 5 can be read.

[0055] Furthermore, the image data captured by the digital camera 3 and the data from at least one information memory 5 (pre-given arrangement characteristics, reference target positions of reference calibration features, correction parameters) are processed by means of an analysis and evaluation unit 6, for example, an analysis and evaluation unit provided in the image processing system 2, for further use. The analysis and evaluation unit 6 can also be part of the digital camera 3 or external. The analysis and evaluation unit 6 is preferably hardware based on a microprocessor, such as a microcontroller. The analysis and evaluation unit 6 and / or the digital camera 3 can also perform certain preprocessing of the captured image data and / or may require (also additional) subsequent processing of the captured images, for example, for filtering, processing, and / or correcting the image data.

[0056] A lighting unit can also be provided (in Figure 1(not shown in the figure) For example, it is integrated in the image processing system 2 or externally to illuminate the calibration device 1 for calibrating the digital camera 3. The lighting unit can be controlled, for example, by the control unit of the digital camera 3. The lighting should be carried out as evenly as possible over the entire surface of the calibration device 1 or at least one calibration feature 4. The calibration device 1 can also be implemented as self-luminous. For example, at least one calibration feature 4 itself emits light of a determined wavelength as evenly as possible (for example, as a dot-shaped light-emitting diode (LED)). Here, it is important that for each LED, a homogeneous bright and preferably circular imaging in the image is achieved. For this purpose, the LED can be sunk into a hole in the calibration device 1, with a ground glass sheet arranged above it.

[0057] The digital camera 3 of the image processing system 2 is configured to read (for example, optically) pre-given arrangement characteristics, the reference target positions of the reference calibration features, and correction parameters from at least one information memory 5 of the calibration device 1, and transmit at least one image of the calibration device 1, the pre-given arrangement characteristics, the reference target positions of the reference calibration features, and the correction parameters to the analysis and evaluation unit 6 of the image processing system 2, for example, through a suitable wired or wireless connection. The analysis and evaluation unit 6 is configured to determine the target positions of at least one additional calibration feature from the pre-given arrangement characteristics and the reference target positions of the reference calibration features, and analyze and evaluate at least one image of the calibration device 1 and the reference target positions of the reference calibration features and the target positions and correction parameters of at least one additional calibration feature from at least one information memory 5 using known calibration methods to determine at least one calibration parameter, for example, by means of an initial calculation and then numerically optimizing the defined parameters. For example, a suitable algorithm is known from "A Flexible New Technique for Camera Calibration" by Zhoung Zhang (1998).

[0058] The digital camera 3 and / or the analysis and evaluation unit 6 are set up to store and use at least one determined calibration parameter in order to correct an image (e.g., of an object) taken with the digital camera 3 (e.g., by means of the known pinhole camera model according to Hartley and Zisserman ("Multiple View Geometry in Computer Vision"), Cambridge University Press (2003)), and according to the known distortion model by Duane C Brown ("Decreasing Distortion of Lenses". In: Optical Measurement Engineering 32.3 (1966), pp. 444-462) or other suitable mathematical models). Alternatively, the digital camera 3 and / or the analysis and evaluation unit 6 are set up to determine at least one characteristic of the object from the taken image and to correct this at least one characteristic with at least one stored calibration parameter. Here, at least one characteristic of the object can in particular describe the position, size, dimensions, orientation or shape of the object. Here, at least one characteristic can be determined, for example, in pixels, where the conversion to metric units (e.g., millimeters) can also be understood as correcting this characteristic with at least one calibration parameter.

[0059] At least one calibration parameter can be stored, for example, in a storage unit integrated in the digital camera 3 or the analysis and evaluation unit 6. It is also possible to store at least one calibration parameter in an external storage unit (e.g., in the cloud). Preferably, at least one information memory 5 contains an identification code that can be read by the digital camera 3 from the at least one information memory 5, where the analysis and evaluation unit 6 can, for example, assign at least one calibration parameter to this identification code in order to save at least one calibration parameter, for example, for a pre-given calibration distance A.

[0060] The analysis and evaluation unit 6 is preferably set up to receive at least one camera parameter from a factory calibration and to analyze and evaluate it in order to determine at least one calibration parameter. The factory calibration is carried out, for example, during the manufacture of the digital camera 3. At least one camera parameter from the factory calibration describes a specific optical characteristic of the digital camera 3 (e.g., the optics). Here, at least one camera parameter can also describe the characteristics of the lighting unit used during the factory calibration (e.g., having different wavelengths). With the aid of at least one camera parameter, imaging errors in the taken image, such as geometric distortion or chromatic aberration, can be corrected. The characteristics, causes and effects of these imaging errors are known and are therefore not discussed in detail here. Preferably, at least one camera parameter is the pre-given focal length of the digital camera 3 of the image processing system 2. At least one camera parameter can also be the pre-given focal length and / or the pre-given aperture of the optics of the digital camera 3. At least one camera parameter from the factory calibration can be provided in an external storage unit (e.g., in the cloud) or saved in the digital camera 3 or the analysis and evaluation unit 6.

[0061] In Figure 2aThe figure shows a digital camera 3 with an image processing system 2 of a calibration device 1 according to the present invention at different calibration distances A from the calibration device 1 1 、A 2 of an exemplary arrangement. Here, for example, the same calibration device 1 is arranged further described below in the order for each calibration distance A 1 、A 2 The digital camera 3 of the image processing system 2 is arranged, for example, fixed or movable in space, for example, on the indoor ceiling above the calibration device 1 (not shown in Figure 2a and Figure 2b ) or on another suitable holding device. The digital camera 3 is positioned such that the calibration device 1 is located in the shooting area of the digital camera 3, so that at least one image of the calibration device 1 can be taken by the digital camera 3 and at least one information memory 5 can be read. In Figure 2a , the objects O 1 、O 2 are arranged in different sizes. The objects O 1 、O 2 can be, for example, packages, where these packages can be arranged on a conveyor belt (not shown in Figure 2a and 2b ) and can move past the digital camera 3 at a distance A O1 、A O2 . The calibration device 1 is positioned at right angles to the camera axis of the digital camera 3 (for example, on one of the objects O 1 、E 2 ) in a plane E Figure 2a and 2b schematically shown), where the digital camera 3 of the image processing system 2 is spaced from each plane E 1 、O 2 by a pre-given calibration distance A 1 、E 2 arranged. Each pre-given calibration distance A 1 、A 2 can extend, for example, from the surface of the calibration device 1 (as shown in 1 、A 2 ) to the digital camera 3 (for example, up to the image plane of the image sensor of the digital camera 3, not shown in Figure 2a ) and Figure 2a and 2b ). During calibration, the calibration device 1 can also be held, for example, by a robotic arm or other suitable holding device (not shown in Figure 2a and 2b ) on one of the objects O 1 、O 2 . Of course, additional objects O1 , O 2 The objects may also be arranged in different sizes or in the same size for calibrating the digital camera 3 of the image processing system 2. For example, there may be ten objects with different sizes. 1 , O 2 , and the smallest object O 1 , O 2 , the largest object O 1 , O 2 and its size is between the smallest and largest object O 1 , O 2 Object O 1 , O 2 The calibration device 1 is used for calibration of the digital camera 3. In order to calibrate the digital camera 3 of the image processing system 2, for example, the calibration device 1 is arranged on three objects O, respectively. 1 , O 2 There are therefore three predefined calibration distances A 1 , A 2 .

[0062] The digital camera 3 of the image processing system 2 records at least one image of the calibration device 1 when the calibration device 1 is used for calibrating the digital camera 3 of the image processing system 2. In this case, the calibration device 1 is, for example, located in the plane E. 1 Positioning in object O 1 On the plane E, the digital camera 3 is 1 Separate calibration distance A 1 In this case, the digital camera 3 reads the predetermined arrangement characteristics, the reference target position of the reference calibration feature and the correction parameters from the at least one information memory 5. The at least one image of the calibration device 1, the predetermined arrangement characteristics, the reference target position of the reference calibration feature and the correction parameters are transmitted to the evaluation unit 6 of the image processing system 1. The evaluation unit 6 determines the target position of at least one further calibration feature from the predetermined arrangement characteristics and the reference target position of the reference calibration feature and evaluates the at least one image of the calibration device 1 and the reference target position of the reference calibration feature, the target position of at least one further calibration feature and the correction parameters from the at least one information memory 5 to determine at least one calibration parameter.

[0063] Preferably, the analysis and evaluation unit 6 receives at least one camera parameter from factory calibration and determines at least one calibration parameter for calibrating the digital camera 3 of the image processing system 2 from the at least one camera parameter, the at least one image of the calibration device 1, and the reference target positions of the reference calibration features, the target positions of at least one additional calibration feature, and the correction parameters from at least one information memory 5. The digital camera 3 and / or the analysis and evaluation unit 6 preferably stores and uses the determined at least one calibration parameter in order to correct the object O 1 、O 2 (for example, the object O in this case 1 ) in the image.

[0064] Preferably, for a plurality of predefined calibration distances A 1 、A 2 between the digital camera 3 of the image processing system 2 and the calibration device 1, at least one calibration parameter is determined for each predefined calibration distance A 1 、A 2 respectively, and the at least one calibration parameter for each predefined calibration distance A 1 、A 2 is stored in the digital camera 3 and / or the analysis and evaluation unit 6. Instead of the calibration distance, at least one calibration parameter can also be determined for different focusing settings or aperture settings of the digital camera 3 or for different wavelengths of the illumination unit. The images of the object O 1 、O 2 taken by the digital camera 3 are corrected with the calibration parameter, and the calibration distance A 1 、A 2 of the calibration parameter is the closest to the distance A 1 、O 2 between the digital camera 3 and the object O 1 、O 2 during the shooting of the object O O1 、A O2 。 Instead, interpolation can also be performed on the stored calibration parameters in order to determine a suitable at least one calibration parameter for the distance A 1 、O 2 between the digital camera 3 and the object O 1 、O 2 during the shooting of the object O O1 、A O2 。 This can be done in the same way for different focusing settings, aperture settings, or illumination settings of the digital camera 3 or for other different settings.

[0065] By using different calibration distances A 1 、A 2(or other settings) respectively determine at least one calibration parameter, and the digital camera 3 of the image processing system 2 can quickly and easily match the changing boundary conditions. For example, during an image processing application, objects O 1 、O 2 (such as packages) of different sizes can be conveyed on a conveyor belt, thereby changing the distance A 1 、O 2 between the objects O O1 、A O2 and the digital camera 3. Another object O Figure 2b is exemplarily shown in 3 , the size of which is between the planes E 1 、E 2 , but closer to the plane E 2 . It is not necessary to recalibrate the digital camera 3 for the object O 3 , but instead, for example, measure the distance A 3 between the digital camera 3 and the object O O3 (for example, by means of a laser distance sensor) and correct the image of the object O 3 taken by the digital camera 3 with the calibration parameter, the calibration distance A 1 、A 2 of which is closest to the distance between the digital camera 3 and the object O 3 during the shooting of the object O 3 . In this case, for example, the calibration distance A 2 is closest to the distance between the digital camera 3 and the object O 3 , whereby the image of the object O 3 taken by the digital camera 3 is corrected with at least one calibration parameter already determined with the calibration distance A 2 . At least one calibration parameter for each calibration distance A 1 、A 2 can also be stored as a so-called recipe in the digital camera 3 and / or the analysis and evaluation unit 6. Alternatively, as described above, the calibration parameter for the distance A 3 between the digital camera 3 and the object O O3 can also be determined by interpolation.

[0066] Finally, an exemplary flow of a method for calibrating the digital camera 3 of the image processing system 2 with the calibration device 1 according to the present invention is shown in Figure 3 . Here, the first step S1 is to position the calibration device 1 in the planes E 1 、E 2 (as shown in Figure 2a ). Here, the digital camera 3 of the image processing system 2 and the planes E 1 、E 2Spacially separate a pre-given calibration distance A 1 、A 2 Arrange. In a second step S2, at least one image of the calibration device 1 is taken by means of a digital camera 3. In a third step S3, the digital camera 3 reads from at least one information memory 5 pre-given arrangement characteristics, the reference target positions of reference calibration features, and correction parameters. Of course, even before the digital camera 3 takes at least one image of the calibration device 1, the pre-given arrangement characteristics, the reference target positions of reference calibration features, and the correction parameters have already been read by the digital camera 3 from at least one information memory 5. In a fourth step S4, the at least one image, the pre-given arrangement characteristics, the reference target positions of reference calibration features, and the correction parameters are transmitted to the analysis and evaluation unit 6 of the image processing system 1. In a fifth step S5, the analysis and evaluation unit 6 determines the target positions of at least one additional calibration feature from the pre-given arrangement characteristics and the reference target positions of reference calibration features, and analyzes and evaluates at least one image of the calibration device 1, the reference target positions of reference calibration features from at least one information memory 5, the target positions of at least one additional calibration feature, and the correction parameters to determine at least one calibration parameter. In a sixth step S6, the digital camera 3 and / or the analysis and evaluation unit 6 stores and uses the determined at least one calibration parameter to correct the images taken by means of the digital camera 3. Alternatively, in the sixth step S6, the digital camera 3 and / or the analysis and evaluation unit 6 may determine at least one characteristic of the object O 1 、O 2 、O 3 from the images of the object O 1 、O 2 、O 3 and stores and uses the determined at least one calibration parameter to correct at least one characteristic of the object O 1 、O 2 、O 3 . Preferably, in the fifth step S5, the analysis and evaluation unit 6 receives at least one camera parameter from a factory calibration in order to determine at least one calibration parameter for calibrating the digital camera 3 of the image processing system 2 from the at least one camera parameter together with at least one image of the calibration device 1, the reference target positions of reference calibration features from at least one information memory 5, the target positions of at least one additional calibration feature, and the correction parameters.

Claims

1. A calibration device (1), in particular a calibration plate, for determining at least one calibration parameter for calibrating a digital camera (3) of an image processing system (2), wherein a calibration feature (4) and at least one information storage device (5) are arranged on the calibration device (1), and the at least one information storage device (5) is inseparably connected to the calibration device (1), wherein the at least one information memory (5) contains a predetermined arrangement property which describes the arrangement of the calibration feature (4) on the calibration device (1), It is characterized in that The at least one information memory (5) contains a reference target position of a reference calibration feature of the calibration feature (4) on the calibration device (1), wherein the predetermined arrangement characteristics and the reference target position of the reference calibration feature can be read from the at least one information memory (5) by the digital camera (3) of the image processing system (2) when the calibration device (1) is used, When using the calibration device (1), an evaluation unit (6) of the image processing system (2) can determine a target position of at least one further calibration feature of the calibration feature (4) on the calibration device (1) from the predetermined arrangement characteristic and the reference target position of the reference calibration feature, and The at least one information memory (5) contains correction parameters for the reference calibration feature and / or the at least one further calibration feature, respectively, the correction parameters describing the relative deviation between the known actual position of the corresponding calibration feature and the target position, wherein when using the calibration device (1), the correction parameters can be read from the at least one information memory (5) by the digital camera (3) of the image processing system (2).

2. The calibration device (1) according to claim 1, characterized in that The correction parameters additionally describe relative deviations in size and / or shape and / or orientation of the reference calibration feature and / or of the at least one further calibration feature.

3. The calibration device (1) according to claim 1 or 2, characterized in that The at least one information storage device (5) can be read out optically by means of the digital camera (3).

4. The calibration device (1) according to claim 3, characterized in that The predetermined arrangement properties, the reference target position of the reference calibration feature and the correction parameters are stored in the at least one information memory (5) as a two-dimensional code, preferably a QR code.

5. The calibration device (1) according to any one of claims 1 to 4, characterized in that The at least one information memory (5) additionally stores an identification code of the calibration device (1).

6. The calibration device (1) according to any one of claims 1 to 5, characterized in that The at least one information memory (5) additionally contains the position of the at least one information memory (5) on the calibration device (1).

7. A device for determining at least one calibration parameter for calibrating a digital camera (3) of an image processing system (2) by means of a calibration device (1) according to claims 1 to 6, wherein the calibration device (1) is arranged in a recording area of ​​the digital camera (3) and the digital camera (3) of the image processing system (2) is configured to record at least one image of the calibration device (1), characterized in that The digital camera (3) of the image processing system (2) is configured to read the predetermined arrangement characteristics, the reference target position of the reference calibration feature and the correction parameters from the at least one information storage device (5) of the calibration device (1), and transmit the at least one image of the calibration device (1), the predetermined arrangement characteristics, the reference target position of the reference calibration feature and the correction parameters to the analysis and evaluation unit (6) of the image processing system (2), and the analysis and evaluation unit (6) is constructed to determine the target position of the at least one other calibration feature from the predetermined arrangement characteristics and the reference target position of the reference calibration feature, and to analyze and evaluate the at least one image of the calibration device (1) and the reference target position of the reference calibration feature, the target position of the at least one other calibration feature and the correction parameters from the at least one information storage device (5) to determine the at least one calibration parameter.

8. The device according to claim 7, characterized in that The information storage device (5) is imaged using an image taken by the digital camera (3).

9. The device according to claim 7 or 8, characterized in that The evaluation unit (6) is configured to receive at least one camera parameter from factory calibration and to evaluate the camera parameter to determine the at least one calibration parameter.

10. The device according to claim 9, characterized in that The at least one camera parameter from the factory calibration is a predefined focal length of the digital camera (3) of the image processing system (2).

11. The device according to any one of claims 7 to 10, characterized in that The digital camera (3) and / or the evaluation unit (6) stores and uses the at least one determined calibration parameter in order to correct images recorded with the aid of the digital camera (3).

12. A method for image correction of an image captured by a digital camera (3) of an image processing system (2) or for correcting at least one characteristic of an object (O1, O2, O3) in an image of the object (O1, O2, O3) captured by the digital camera (3) of the image processing system (2), wherein a calibration device (1) according to claims 1 to 6 is positioned in the capturing area of ​​the digital camera (3) and the digital camera (3) of the image processing system (2) captures at least one image of the calibration device (1), characterized in that The digital camera (3) reads the predetermined arrangement characteristics, the reference target position of the reference calibration feature and the correction parameters from the at least one information memory (5) and transmits them to an evaluation unit (6) of the image processing system (1). The analysis and evaluation unit (6) determines the target position of at least one other calibration feature from the predetermined arrangement feature and the reference target position of the reference calibration feature, and analyzes and evaluates the at least one image of the calibration device (1) and the reference target position of the reference calibration feature, the target position of the at least one other calibration feature and the correction parameter from the at least one information memory (5) to determine the at least one calibration parameter, and The digital camera (3) and / or the analysis and evaluation unit (6) stores and uses the at least one calibration parameter determined in order to correct an image captured by means of the digital camera (3), or the digital camera (3) and / or the analysis and evaluation unit (6) determine at least one characteristic of the object (O1, O2, O3) from an image of the object (O1, O2, O3) captured by means of the digital camera (3), and stores and uses the at least one calibration parameter determined in order to correct the at least one characteristic of the object (O1, O2, O3).

13. The method according to claim 12, characterized in that The analysis and evaluation unit (6) receives at least one camera parameter from a factory calibration and determines from the at least one camera parameter together with the at least one image of the calibration device (1) and the reference target position of the reference calibration feature from the at least one information memory (5), the target position of the at least one other calibration feature and the correction parameters, the at least one calibration parameter of the digital camera (3) for calibrating the image processing system (2).

14. The method according to any one of claims 12 or 13, characterized in that The digital camera (3) and / or the evaluation unit (6) stores and uses the at least one determined calibration parameter in order to correct images of objects (O1, O2, O3) recorded with the aid of the digital camera (3).

15. The method according to any one of claims 12 to 14, characterized in that For each predetermined calibration distance (A, A1, A2) of the digital camera (3) of the image processing system (2) relative to the calibration device (1), at least one calibration parameter is determined, and the at least one calibration parameter for each predetermined calibration distance (A, A1, A2) is stored in the digital camera (3) and / or the evaluation unit (6), and An image of an object (O1, O2, O3) captured by means of the digital camera (3) is corrected using the at least one stored calibration parameter, wherein the calibration distance (A, A1, A2) of the calibration parameter is closest to the distance (A) of the digital camera (3) relative to the object (O1, O2, O3) during the capturing of the object (O1, O2, O3). O1 , A O2 , A O3 ) or interpolating the stored calibration parameters to obtain the distance (A) of the digital camera (3) relative to the object (O1, O2, O3) during photographing the object (O1, O2, O3) O1 , A O2 , A O3 ) determine at least one suitable calibration parameter.

Images