Method and apparatus for calibrating a camera using a diffractive optical element

Abstract

A method for calibrating a camera (102) using a diffractive optical element (104), the method comprising: detecting a diffraction pattern projected onto an image sensor (112) of the camera (102) using the diffractive optical element (104), wherein the diffractive optical element (104) and the camera (102) are located at an angular position relative to each other; further detecting another diffraction pattern projected onto the image sensor (112) of the camera (102) using the diffractive optical element (104), wherein the diffractive optical element (104) and the camera (102) are located at an angular position relative to each other other than the angular position; and determining a calibration rule for calibrating the camera (102) using these diffraction patterns.

Description

Technical Field

[0001] The starting point of this invention is the apparatus or method described in the preamble of the independent claim. The subject matter of this invention is also a computer program. Background Technology

[0002] For example, cameras used in vehicles must be calibrated.

[0003] US3912395A describes a geometric camera calibration using diffractive optical elements (DOE). Summary of the Invention

[0004] Against this backdrop, utilizing the proposed solution, a method for calibrating a camera using diffractive optical elements according to the main claim is proposed, along with an apparatus for using this method, and finally, a corresponding computer program. Advantageous extensions and improvements to the apparatus described in the independent claim are possible through the measures listed in the dependent claims.

[0005] The described scheme enables calibration of cameras with a large field of view (FOV).

[0006] A method for calibrating a camera using diffractive optical elements includes the following steps: The diffraction pattern is detected, which is projected onto the image sensor of a camera using a diffraction optical element, while the diffraction optical element and the camera are located at an angular position relative to each other; Additionally, another diffraction pattern is detected, which is projected onto the image sensor of the camera using the diffraction optics, while the diffraction optics and the camera are located at a different angular position relative to each other than the stated angular position; and Using these diffraction patterns, a calibration rule is determined for calibrating the camera.

[0007] The camera may relate to, for example, an image detection device for detecting the interior space or surrounding environment of a vehicle. Optionally, the camera may be used in other application areas, such as as a surveillance camera on a building. The camera may include an image sensor (also called a camera sensor) and optics. The optical characteristics of cameras that are identical may deviate from each other, for example, due to manufacturing tolerances. Such deviations can be identified by calibrating the camera and compensated for using the calibration rules. Exemplarily, the calibration rules can be used to ensure that the pixel coordinates of the images recorded by the camera are correctly converted to real-world perspectives, and vice versa. To determine the calibration rules, known methods, such as geometric camera calibration, can be employed. The diffractive optical element may have a predetermined microstructure through which a beam of light incident on the diffractive optical element can be diffracted into multiple diffraction orders, thereby producing the diffraction pattern, such as a matrix of points. Due to the optical characteristics of the camera, the diffraction pattern detected by the camera typically deviates from a known ideal reference pattern associated with the microstructure of the diffractive optical element. The calibration rules can be determined by comparing the detected diffraction pattern with a known reference pattern. The different angular positions between the diffractive optical element and the camera can be selected such that the detected diffraction patterns, when combined, cover the entire area of ​​the image sensor. In this way, the calibration rule can cover all pixels of the image sensor. Different angular positions can be achieved, for example, by tilting the optical axis of the camera relative to the plane of the diffractive optical element.

[0008] In the detection step, the diffraction pattern can be projected onto a segment of the image sensor. In a further detection step, an additional diffraction pattern can be projected onto another segment of the image sensor, which is at least partially different from the first segment. In this way, the entire image sensor can be gradually covered by multiple detections and changes in angular position during the process. According to an embodiment, two or more diffraction patterns can be detected at their respective angular positions.

[0009] During the motion step, the camera and the diffractive optical element can be moved relative to each other to convert the angular position to the other angular position. For this purpose, the camera can be moved while the diffractive optical element is stationary, or vice versa. Alternatively, the camera and the diffractive optical element can be moved simultaneously. Advantageously, the motion step can be performed automatically.

[0010] During the loading step, the diffractive optical element can be loaded with electromagnetic radiation from the side opposite to the camera. The incident radiation can be diffracted by the diffractive optical element, thereby projecting the diffraction pattern onto the image sensor. Exemplarily, the electromagnetic radiation can be provided as a laser beam.

[0011] The diffraction angle of the diffraction optical element can be smaller than the field of view of the camera. Advantageously, inexpensive diffraction optical elements with high quality and relatively small diffraction angles in terms of field of view can be used. For example, the field of view of the camera can be greater than 180°. Thus, the method can be used, for example, to calibrate parking assistance cameras.

[0012] The additional detection steps can be repeated multiple times to detect multiple additional diffraction patterns projected onto the image sensor of the camera using the diffraction optics, while the diffraction optics and the camera are located at multiple additional angular positions relative to each other, different from the stated angular position. Thus, the diffraction patterns can be projected onto the image sensor until the entire area of ​​the image sensor is covered at least once.

[0013] In the defined steps, the calibration rule can be determined by comparing the diffraction pattern with a reference pattern assigned to the diffraction optical element. Here, information about the current angular position in which the diffraction pattern is detected can be considered to assign the camera's pixel coordinates to the corresponding viewing angle.

[0014] For example, the method can be implemented in software or hardware, or in a hybrid form of software and hardware, such as in a control device.

[0015] The proposed solution also creates an apparatus configured to perform, manipulate, or implement various steps of variations of the method proposed herein within a corresponding device. Similarly, this embodiment of the invention, in the form of an apparatus, allows for the rapid and efficient resolution of the task on which the invention is based.

[0016] Therefore, the device may have at least one computing unit for processing signals or data, at least one storage unit for storing signals or data, at least one interface leading to a sensor or actuator for reading sensor signals from the sensor or outputting data signals or control signals to the actuator, and / or at least one communication interface for reading or outputting data embedded in a communication protocol. The computing unit may exemplarily be a signal processor, a microcontroller, or the like, and the storage unit may be a flash memory or magnetic storage unit. The communication interface may be configured to wirelessly and / or wiredly read or output data, wherein a communication interface capable of reading or outputting wired data may exemplarily read or output such data electrically or optically from a corresponding data transmission line.

[0017] In this invention, the device can be understood as an electrical device that processes sensor signals and outputs control signals and / or data signals accordingly. The device may have an interface, which may be constructed in hardware and / or software. In the case of a hardware construction, the interface may, for example, be part of a so-called system ASIC that includes various functions of the device. However, it is also possible that the interface is a dedicated integrated circuit, or at least partially composed of discrete components. In the case of a software construction, the interface may be a software module, which, exemplarily, exists on a microcontroller along with other software modules.

[0018] Also advantageous are computer program products or computer programs having program code that can be stored on a machine-readable carrier or storage medium (such as semiconductor memory, hard disk memory or optical memory) and used to perform, implement and / or manipulate the steps of a method according to one of the embodiments described above, especially when the program product or program is running on a computer or device. Attached Figure Description

[0019] Embodiments of the proposed solution are shown in the accompanying drawings and explained in more detail in the following description. The drawings show: Figure 1 : A schematic diagram of an embodiment of a device for calibrating a camera; Figure 2 Simulation of an embodiment of diffraction patterns detected by a camera; Figure 3 Simulation of an embodiment of a diffraction pattern of a diffraction optical element with a large diffraction angle; Figure 4 The output of an example of the asin function; Figure 5 Illustration of an embodiment of a device for calibrating a camera; Figure 6An example of a diffraction pattern centered within an image; Figure 7 Example of a diffraction pattern offset within an image; Figure 8 Examples of additional diffraction patterns offset within an image; Figure 9 Examples of additional diffraction patterns offset within an image; Figure 10 Examples of additional diffraction patterns offset within an image; Figure 11 Examples of additional diffraction patterns offset within an image; Figure 12 Examples of additional diffraction patterns offset within an image; Figure 13 Examples of additional diffraction patterns offset within an image; Figure 14 Examples illustrating the difference between beam adjustment calibration in front of a diffractive optical element and beam adjustment calibration in front of a physical 3D target; and Figure 15 : A flowchart of a method according to one embodiment. Detailed Implementation

[0020] In the following description of advantageous embodiments of the invention, the same or similar reference numerals are used for elements shown in different figures and which have similar functions, wherein repeated descriptions of these elements are omitted.

[0021] Figure 1 A schematic diagram of an embodiment of a device 100 for calibrating a camera 102 is shown. The device 100 includes a diffractive optical element 104 disposed in front of the camera 102. The diffractive optical element 104 is configured to generate a projection 106 of a diffraction pattern when loaded with electromagnetic radiation 108. The projection 106 here consists of a plurality of diffracted beams. Exemplarily, the electromagnetic radiation 108 is shown as a laser beam emitted by a laser 110. The laser 110 or other suitable radiation source may optionally be part of the device 100. The electromagnetic radiation 108 is incident on a surface of the diffractive optical element 104 facing away from the camera 102. The projected diffraction pattern, when using the diffractive optical element 104, is detected by an image sensor 112 of the camera 102.

[0022] For example, in Figure 1In the state shown, the image sensor 112 is oriented parallel to the diffractive optical element 104. The propagation direction of the electromagnetic radiation 108 here exemplary follows the optical axis of the camera 102. In other states, the image sensor 112 and the diffractive optical element 104 are oriented in different ways relative to each other.

[0023] According to this embodiment, projection 106 covers only a portion of the area of ​​image sensor 112. To cover the entire area of ​​image sensor 112, diffractive optical element 104 and camera 102 are moved relative to each other. Exemplarily, for this purpose, camera 102 is tilted 114 while diffractive optical element 104 remains stationary.

[0024] Therefore, it is possible to continuously detect diffraction patterns projected onto different areas of the image sensor 112 when using the diffractive optical element 104, and provide them as images. Here, the number of diffraction patterns detected can be arbitrarily selected, wherein, according to one embodiment, relative motion between the camera 102 and the diffractive optical element 104 is performed between two detection processes respectively.

[0025] According to one embodiment, device 100 includes determining means 116 configured to determine calibration rules for calibrating camera 102 using a detected diffraction pattern.

[0026] According to one embodiment, to expand the covered field of view, the camera 102 is rotated in front of the diffractive optical element 104. This allows for fully automatic calibration of the camera 102 using the diffractive optical element 104. To this end, the markers detected in the image captured by the camera 102 (in this case, markers of the diffraction pattern) are automatically compared with the corresponding diffraction orders of the diffractive optical element 104. Exemplarily, the corresponding steps are performed using the determining device 116.

[0027] Camera 102 can be used, for example, in a vehicle for driver assistance systems or autonomous driving systems, and provides a variety of functions, including identifying objects in a scene using various object recognition algorithms. For many of these algorithms, the world-to-camera or camera-to-world transformation must be known, which means the optical parameters of camera 102 should be known. These parameters include focal length, principal point, and parameters of optical distortion, such as radial and non-radial distortion effects.

[0028] The intrinsic parameter calibration and optical distortion of camera 102 can be measured during the production stage of camera 102. A space-saving and low-cost method for camera calibration is based on the use of diffractive optical element (DOE) 104. Figure 1An embodiment of the corresponding basic structure consisting of a laser 110, a diffractive optical element 104, and a camera 102 for geometric camera calibration of camera 102 is shown.

[0029] If a monochromatic plane wave (such as a laser beam) illuminates the diffractive optical element 104, the incident beam is diffracted into multiple diffraction orders by the precisely fabricated microstructure of the diffractive optical element 104. The Fraunhofer diffraction pattern is visible on a far-field screen. By using powerful optical field simulation tools, it is possible to calculate the desired microstructure to generate arbitrary diffraction patterns.

[0030] For camera calibration, diffractive optical element 104 is placed directly in front of camera 102. The camera optics then act as Fourier lenses. This results in a far-field diffraction pattern being projected onto the image plane. (As shown...) Figure 2 As shown, the plane wave at each diffraction angle is projected as a single point onto the image plane. Here, the pattern on the image sensor 112 is similar to a pattern arranged on a screen at infinity. Of course, the diffraction pattern on the image plane, i.e., the image sensor 112 (also called the camera sensor), will also change due to lens distortion. This fact is used for camera calibration.

[0031] Figure 2 A simulation of an embodiment of a diffraction pattern 200 detected by a camera is shown. Exemplarily, the diffraction pattern 200 relates to the use of a reference... Figure 1 The pattern that can be projected in the case of the described diffractive optical element.

[0032] Known methods exist for performing geometric camera calibration using diffractive optical elements. Currently, DOE suppliers can achieve diffraction angles up to + / - 70°. This allows calibration of cameras with a full field of view (FFOV) of up to 140°. For cameras with an FFOV exceeding 140°, such as parking assistance cameras covering over 200° FFOV or larger, calibration has so far required classical methods using physical 3D targets. Even if larger diffraction angles become achievable in the future, the physical limit of DOE remains 180° FFOV, which is still insufficient for the aforementioned parking assistance cameras.

[0033] Furthermore, the design and manufacturing costs of DOEs with large diffraction angles are significantly higher compared to DOEs with relatively small diffraction angles.

[0034] Another aspect is the distribution of the diffraction angle. The diffraction angle α (for the perpendicular incident angle) follows this rule: Where m is the diffraction order, g is the grating constant of the DOE, and λ is the laser wavelength. Therefore, the diffraction angles are not evenly distributed in angular space.

[0035] For example, it can be Figure 2 The diffraction pattern 200 shown is used as a reference pattern for the diffraction optical elements used to calibrate the camera.

[0036] Figure 3 A simulation of an embodiment is shown, in which a diffraction pattern 300 of a diffraction optical element with a large diffraction angle is captured by a camera with a large field of view.

[0037] For a DOE designed specifically for a wavelength of 632.8 nm, if the line count is 56.4 1 / mm and it has 28 x 28 diffraction orders (maximum diffraction angle = 89.5°), the diffraction pattern 300 on the parking camera will look like this. Figure 3 As shown.

[0038] Figure 4 The output of an example of the asin function 450 with parameters in the range [0,1] is shown.

[0039] based on Figure 4 The asin function shown increases the angular distance between diffraction orders when the diffraction order is large, which leads to very sparse coverage at larger angles.

[0040] This means that even if the diffraction angle may be greater than 70° in the future, it will not be enough to calibrate a camera with a large field of view using a DOE that only captures a single image.

[0041] Currently, DOEs with diffraction angles up to + / - 60° or + / - 70° are used for camera calibration. Therefore, camera calibration is currently limited to cameras with a maximum FFOV of 120°, and at most 140°.

[0042] The scheme described herein creates a technology that extends the potential of DOE-based camera calibration to cameras with very large FOVs.

[0043] Figure 5 An illustration shows an embodiment of a device 100 for calibrating a camera 102 using a diffractive optical element 104. This device 100 exemplarily corresponds to a reference. Figure 1 The device is shown schematically.

[0044] To enable relative movement between camera 102 and diffractive optical element 104, camera 102 is exemplarily held by alignment device 560. Alignment device 560 enables movement of camera 102 relative to diffractive optical element 104 to convert the current angular position between camera 102 and diffractive optical element 104 into another angular position. According to one embodiment, at each angular position, image is captured using camera 102. For example, a control device is used to manipulate alignment device 560 and camera 102, allowing camera 102 to move between consecutive captures.

[0045] The camera 102, which is automatically rotated in front of the diffractive optical element 104, is capable of performing beam adjustment calibration.

[0046] The idea behind this technique is to combine a bundle adjustment scheme with a diffractive optical element 104 to calibrate a camera 102 with a large field of view (FOV). Multiple images are captured for this purpose, with the orientation of the camera 102 relative to the diffractive optical element 104 changed between image captures. Importantly, the lens of the device under test (here, camera 102) is always centered behind the diffractive optical element 104 at a very small distance to achieve optimal image quality.

[0047] By rotating the camera 102 or optionally rotating the diffractive optics 104 (e.g., a structure consisting of a laser beam expander and a DOE), different portions of the sensor / FOV of the camera 102 can be covered with each captured image. The number of images required to cover the entire field of view may vary depending on the field of view of the camera 102 and the diffraction angle of the diffractive optics 104.

[0048] For example, a diffractive optical element 104 with a diffraction angle of + / - 15° is used to calibrate a camera 102 with a horizontal FOV (hFOV) of 90°.

[0049] In mass production, it is necessary to automatically position the camera 102 in different orientation directions. This can be achieved by an orientation device 560 (e.g., in the form of a high-precision robotic arm).

[0050] In the following Figures 6 to 13 Nine images are shown in the figure, which are used for bundle adjustment calibration according to one embodiment.

[0051] Figure 6 An embodiment of a diffraction pattern 600 centered within image 670 is shown. Image 670 is exemplarily a diffraction pattern 600 centered within image 670. Figure 5 The footage was taken by the camera shown.

[0052] In image 670, the diffraction pattern 600 is centered within image 670. It can be clearly seen that the observable diffraction order does not cover the entire field of view of the camera.

[0053] Beyond the diffraction orders used, higher-order diffraction orders with lower intensities can be observed. Due to the low intensity of these diffraction orders, robust detection is not possible for a sufficient number of orders.

[0054] Figure 7 An embodiment of a diffraction pattern 700 arranged offset within an image 760 is shown.

[0055] Figure 8 An embodiment of an additional diffraction pattern 800 arranged offset within image 860 is shown.

[0056] Figure 9 An embodiment of an additional diffraction pattern 900 arranged offset within image 960 is shown.

[0057] Figure 10 An embodiment of an additional diffraction pattern 1000 arranged offset within image 1060 is shown.

[0058] Figure 11 An embodiment of an additional diffraction pattern 1100 arranged offset within image 1160 is shown.

[0059] Figure 12 An embodiment of an additional diffraction pattern 1200 arranged offset within image 1260 is shown.

[0060] Figure 13 An embodiment of an additional diffraction pattern 1300 arranged offset within image 1360 is shown.

[0061] With the help of Figures 6 to 13 The diffraction patterns 600, 700, 800, 900, 1000, 1100, 1200, and 1300 shown do not each cover the entire field of view of the camera and therefore do not cover their respective entire images 670, 770, 870, 970, 1070, 1170, 1270, and 1370. However, by varying the angular positions between the camera and the diffractive optical elements (at which images 670, 770, 870, 970, 1070, 1170, 1270, and 1370 are captured), each region of the camera's entire field of view is covered at least once by one of the diffraction patterns 600, 700, 800, 900, 1000, 1100, 1200, and 1300.

[0062] According to one embodiment, based on Figures 6 to 13The nine images shown (670, 770, 870, 970, 1070, 1170, 1270, 1370) show the camera calibrated to its maximum field of view (FFOV) of 93°. This is three times the angle allowed by the DOE diffraction angle when only one of these images (670, 770, 870, 970, 1070, 1170, 1270, 1370) is captured. By capturing even more images, a camera with a larger FOV can be calibrated using the same DOE. Since rotating the camera or DOE and capturing images is very time-consuming, it is certainly meaningful to find the optimal trade-off between increasing the DOE diffraction angle and the number of locations required to cover the entire FOV.

[0063] Figure 14 This illustrates the use of diffractive optical elements in front of ( Figures 6 to 13 Example 1480 comparing the difference between bundle adjustment calibration (as shown in the image) and calibration in front of a physical 3D target.

[0064] The results of the aforementioned calibration were therefore compared with those of calibration using a bundle adjustment camera with a physical 3D target. Figure 14 The differences in the 3D->2D projection of the two obtained intrinsic parameter sets are shown.

[0065] Here, virtual 3D coordinates are projected onto a virtual 2D sensor, using both intrinsic parameter sets. The figure illustrates the difference in 2D coordinates.

[0066] A difference of approximately 1.5 pixels corresponds to an angular error of 0.055°, which only occurs at the outermost edges of the image. This demonstrates excellent consistency between the two calibration methods.

[0067] Figure 15 A flowchart of a method for calibrating a camera using diffractive optical elements according to one embodiment is shown, as has been exemplarily described with reference to the preceding figures.

[0068] In step 1501, when using the image sensor of a camera, the diffraction pattern projected onto the image sensor of the camera when using the diffractive optical element is detected, while the diffractive optical element and the camera are in an angular position relative to each other.

[0069] In step 1503, when using the image sensor of the camera, an additional diffraction pattern is detected. This additional diffraction pattern is projected onto the image sensor of the camera using a diffraction optical element, while the diffraction optical element and the camera are located at angular positions relative to each other, different from the angular positions in step 1501. Optionally, step 1503 is repeated to further detect multiple diffraction patterns using the image sensor of the camera. These diffraction patterns are projected onto the image sensor of the camera using a diffraction optical element, while the diffraction optical element and the camera are continuously located at new angular positions relative to each other, each of which is different from the previously adopted angular position.

[0070] In step 1505, using the detected multiple diffraction patterns, a calibration rule for calibrating the camera is determined. For example, this is done by comparing the diffraction patterns with a reference pattern assigned to the diffractive optical element.

[0071] Optionally, between steps 1501, 1503, and each subsequent step 1503, if necessary, step 1507 may be performed, in which the camera and the diffractive optical element are moved relative to each other to convert their current angular positions into new angular positions.

[0072] Optionally, the method includes step 1509, in which the diffractive optical element is loaded with electromagnetic radiation.

[0073] This method allows for the calibration of any camera using a DOE that covers only a small portion of the camera's field of view (FOV). By rotating the camera or DOE settings, the covered FOV can even be extended to over 180°, exceeding the physical limits of single-shot DOE calibration.

[0074] Advantageously, the above operations make it possible to calibrate a large FOV camera using a DOE, even if the DOE's diffraction angle does not cover the entire FOV of the camera. This is achieved by aligning the camera with multiple different orientations relative to the DOE and capturing multiple images for bundle adjustment calibration.

[0075] Here, the concept of bundle adjustment calibration is to capture images of the 3D scene from multiple perspectives. Different camera and 3D scene arrangements, along with a large number of images, significantly increase the number of observations. For each image, an independent position and orientation in space are determined. However, all images collectively determine only one set of intrinsic calibration values ​​that allow description of the camera's 3D-to-2D projection. Therefore, this set of calibration values ​​must be valid for all images simultaneously. This achieves a high degree of decoupling between the intrinsic calibration values ​​and the camera's position and orientation in space, meaning higher confidence in the determined parameters. Simultaneously, as mentioned earlier, if the 3D scene itself does not fill the entire camera's field of view, the covered field of view can be expanded.

[0076] This technology allows for the calibration of cameras with a field of view (FOV) greater than 180° using the DOE (Design of Equipment).

[0077] Even if a DOE with a large enough diffraction angle to cover the camera's field of view is available, it may be more economical to use a DOE with a smaller FOV and rotate the camera.

[0078] Since some important quality parameters of a DOE decrease as the diffraction angle increases, such as the uniform intensity distribution on the diffraction order, it may be advantageous to use a high-quality DOE with a smaller diffraction angle and cover the entire FOV by rotating the camera or DOE for extremely high-quality camera calibration.

[0079] Compared to currently used physical 3D targets, this method can significantly reduce the cost and size of test stands used for calibrating large FOV cameras.

[0080] Compared to currently used physical 3D testing platforms, all object information originates from a virtual, infinitely distant location. This enables more precise camera calibration for targets at distances greater than 2 meters.

[0081] As mentioned earlier, the DOE device can also be rotated instead of using a camera. If this option is preferred, it must be ensured that the angle of incidence of the laser onto the DOE does not change or is persistently redefined. This means that the laser itself, or (e.g., if using fiber optics) the coupling and beam-shaping optics, should rotate with the DOE. Regardless of which option is used, it must be ensured that the lens never comes into contact with the surface of the DOE, as this would damage the DOE's microstructure.

[0082] Generally speaking, the described scheme is well used for calibrating large FOV cameras, especially those with an FFOV > 180°.

Claims

1. A method for calibrating a camera (102) using a diffractive optical element (104), wherein, The method includes the following steps: Detect (1501) a diffraction pattern (600) which is projected onto the image sensor (112) of the camera (102) using a diffraction optical element (104), while the diffraction optical element (104) and the camera (102) are located at an angular position relative to each other; Additionally, (1503) another diffraction pattern (700) is detected, which is projected onto the image sensor (112) of the camera (102) using the diffraction optical element (104), while the diffraction optical element (104) and the camera (102) are located at an angular position relative to each other that is different from the stated angular position; and Using these diffraction patterns (600; 700), a calibration rule (1505) for calibrating the camera (102) is determined.

2. The method according to claim 1, wherein, In the detection step (1501), the diffraction pattern (600) is projected onto a segment of the image sensor (112), and in the additional detection step (1503), the additional diffraction pattern (700) is projected onto another segment of the image sensor (112), the other segment being at least partially different from the first segment.

3. The method according to any one of the preceding claims, comprising the step (1507) of moving the camera (102) and the diffractive optical element (104) relative to each other to convert the angular position into the other angular position, wherein, The steps of the movement (1507) are executed automatically.

4. The method according to any one of the preceding claims, comprising the step (1509) of loading the diffractive optical element (104) with electromagnetic radiation (108) from the side opposite to the camera (102).

5. The method according to any one of the preceding claims, wherein, The diffraction angle of the diffraction optical element (104) is smaller than the field of view of the camera (102).

6. The method according to any one of the preceding claims, wherein, The field of view of the camera (102) is greater than 180°.

7. The method according to any one of the preceding claims, wherein, The additional detection step (1503) is repeated multiple times to detect multiple additional diffraction patterns (800; 900; 1000; 1100; 1200; 1300), which are projected onto the image sensor (112) of the camera (102) using the diffraction optical element (104), while the diffraction optical element (104) and the camera (102) are located at multiple additional angular positions relative to each other, different from the stated angular position.

8. The method according to any one of the preceding claims, wherein, In the determined step (1505), the calibration rule is determined by comparing the diffraction pattern (600; 700) with a reference pattern assigned to the diffractive optical element (104).

9. A device (100) for calibrating a camera (102), wherein, The device (100) is configured to perform and / or manipulate the steps of the method according to any one of the preceding claims in a respective unit.

10. A computer program configured to perform and / or manipulate the steps of the method according to any one of claims 1 to 8.

Images