Method and camera for correcting a geometric imaging aberration in an image recording

Abstract

For an accurate correction of influences on the imaging accuracy of a camera that occur during the operation of the camera, provision is made for a deformation (V) of the transparent cover (5) and / or the sensor layer (4) which is caused by a pressure difference (D) between an interior pressure (pi) in the sensor recess (3) and an ambient pressure (P2) of the camera (10) to be modelled by a deformation model (20), for a pressure difference (D) currently prevalent at the time of recording an image (B') with the camera (10) to be determined on the basis of the interior pressure (p1) and ambient pressure (p2) of the camera (10), the pressures being known at the time of recording the image (B'), and the determined pressure difference (D) to be used with the deformation model (20) to determine a deformation (V) of the transparent cover (5) and / or the sensor layer (4) at the time of recording the image (B'), for a displacement of the imaging of the object point (G) onto an arising actual picture point (P") in the image recording (B') caused by the deformation (V) to be determined as geometric imaging aberration (A) from the determined deformation (V) and the known geometric dimensions of the image sensor (1), and for the determined geometric imaging aberration (A) to be corrected in the image recording (B') for the purpose of determining a corrected image recording (B).

Description

[0001] Method and camera for correcting a geometric aberration in an image recording

[0002] The present invention relates to a method for correcting a geometric aberration of a camera with an image sensor formed by a sensor housing, a sensor layer arranged in a sensor recess of the sensor housing and having a plurality of pixels, and a transparent cover hermetically sealing the sensor housing to the outside. A point of an object captured by the camera is imaged onto a pixel of the sensor layer. The invention also relates to a correspondingly designed camera.

[0003] Digital cameras use image sensors, such as CCD (Charge Coupled Device) sensors or Active Pixel Sensors (often also called CMOS sensors), to capture a digital image. The image sensor is located in the image plane of the digital camera and converts the incoming light rays (photons) into electrical signals that are captured and represent the digital image. An image sensor typically consists of a suitable sensor housing, usually made of a ceramic material, in which an electro-optical sensor consisting of a thin electro-optical sensor layer is arranged, and a transparent cover that closes the sensor housing. The sensor layer is usually glued into the housing and connected to the housing.To enable sharp image capture, it is essential that the image sensor is positioned and aligned in a defined manner relative to the digital camera's optical system to enable a precise image of the object captured with the digital camera in the image plane. Digital camera applications in large-format photography, geomatics, or photogrammetry, in particular, place high demands on this image accuracy.

[0004] Therefore, it is already known to position the image sensor with high precision in a digital camera. For example, WO 2021 / 069382 A1 describes a sensor mount that is adapted to an image sensor to enable highly precise positioning of the image sensor in a digital camera using the sensor mount. EP 1 475 960 A2 describes the alignment of an image sensor cover to the image sensor. The disadvantage of this is that the initial alignment or positioning is fixed and cannot be subsequently changed.

[0005] It is therefore already known to arrange an image sensor in a digital camera in a movable or adjustable manner. This is demonstrated, for example, by EP 2 485 474 A1. On the one hand, such an adjustment is structurally complex. On the other hand, such a mechanical adjustment hardly allows the highly precise alignments or positioning required for the aforementioned applications to be achieved. Last but not least, with an active adjustment, a deviation from a desired position must also be determined in order to compensate for this deviation through the adjustment, which further increases the complexity.

[0006] DE 100 13 567 A1 pursues a similar approach, although it also results in the same difficulties. This document discloses how to counteract focus drift in order to improve the image sharpness of an aerial camera by shifting the focus as a function of temperature and air pressure. The required focus shift is modeled using a correction table, and piezoelectric translators are controlled by a suitable control signal to shift the focal plane along the optical axis counter to the focus drift.

[0007] Influences on the imaging accuracy that occur during use of the digital camera can either not be compensated for at all with the known approaches, as in WO 2021 / 069382 A1 or EP 1 475 960 A2, or only to a limited extent or with great effort, as in EP 2 485 474 A1 or DE 100 13 567 A1.

[0008] There is therefore a need for a simple, yet accurate correction of influences on the imaging accuracy of a digital camera that occur during operation of the digital camera.

[0009] This problem is solved for the method and the digital camera mentioned at the beginning with the features of the independent claims.

[0010] It was recognized that the image sensor, in particular the transparent cover and / or the sensor layer, is mechanically deformed due to a pressure difference acting at the time of image acquisition between the pressure in the sensor recess of the image sensor and the ambient pressure of the camera. This deformation leads to a shift in the image of an object point recorded with the camera in the image (or on the sensor layer) in the form of a geometric aberration. This shift can be modeled so that the shift can be determined depending on the pressure difference acting at the time of image acquisition and corrected during image acquisition. A deformation model is used for modeling that models the relationship between the pressure difference acting at the time of image acquisition and the deformation.The deformation model can be pre-calibrated for a specific image sensor, for example based on measurements on the image sensor, and can then be used in camera operation.

[0011] In this way, previously uncorrected geometric aberrations in the image acquisition can be corrected. The quality of the image acquisition can thus be improved. The geometric aberration is advantageously determined as the shift of the image of the object point onto an expected pixel with an undeformed transparent cover and / or undeformed sensor layer compared to the actual image of the object point onto the resulting actual pixel due to the deformation. Since these are simple geometric relationships on the image sensor, taking the law of refraction into account, this can be accomplished easily.

[0012] If the geometric aberration also takes into account an offset caused by optical refraction on the undeformed transparent cover, the quality of the image acquisition can be further improved. This allows a previously known geometric aberration to be corrected in the same step.

[0013] The correction of the geometric aberration in the image recording can be carried out in a simple manner by shifting the image of the object point onto the actual image point resulting from the image recording in the corrected image recording to another image point by the geometric aberration.

[0014] The present invention will be explained in more detail below with reference to Figures 1 to 6, which show exemplary, schematic and non-limiting advantageous embodiments of the invention.

[0015] Fig.1 shows an embodiment of an image sensor with sensor housing, sensor layer and transparent cover,

[0016] Fig.2 a geometrical aberration caused by the deformation of the transparent cover,

[0017] Fig.3 a geometrical imaging error caused by the deformation of the sensor layer,

[0018] Fig.4 the use of the image sensor in a camera,

[0019] Fig.5 a scheme for correcting the geometric imaging error and

[0020] Fig.6 an arrangement for determining or calibrating the deformation model.

[0021] A typical embodiment of an image sensor 1 is shown in Fig. 1. A sensor layer 4, which forms an electro-optical sensor, is arranged in a sensor recess 3 of a sensor housing 2. The sensor housing 2 is usually made of a ceramic material. The sensor housing 2 is hermetically sealed from the outside world by a transparent cover 5, usually made of glass. The sensor recess 3 is surrounded by the sensor housing 2 except for the part closed by the transparent cover 5. The sensor layer 4 is normally connected to the sensor housing 2, for example, glued to the bottom of the sensor recess 3, usually opposite the transparent cover 5. Necessary, sufficiently known electrical connections for electrically contacting the sensor layer 4 from the outside are not shown in Fig. 1 for the sake of simplicity.

[0022] The sensor layer 4 comprises an arrangement of a plurality of electro-optical detectors, each of which forms a pixel of the sensor layer 4 and thus also a pixel of the image recording B.

[0023] The sensor layer 4 can also be designed with multiple channels, for example, as an RGB sensor. The individual electro-optical detectors of the sensor layer 4 can also each be provided with different color filters in order to be able to image several separate wavelength ranges of light with one image sensor 1. Such designs of an image sensor 1 are well known.

[0024] Due to the known optical properties of the optical system of the camera 10 with the image sensor 1 and also due to the known geometry of the image sensor 1 and the camera 10 and the positioning of the image sensor 1 in the camera 10, each object point G of an object recorded with a camera 10 with image sensor 1 (see also Fig. 4) is imaged onto an image point P of the sensor layer 4 in the image plane 7, and can therefore be assigned to an image point P of the camera 10. This is indicated in Fig. 1. The optical system of the camera 10 comprises an objective lens 11 and can also comprise other optical components such as an aperture, optical filters, a shutter, etc. The projection center O of the objective lens 11 (not shown in Fig. 1) can be arranged at a distance of the known focal length fß from the image plane 7. The sensor layer 4 of the image sensor 1 is arranged in the image plane 7.

[0025] The projection center O does not have to be arranged centrally to the sensor layer 4, but the sensor layer 4 can also be arranged off-center and / or asymmetrically to the projection center O. It is also conceivable that several image sensors 1 are arranged in a camera 10, which can also lead to an image sensor 1, or its sensor layer 4, being arranged eccentrically to the projection center O and to the optical axis, which runs through the projection center O and normal to the image plane 7.

[0026] A light ray 6, which emanates from an object point G on the recorded object and passes through the projection center O, is refracted at the transparent cover 5 according to the known law of refraction and strikes the image plane 7, depending on the known refractive indices present (e.g. transition from air to glass at the outer boundary surface 5a and from glass to air at the inner boundary surface 5b), offset by the offset L at the image point P' instead of at the actual image point P. The lens 11, however, naturally bundles in a known manner all light rays emanating from the object point G on the recorded object which pass through the lens 11 (see Fig.4), onto the image point P. For the sake of simplicity, however, only the light ray 6 passing through the projection center O is shown in Fig.1.

[0027] It should also be noted that the refractive indices at the outer interface 5a of the transparent cover 5 do not necessarily have to be the same as at the inner interface 5b of the transparent cover 5 into the sensor housing 2. However, the refractive indices at both optical interfaces 5a, 5b can be assumed to be known.

[0028] The offset L results from the known refractive indices according to the law of refraction and the known geometric dimensions of the image sensor 1, in particular the thickness of the transparent cover 5 and the distance between the inner interface 5b of the transparent cover 5 and the sensor layer 4.

[0029] It is known to mathematically map the influence of optical refraction at the transparent cover 5 of the image sensor 1 on the recorded image and to compensate for it in the image recording B, for example, by shifting the image of the object point G from the image point P' in the image recording B' by the offset L to the image point P of the image recording B (Fig. 1). This geometric imaging error caused by the refraction becomes greater the more the angle of incidence δi of the light beam 6 deviates from the surface normal of the transparent cover 5. The offset L is therefore usually not constant across the entire image plane 7.

[0030] The image sensor 1, or the sensor layer 4, is generally two-dimensional, so that the offset L in the image plane 7 arises from refraction in two coordinate directions (e.g., x and y, see Fig. 6). The offset L at an image point P in the image plane 7 (in both coordinate directions) depends on the angle of incidence θi of a light beam 6, the existing refractive indices and the geometry of the image sensor 1, and possibly also on the position of the image sensor 1 in the camera 10 (particularly if the image sensor is arranged eccentrically to the optical axis).

[0031] However, in (high-)precision applications, such as geomatics or photogrammetry, it was found that despite compensation of the offset L caused by refraction, a recognizable geometric aberration A remains in the image acquisition B, which limits the quality of the image acquisition B. It was recognized that this remaining geometric aberration A is due to the design of the image sensor 1.

[0032] After the sensor recess 3, in which the sensor layer 4 is arranged, is hermetically sealed by the transparent cover 5, a specific gaseous medium and a specific internal pressure Pi are present in the sensor recess 3 due to production reasons. If, for example, an image is taken from an aircraft using the image sensor 1, an ambient pressure p2 prevails in the area of ​​the camera 10 with the image sensor 1 in the aircraft, which will be different from the internal pressure pi. Since the production of the image sensor 1 will take place at a significantly lower altitude than the flight altitude of the aircraft during image capture, the internal pressure pi trapped in the sensor recess 3 will usually be greater than the ambient pressure p2 when using the camera 10. The internal pressure pi can of course also be lower than the ambient pressure p2 when using the camera 10. In this case, the pressure difference D can lead to deformation of the transparent cover 5.If the internal pressure pi in the sensor recess 3 is greater than the ambient pressure p2, the transparent cover 5 will, for example, bulge outward. If the sensor housing 2 is not sufficiently rigid, the sensor housing 2 and thus also the sensor layer 4 can be deformed, for example, bulge outward, if the internal pressure pi is greater than the ambient pressure p2.

[0033] The pressure difference D is preferably expressed as a pressure difference (P2-P1), also as a power of the pressure difference, or as a quotient (P2 / P1) of the two pressures, although the pressure difference D could also be expressed differently. The pressure difference D could also be expressed as the difference between the pressure difference at a first altitude and the pressure difference at a different altitude.

[0034] However, the internal pressure pi can also be influenced by other physical effects, in particular by the temperature in the area of ​​the image sensor 1. According to Amontons' law, the pressure of a gas confined in a constant volume depends on the temperature T, according to the relationship Here, pi denotes the pressure at temperature Ti, which corresponds to the prevailing internal pressure pi, To denotes an initial temperature (for example, the temperature prevailing during the manufacture of the image sensor or a certain assumed temperature), and Yo denotes the volume expansion coefficient at the initial temperature To, which can be assumed to be known for the gas enclosed in the image sensor 1. The temperature T can be measured, for example, with a temperature sensor 13 on the image sensor 1. pp

[0035] Amontons’ law can also be applied in a simplified form — = — T 1 o become a TA.

[0036] For the application of Amontons' law, the volume in image sensor 1 can be considered approximately constant, because image sensor 1 will only deform slightly. However, this influence could also be modeled, for example, by applying the well-known thermal equation of state for ideal gases. The effects of deformations caused by the pressure difference D in image sensor 1 are explained with reference to Figures 2 and 3, where the effects are greatly exaggerated for illustrative purposes.

[0037] In Fig. 2, a portion of the transparent cover 5 is shown in dashed lines in an undeformed initial position. A light ray 6 arriving from an object point G would, due to the law of refraction, be refracted at the two interfaces 5a, 5b on the transparent cover 5 and would leave the transparent cover 5 as a light ray 6', which would strike the sensor layer 4 at the image point P'. Due to a deformation of the transparent cover 5 caused by the pressure difference D - in the example shown, the transparent cover 5 is curved outwards - the orientations of the normals to the interfaces 5a, 5b change. This also changes the angle δi of the incoming light ray 6 to the normal to the outer interface 5a and also the angle δ3 of the outgoing light ray 6" to the normal to the inner interface 5b. The light ray 6 is therefore refracted differently due to the deformation compared to an undeformed transparent cover 5.The emerging light beam 6" therefore hits the sensor layer 4 at a different image point P", which leads to a geometric aberration A. The object point G is thus no longer imaged onto the image point P' (as expected), but onto another actual image point P" of the sensor layer 4.

[0038] The use of reference numerals 6', 6" for the refracted, emerging light rays is for explanatory purposes only. However, light beam 6 remains the same.

[0039] In Fig. 3, a portion of the undeformed sensor housing 2 with the undeformed sensor layer 4 is shown in dashed lines. A light beam 6 arriving from an object point G would strike the undeformed sensor layer 4 at the image point P'. However, due to a deformation of the sensor layer 4 caused by the pressure difference D (in the example shown, the sensor layer 4 is curved outwards), the light beam 6 strikes the sensor layer 4 at the image point P', which also leads to a geometric imaging error A. The object point G is thus no longer imaged onto the image point P' (as expected), but onto another actual image point P' of the sensor layer 4.

[0040] A geometric imaging error A is therefore understood as a shift of the image of an object point G from an expected image point P' of the sensor layer 4 to another actual image point P" of the sensor layer 4.

[0041] The two effects of deformation explained with reference to Figs. 2 and 3 can occur together (which will usually be the case), but can also occur individually. The geometric aberration A can therefore be caused by one of these effects or by both effects. The resulting geometric aberrations A manifest themselves, for example, in a distortion and / or distortion of the recorded image. Such a geometric aberration A can arise at each pixel P of the image plane 7 or only at certain pixels P, but will not be constant across the entire image plane 7.

[0042] Knowledge of a deformation V of the transparent cover 5 and / or the sensor layer 4 in the sensor housing 2 due to a pressure difference D between the internal pressure pi in the sensor recess 3 and the ambient pressure p2 of the camera 10, which prevail at the time of an image recording made with the camera 10 with the image sensor 1, can be used to correct a geometric imaging error A caused thereby in the image recording B' in order to obtain a corrected image recording B. For this purpose, an evaluation unit 21 can be provided, which can also be part of the camera 10.

[0043] The time of image capture B' can be considered, for example, to be the start or end of exposure or a time between the start and end of exposure.

[0044] A geometric aberration A can be determined and corrected for each pixel P of the digital image, but does not have to be determined and corrected for each pixel P.

[0045] The following procedure can be followed, also with reference to Fig.4 and 5.

[0046] The internal pressure pi in the image sensor 1 at the time of image acquisition B' can be assumed to be known. For example, the internal pressure pi corresponds to the prevailing pressure at the time of manufacture of the image sensor 1, which can also be approximated using standard conditions (To=273K, pi=1, 01325 bar). Any temperature influence on the internal pressure pi can be determined, for example, using Amontons' law. For this purpose, the current temperature at the image sensor 1 at the time of image acquisition B' can be recorded by a temperature sensor 13, and a current internal pressure pi can be calculated.

[0047] Likewise, the ambient pressure p2 at the image sensor 1 or at the camera 10 can be assumed to be known at the time of image capture B'. For example, the camera 10 is mounted on a vehicle 12, such as an aircraft. The ambient pressure p2 can be detected by a pressure sensor, for example in the case of a road vehicle, or can be determined from the known flight altitude h using the barometric altitude formula, for example in the case of an aircraft. Of course, the ambient pressure p2 could also be detected in an aircraft using a pressure sensor. If the camera 10 is arranged in a pressurized cabin of an aircraft, the ambient pressure P2 corresponds, for example, to the pressure in the pressurized cabin. However, the camera 10 does not necessarily have to be arranged on a vehicle 12.

[0048] This means that the current pressure difference D at the time of image acquisition B' can be assumed to be known or determined.

[0049] A deformation model 20 is used that models a deformation V of the transparent cover 5 and / or the sensor layer 4 caused by a pressure difference D between the internal pressure p1 in the sensor recess 3 of the image sensor 1 and the ambient pressure p2 of the image sensor 1 or the camera 10. The deformation V can be generally expressed by the deformation model 20 as a function f of the pressure difference D, i.e., V=f(D). The deformation model 20 describes the relationship between the pressure difference D and the deformation V of the transparent cover 5 and / or the sensor layer 4.

[0050] The deformation model 20 can be implemented, for example, in an evaluation unit 21 (Fig. 5). The evaluation unit 21 can be a computer or microprocessor-based hardware. An integrated circuit design, such as an FPGA (Field Programmable Gate Array) or ASIC (Application Specific Integrated Circuit), is also conceivable. The deformation model 20 can be implemented as software running on a microprocessor, but can also be stored in tabular form or as a mathematical formula or function.

[0051] How the deformation model 20 is specifically implemented is not relevant for the invention and a wide variety of models are conceivable, as will be explained in more detail below.

[0052] Using the deformation model 20, a deformation V of the transparent cover 5 and / or the sensor layer 4 at the time of image acquisition B' can be determined from the known pressure difference D at the time of image acquisition B'.

[0053] From the determined deformation V and the known geometric dimensions of the image sensor 1, it can be determined at which pixel P" of the sensor layer 4 a light beam 6 captured by the camera 10 actually strikes the sensor layer 4 and at which pixel P' the light beam 6 should strike the sensor layer 4 without deformation V. Since these are simple geometric relationships, this can be easily determined. This determined shift between the actual pixel P" and the expected pixel P', which represents the geometric aberration A of the pixel P, can be corrected in the image recording B' to correct the geometric aberration A. To do this, the image of an object point G is shifted from the pixel P" in the image recording B', for example, by the geometric aberration A to the pixel P' of the corrected image recording B.In this process, the offset L, which also occurs with an undeformed image sensor 1 (see Fig. 1 above), could also be corrected immediately. However, this offset L could also be corrected later. This offset L could also be taken into account immediately in the geometric aberration A, so that the geometric aberration A is used to correct the image point P.

[0054] Although the geometric aberration A of an image point P is described in the description as a displacement in one coordinate direction (x or y) for the sake of simplicity, it is obvious that, due to the deformation V, geometric aberrations A will usually occur in both coordinate directions of the sensor layer 4 for an image point P. When reference is made to a correction of the geometric aberration A, this means both, namely a displacement in only one coordinate direction or a displacement in both coordinate directions of the sensor layer 4 (which will usually be the case).

[0055] The correction of the geometric aberration A in an image B' due to a deformation V of the transparent cover 5 and / or the sensor layer 4 can be performed in a correction unit 22. The correction unit 22 can be integrated into the evaluation unit 20 and can be implemented, for example, in the form of software. However, the correction unit 22 can also be hardware (computer, microprocessor-based hardware, integrated circuit) separate from the evaluation unit 21.

[0056] The image B corrected for the geometric aberration A can then be further used as required, for example by subjecting the corrected image B to further optical corrections, for example a known deblurring, or by evaluating the corrected image B according to the application.

[0057] The correction of a geometric aberration A of an image point P" can be performed immediately after image acquisition B', or at a later time. If the correction is performed at a later time, the pressure difference D prevailing at the time of image acquisition B' is also stored with image acquisition B' in order to be able to determine the deformation V later.

[0058] To determine the deformation model 20, the following procedure can be followed with reference to Fig.6.

[0059] The image sensor 1 is arranged in a pressure chamber 30 in which a specific pressure PD is set. If necessary, a specific temperature can also be set in the pressure chamber 30, or the temperature can be kept constant. A measuring device 31 is provided in or on the pressure chamber 30, with which a deformation V of the transparent cover 5 and / or the sensor layer 4, or of the part of the sensor housing to which the sensor layer 4 is attached, is measured.

[0060] The deformation V can be measured using any suitable direct or indirect, preferably contactless, measuring device 31. Examples of suitable measuring devices include mechanical measuring devices that determine a shape by means of tactile scanning, such as micrometers or calipers; optical measuring systems that have image, line, and / or pixel sensors, such as 1D, 2D, and 3D laser displacement sensors; measuring devices with imaging sensors, wherein the deformation is determined by image analysis; and measuring devices that combine the above measuring methods. A preferred example of a measuring device 31 is an optomechanical measuring device in which the positioning of the image sensor 1 is mechanical and the measurement is performed optically. The optomechanical measurement of the image sensor 1 can be performed, for example, using known industrial optical 3D measuring systems.An example of a usable optomechanical measuring device is the device called “Vertex 251 UC” manufactured by the American company Micro Vu, Windsor, CA.

[0061] In order to be able to carry out the deformation V of the transparent cover 5 by means of an optical method, visible points can also be arranged on the transparent cover 5 during the measurement.

[0062] The measuring device 31 measures the deformation V of the transparent cover 5 and / or the sensor layer 4 at a plurality of measuring points M (indicated in Fig. 6 for the transparent cover 5) at a specific pressure difference D set in the pressure chamber 20. The majority of the measuring points M are distributed in the plane of the transparent cover 5 and / or the sensor layer 4 (in Fig. 6, for example, in the xy plane) in order to also map the deformation V two-dimensionally. The position of the measuring points M can, for example, be specified relative to a coordinate system on the image sensor 1.

[0063] However, the measurement variable used to map the deformation V is not important. For example, the deformation V could be measured as the distance (e.g. in the z-direction in Fig. 6) between an undeformed reference position and an actual position of a measuring point on the transparent cover 5 and / or the sensor layer 4. From this, a normal to the interfaces 5a, 5b of the deformed transparent cover 5 could also be determined for the transparent cover 5. A difference between a normal to the interfaces 5a, 5b in an undeformed reference position and at a specific pressure difference D could also be determined and mapped to the deformation V. A normal can, for example, be specified as a vector in a coordinate system of the image sensor 1.

[0064] The deformations V measured at the measuring points M could be stored in tabular form and used as deformation model 20. Interpolation between the measuring points M could then be performed using the deformation model 20.

[0065] However, it is of course also possible to use 20 mathematical functions as the deformation model, which approximate the determined deformations V at the measuring points M. The measuring points M or a selection of the measuring points M are to be interpolated using the mathematical function. For this purpose, it can be advantageous to arrange the measuring points M in a grid pattern in the plane. For example, polynomials of a certain degree (e.g., 6th degree polynomials) or free-form curves such as Bezier curves, piecewise polynomial functions (splines), etc. could be used as mathematical functions. The approximation of given support points of a curve (measurement points M) using such mathematical functions is well known and does not need to be explained in more detail here. This can of course also be extended to surfaces, i.e. a two-dimensional approximation.For this purpose, two-dimensional mathematical functions, such as a two-dimensional polynomial of a certain degree or a free-form surface, are used to model a surface.

[0066] To map the deformation V of the entire transparent cover 5 and / or the sensor layer 4 in the deformation model 20, several mathematical functions can be used, each covering a specific area of ​​the transparent cover 5 and / or the sensor layer 4. For example, several mathematical functions distributed in the y-direction are used, with each mathematical function being applied to a specific y-coordinate and approximating several measurement points M in the x-direction to this y-coordinate. This can also be done two-dimensionally if several mathematical functions are used distributed in the x-direction and distributed in the y-direction. With such two-dimensional modeling, the accuracy of the deformation model 20 can be increased.

[0067] The mathematical functions in both directions could be linked multiplicatively. At any position on the image sensor 1 covered by the mathematical functions, a deformation can be determined using each of the two mathematical functions, which are then multiplied to obtain the total influence at that position.

[0068] In an exemplary implementation of the deformation model 20, 6th-degree polynomials (or even a different degree) could be used to approximate the measurement points M in a specific direction, e.g., the x- or y-direction. This would result in several such polynomials in the xy-plane of the image sensor 1, which model the deformation V in the plane. The measurement points M are, of course, chosen appropriately for this purpose. The polynomials can be used to interpolate between the measurement points. The polynomials could be stored in the form of their polynomial parameters, which are determined during curve approximation.

[0069] To increase accuracy, 6th-degree polynomials (or even a different degree) could be used for both directions, e.g., the x- and y-directions. This would allow the deformation V to be modeled more accurately in the surface.

[0070] If a mathematical function is used to approximate the deformation V, it is obvious that the deformation V of the transparent cover 5 and the sensor layer 4, when both are modeled, are usually approximated by different functions. "Different" can also mean that the same function class (e.g., 6th-order polynomial) is used, but the function parameters are different.

[0071] The deformation V of the transparent cover 5 and / or the sensor layer 4 is modeled by the deformation model 20, preferably for different pressure differences D. For this purpose, the deformation V can be measured, for example, for different pressure differences D with the measuring device 31 and modeled in the deformation model 20 for each pressure difference D used. In use, interpolation can be carried out between determined deformations V to certain pressure differences D, or the deformation V for the pressure difference D closest to the current pressure difference D stored in the deformation model 20 is used.

Claims

Patent claims 1. Method for correcting a geometric imaging error (A) of a camera (10) with an image sensor (1) formed by a sensor housing (2), a sensor layer (4) arranged in a sensor recess (3) of the sensor housing (2) with a plurality of pixels (P) and a transparent cover (5) hermetically sealing the sensor housing (2) to the outside, wherein an object point (G) of an object recorded by the camera (10) is imaged onto a pixel (P) of the sensor layer (4), characterized in that a deformation (V) of the transparent cover (5) and / or the sensor layer (4) caused by a pressure difference (D) between an internal pressure (pi) in the sensor recess (3) and an ambient pressure (P2) of the camera (10) is modeled by a deformation model (20),that a currently prevailing pressure difference (D) at the time of image acquisition (B') with the camera (10) is determined based on the internal pressure (pi) and ambient pressure (P2) of the camera (10) known at the time of image acquisition (B'), and that with the determined pressure difference (D) a deformation (V) of the transparent cover (5) and / or the sensor layer (4) at the time of image acquisition (B') is determined with the deformation model (20), that from the determined deformation (V) and the known geometric dimensions of the image sensor (1), a displacement of the image of the object point (G) to a resulting actual image point (P") in the image acquisition (B') caused by the deformation (V) is determined as a geometric imaging error (A), and that the determined geometric imaging error (A) in the image acquisition (B') is corrected to determine a corrected image acquisition (B).

2. Method according to claim 1, characterized in that the geometric imaging error (A) is determined as the displacement of the image of the object point (G) onto an expected image point (P') with an undeformed transparent cover (5) and / or undeformed sensor layer (4) to an actual image of the object point (G) onto the resulting actual image point (P") resulting from the deformation (V).

3. Method according to claim 1 or 2, characterized in that the geometric imaging error (A) also takes into account an offset (L) caused by optical refraction at the undeformed transparent cover (5).

4. Method according to one of claims 1 to 3, characterized in that the geometric imaging error (A) in the image acquisition (B') is corrected by mapping the object point (G) onto the resulting image in the image acquisition (B'). The actual image point (P") in the corrected image (B) is shifted to another image point (P, P') by the geometric aberration (A).

5. Camera with an image sensor (1) formed by a sensor housing (2), a sensor layer (4) arranged in a sensor recess (3) of the sensor housing (2) with a plurality of pixels, and a transparent cover (5) hermetically sealing the sensor housing (2) to the outside, and with an evaluation unit (21) for processing an image (B') taken with the camera (10), wherein an object point (G) of an object recorded with the camera (10) is mapped onto a pixel (P) of the sensor layer (4), characterized in that a deformation model (20) is implemented in the evaluation unit (21) which models a deformation (V) of the transparent cover (5) and / or the sensor layer (4) caused by a pressure difference (D) between an internal pressure (pi) in the sensor recess (3) and an ambient pressure (P2) of the camera (10),that the evaluation unit (21) determines a currently prevailing pressure difference (D) at the time of image acquisition (B') with the camera (10) based on the internal pressure (pi) and ambient pressure (P2) of the camera (10) known at the time of image acquisition (B') and determines a deformation (V) of the transparent cover (5) and / or the sensor layer (4) at the time of image acquisition (B') from the determined pressure difference (D) with the deformation model (20), that the evaluation unit (21) determines a displacement of the image of the object point (G) to a resulting actual image point (P") in the image acquisition (B') caused by the deformation (V) as a geometric imaging error (A), and that a correction unit (22) is provided which corrects the determined geometric imaging error (A) in the image acquisition (B') to determine a corrected image acquisition (B).

6. Camera according to claim 5, characterized in that the evaluation unit (21) determines the geometric imaging error (A) as the displacement of the image of the object point (G) onto an expected image point (P') with an undeformed transparent cover (5) and / or undeformed sensor layer (4) to an actual image of the object point (G) onto the resulting actual image point (P") resulting from the deformation (V).

7. Camera according to claim 5 or 6, characterized in that the evaluation unit (21) also takes into account in the geometric imaging error (A) an offset (L) caused by an optical refraction at the undeformed transparent cover (5).

8. Camera according to one of claims 5 to 7, characterized in that the correction unit (22) corrects the geometric imaging error (A) in the image acquisition (B') by the correction unit (22) adjusting the imaging of the object point (G) onto which The actual image point (P") resulting from the image capture is shifted in the corrected image capture (B) by the geometric aberration (A) to a different image point (P, P').

Images