Method for camera-based determination of a distance in a vehicle interior and vehicle with a camera

Abstract

Method for camera-based determination of a distance of at least one area of ​​an object (2.1, 2.2) located in a vehicle interior (1) to a camera (3) capturing at least a section of the vehicle interior (1), wherein the camera (3) records at least a first (image 1) and a second image (image 2), at least the section of the vehicle interior (1) is illuminated by two light sources (4.1, 4.2) during the recording of the images (image 1, image 2), and the first (image 1) and second (image 2) are combined to obtain distance information, comprising at least the following method steps: - Recording the first (Image1) and second image (Image2), wherein at least the section of the vehicle interior (1) is illuminated by a first light source (4.1) during the recording of the first image (Image1) and by a second light source (4.2) spaced apart from the first light source (4.1) during the recording of the second image (Image2); - Combining the first (Image1) and second (Image2) images to generate at least one delta image (Delta1, Delta2); - Determine at least one region of interest (5) in at least one delta image (Delta1, Delta2); - Determining a quantitative brightness value for the region of interest (5), wherein the brightness level correlates with the distance sought, wherein - to generate a delta image (Delta1, Delta2) a difference or quotient is generated from the first (image1) and second image (image2), and wherein - to generate a first delta image (Delta1) the calculation operation: Image1 / Image2 is performed and to generate a second delta image (Delta2) the calculation operation: Image2 / Image1 is performed, or to generate the first delta image (Delta1) the calculation operation: Image1 - Image2 and to generate the second delta image (Delta2) the calculation operation: Image2 - Image1 is performed; and the first (Delta1) and second delta images (Delta2) are each divided vertically in a horizontal image center into two equal sections (6.1, 6.2) and to generate a total delta image (DeltaTotal) a left section (6.1) of the first delta image (Delta1) is used as the left half (L) of the total delta image (DeltaTotal) and a right section (6.2) of the second delta image (Delta2) is used as the right half (R) of the total delta image (DeltaTotal).

Description

[0001] The invention relates to a method for determining, using a camera, the distance of at least one area of ​​an object located in a vehicle interior to a camera of the type defined in more detail in the preamble of claim 1, and to a vehicle with a camera for carrying out the method.

[0002] Modern vehicles often include not only environmental sensors that detect objects in the vehicle's vicinity, but also sensors for monitoring the vehicle interior. This allows for the detection and tracking of people and objects inside the vehicle, enabling the provision of comfort and safety features. For example, it can detect whether a driver is tired or even falling asleep at the wheel, monitor the driver's attention distribution to determine whether they are paying sufficient attention to the road, or allow entertainment functions to be operated via gesture control.

[0003] To provide these functionalities, sufficiently accurate detection and tracking of people or objects inside the vehicle is required. This particularly involves determining the distance of the people or objects to an interior sensor, such as a camera. Determining distances makes it possible to detect partial or complete obstruction of the driver's field of vision, as well as to ascertain the relative position of objects inside the vehicle and the relative distance between objects.

[0004] Typical methods for distance measurement include the use of stereo cameras, structured light illumination, and / or time-of-flight measurements, for example, of an electromagnetic wave. However, the sensor technology used for these methods is comparatively complex and therefore expensive. Consequently, a relatively large amount of space is required to install the sensor technology in the vehicle, as well as a comparatively high computing capacity to process the resulting sensor data.

[0005] Artificial intelligence can also be used to estimate distances from two-dimensional images. However, such methods are only suitable for known objects, i.e., objects whose dimensions are known, such as the head of a driver.

[0006] There is therefore a need to be able to detect distances between objects in a vehicle interior using the most cost-effective sensor technology and simple algorithms possible.

[0007] German patent application DE 10 2015 010 421 A1 discloses a method for the three-dimensional detection of a vehicle interior. To generate distance information, structured light is projected into the vehicle interior, and light reflected from objects within the interior is detected by a sensor device, such as a camera. At least two images are captured: one during a period when the lights are on and one during a period when they are off. The two images are compared to filter out disruptive glare effects, i.e., excessive background brightness. The vehicle interior can also be illuminated by a second light source with unstructured light. The method disclosed in the patent application can also be used to track the movements of objects. This enables, for example, the recognition of gestures for gesture control.However, the hardware required to generate the structured light and to evaluate the generated sensor signals is comparatively complex and therefore expensive.

[0008] Furthermore, US patent 2015 / 0226553 A1 discloses a motion sensor device with multiple light sources. The motion sensor device captures a first image of a scene while the scene is illuminated by a first light source. Subsequently, a second image of the scene is captured while the scene is illuminated by a second light source. The difference between the two images is then used to determine the distance of an object located within the scene.

[0009] Furthermore, US patent 6,600,168 B1 discloses a high-speed laser-based 3D imager. This imager can project different light patterns onto a scene and capture them sequentially with a camera to determine distances from a layered sequence of images.

[0010] Furthermore, DE 11 2009 004 059 T5 discloses a method for removing blur from an image and a recording medium on which the method is recorded. In this process, blur information is estimated and, in a sharpening step, a blurry image is sharpened taking the blur information into account.

[0011] Furthermore, US 2007 / 0177817 A1 reveals region-based image noise reduction.

[0012] Furthermore, WO 2017 / 179511 A1 discloses a device and a method for processing information to detect the position of an object.

[0013] Furthermore, US patent 2013 / 0182077 A1 reveals improved contrast for object recognition and characterization through optical imaging. In this process, a camera captures images of a scene. The scene is actively illuminated during image capture. Pixels corresponding to an object are identified in the images, and a 3D model of the object is generated based on this identification.

[0014] The present invention is based on the objective of providing an improved method for camera-based determination of the distance of at least one area of ​​an object located in a vehicle interior to a camera, which on the one hand allows the use of comparatively simple and therefore cost-effective hardware components and whose algorithm for processing sensor signals is comparatively simple.

[0015] According to the invention, this problem is solved by a method for camera-based determination of the distance between at least a region of an object located in a vehicle interior and a camera capturing at least a section of the vehicle interior, comprising the features of claim 1. Advantageous embodiments and further developments, as well as a vehicle with a camera, two light sources, and a processing unit, are described in the dependent claims.

[0016] In a method for camera-based determination of a distance of the type mentioned above, at least the following method steps are carried out according to the invention: - Recording a first and a second image with the camera, wherein at least the section of the vehicle interior is illuminated by a first light source during the recording of the first image and by a second light source spaced apart from the first light source during the recording of the second image; - Combining the first and second images to generate at least one delta image; - Determine at least one region of interest in at least one delta image; - Determining a quantitative brightness value for the region of interest, where the brightness level correlates with the distance being sought, wherein to generate a delta image, a difference or quotient is created from the first and second images, and wherein To generate a first delta image, the arithmetic operation: Image1 / Image2 is performed, and to generate a second delta image, the arithmetic operation: Image2 / Image1 is performed; or to generate the first delta image, the arithmetic operation: Image1 - Image2 is performed, and to generate the second delta image, the arithmetic operation: Image2 - Image1 is performed; and the first and second delta images are each divided vertically into two equal sections at a horizontal image center, and to generate a complete delta image, a left section of the first delta image is used as the left half of the complete delta image, and a right section of the second delta image is used as the right half of the complete delta image.

[0017] The method according to the invention allows the use of relatively simple hardware components such as a monocular camera and two light sources. No special requirements are placed on the two light sources, except that they are configured to emit a specific minimum brightness within a defined activation time. Furthermore, the first and second images can be combined using relatively simple arithmetic operations, which allows the delta image to be generated relatively quickly, even with less powerful hardware. To determine the desired distances, no further computationally intensive operations are required, apart from evaluating the brightness value. This, however, represents a particularly simple and therefore quick calculation step.

[0018] The method according to the invention is based on the idea that an object located in the vehicle interior, or at least an illuminated section of the object, appears differently bright in the respective camera image depending on its position in the vehicle interior during illumination by the first and second light sources. A corresponding difference in brightness between the first and second images then correlates with the desired distance of the object from the camera.

[0019] The brightness with which an object appears in a given image depends, among other factors such as the object's reflectivity or the luminosity of the two light sources, on the object's distance from those light sources. Since the position of the light sources within the vehicle interior and their respective luminosities are known, it is possible to determine the object's distance from the camera based on the resulting brightness values, or rather, the brightness difference value of the delta image.

[0020] The area of ​​interest can be of any size, meaning it can correspond to any percentage of the delta image. Multiple areas of interest can be defined or evaluated for at least one delta image. For example, an area of ​​interest can represent 5 percent, 10 percent, 20 percent, or even fractions or multiples thereof of a delta image. The entire delta image can also be rasterized into adjacent areas of interest. These areas of interest are then distributed two-dimensionally adjacent to each other on the delta image. For example, an area of ​​interest can have a square base. However, an area of ​​interest can also be rectangular or any polygonal shape.In particular, areas of interest lie in the image areas of the delta image where objects are expected in the vehicle interior, such as a typical occupant volume in the vehicle interior.

[0021] To generate a delta image, a difference or quotient is calculated from the first and second images. For example, the first image can be divided by the second image, or vice versa. Alternatively, the second image can be subtracted from the first, or vice versa. The delta image thus only reflects the brightness differences resulting from the varying illumination of the vehicle interior using the first and second light sources. Calculating a difference or quotient from two images is also feasible with relatively low-powered hardware, meaning it can be performed quickly within a defined maximum timeframe.

[0022] It is intended that, to generate a first delta image, the calculation operation: Image1 / Image2 is performed, and to generate a second delta image, the calculation operation: Image2 / Image1 is performed, or, to generate the first delta image, the calculation operation Image1-Image2 is performed, and to generate the second delta image, the calculation operation Image2-Image1 is performed, and that the first and second delta images are each divided vertically into two equal sections in a horizontal image center, and that, to generate a total delta image, a left section of the first delta image is used as the left half of the total delta image and a right section of the second delta image is used as the right half of the total delta image.

[0023] For example, the first light source is positioned to the left of the camera from the camera's perspective and is therefore located inside the vehicle on the passenger side, while the second light source is positioned to the right of the camera from the camera's perspective and is therefore located on the driver's side. Generating the combined delta image and using it to determine quantitative brightness differences allows for increased robustness of the algorithms used to determine brightness differences and reduces the complexity of the evaluation.

[0024] When dividing image 1 by image 2, brightness values ​​greater than 1 are generally generated for objects located in the left half of the image, and brightness values ​​less than 1 for objects located in the right half. This is because the first light source, located on the passenger side, was activated when image 1 was generated (thus, an object located to the left of the camera's optical axis appears brighter in image 1 than in image 2). This makes a direct comparison of the left and right halves of the image difficult. However, if, for example, image 1 / image 2 is used for the left half of the overall delta image and image 2 / image 1 for the right half, then the half of the overall delta image used for each half is the one on which the light source was active when the respective image was generated.This makes it easier to compare the two halves of the image, since objects in both the left and right halves of the image will produce brightness values ​​> 1 in the overall delta image.

[0025] This statement can be applied to the generation of the first and second delta images by subtracting the images Image1 and Image2, whereby instead of brightness values ​​above and below 1, brightness values ​​above and below 0 are obtained.

[0026] In general, it would also be possible to use image 2 / image 1 for the left half of the overall delta image and image 1 / image 2 for the right half (accordingly, brightness values ​​less than 1 are obtained for both halves of the image), or to use image 2-image 1 for the left half of the overall delta image and image 1-image 2 for the right half (resulting in negative brightness values).

[0027] Generating the first and second delta images by difference calculation allows for the detection of relatively small brightness differences, thus improving the accuracy of the distance determination method. However, the delta image generated by difference calculation exhibits values ​​close to 0, resulting in comparatively high noise levels in the overall delta image. This complicates the evaluation and, consequently, the distance determination. Distance determination is further complicated by reflective surfaces. In contrast, generating the first and second delta images by ratio calculation allows for a more robust evaluation, as the brightness values ​​in the delta image are close to 1, resulting in less pronounced noise.

[0028] An advantageous further development of the method provides that two light sources lying in a common horizontal plane are used to illuminate at least the section of the vehicle interior. These light sources preferably lie in a common vertical and / or horizontal plane with the camera, and in particular, the two light sources have the same horizontal distance to the camera. This symmetrical arrangement of light sources and camera simplifies the effort required to evaluate the generated images and thus to determine the distance. This effect can be further enhanced if the camera is positioned parallel to the vehicle's longitudinal axis within the vehicle interior. With the same luminosity of both light sources, it is then no longer necessary to consider their installation position within the vehicle interior when evaluating the brightness difference in the differential image.

[0029] The difference in brightness of the light reflected from an object during successive illumination by the first and second light sources depends, firstly, on the object's orthogonal distance to the vertical plane in which the light sources and camera are located, i.e., a distance along the vehicle's longitudinal axis. If the object is closer to this vertical plane, it will exhibit a greater difference in brightness in the delta image than an object located further away from the vertical plane. Secondly, due to the symmetrical installation of the light sources and camera, the difference in brightness also depends on the object's orthogonal distance to the camera's optical axis, lying in a horizontal plane. If the object lies on the optical axis, it is illuminated equally by both light sources, resulting in no difference in brightness in the delta image.However, this is not a disadvantage for determining the distance between vehicle occupants and the camera, since vehicle occupants, or body parts of vehicle occupants, are typically located on their vehicle seat, which has a sufficient distance to the vehicle's longitudinal axis, resulting in relevant objects having a sufficient distance to the optical axis of the camera.

[0030] Distance determination methods using a stereo camera and / or structured light often fail to determine the distance to objects located close to the sensor with insufficient accuracy, or even at all. This can occur, for example, because the object is only within the detection range of one of the stereo cameras, or because the light pattern only partially covers the object. However, the method according to the invention reliably enables distance determination with sufficient accuracy, even for objects located very close to the camera.

[0031] According to a further advantageous embodiment of the method, an offset is added to or subtracted from at least one of the first and second images during the calculation. For example, a value of 0.01 can be chosen as the offset. The offset can be added to or subtracted from the first and / or the second image. Using the offset, division by zero can be avoided. This prevents the formation of a quotient between the first and second images.

[0032] A further advantageous embodiment of the method provides that, to determine the quantitative brightness value, brightness values ​​above a defined maximum and below a defined minimum are ignored, and in particular, an average brightness value is determined for the region of interest. If reflections occur in at least one of the images produced by the camera, the area of ​​the image containing the reflections will exhibit particularly high brightness. This brightness can be mistaken for a particularly short or long distance. Similarly, objects can also cast shadows, resulting in particularly dark regions in an image. An upper and lower limit, i.e., the maximum and minimum brightness, can then be defined, whereby brightness values ​​above or below these limits are ignored for distance determination.This prevents light-reflecting objects from being mistaken for a particularly close object and particularly dark objects from being mistaken for a particularly distant object (or vice versa, depending on how the first and second delta images are generated).

[0033] Calculating an average brightness value for individual regions of interest enables a particularly fast and simple determination of distances for these regions. This eliminates the need to determine a distance value for each individual pixel of the delta image. As a result, the computational effort of the method according to the invention can be reduced even further.

[0034] To evaluate the brightness values, a brightness histogram can be created for the individual regions of interest, for example.

[0035] According to a further advantageous embodiment of the method, at least one filter is applied to at least one section of a delta image, in particular a blur filter, Fourier filter, or bilateral filter. Different image areas can also be processed with different filters. Reflections and / or artifacts in the images can be reduced using filters.

[0036] A further advantageous embodiment of the method according to the invention provides for the use of artificial intelligence to evaluate a delta image. The use of artificial intelligence, such as neural networks, increases the accuracy of the distance calculation, but also increases the complexity of the computational operations to be performed.

[0037] Object recognition is preferably applied to the first and / or second image. The first and / or second image generated by the camera can thus also be analyzed for object recognition, for example, to support gesture control of vehicle functions. Proven image recognition algorithms can be used for this purpose, enabling, for example, the detection of edges, surfaces, or similar features.

[0038] According to the invention, a vehicle equipped with a camera, two light sources, and a processing unit is configured to carry out the method described above. The vehicle can be any type of vehicle, such as a car, truck, van, bus, or the like. The camera is preferably a monocular camera, as monocular cameras have a particularly simple design compared to complex camera systems, such as stereo cameras, and are therefore cost-effective. Of course, it is generally possible to use more than one camera to monitor the vehicle interior. The two light sources are preferably arranged in a common horizontal plane. In particular, the camera is also located in this horizontal plane.The effort required to carry out the method according to the invention can be reduced if the two light sources are equidistant from the camera in the horizontal plane. For example, the two light sources and the camera are integrated into a vehicle's dashboard. The processing unit can be, for example, a central on-board computer, a telematics unit, a control unit of a vehicle subsystem, or the like.

[0039] Further advantageous embodiments of the inventive method for camera-based distance determination and of the vehicle also result from the exemplary embodiments, which are described in more detail below with reference to the figures.

[0040] This shows: Fig. 1 a schematic representation of a vehicle according to the invention with a camera capturing the interior of the vehicle; Fig. 2. A schematic representation of an arrangement consisting of a camera and two light sources for producing a first and a second image of the vehicle interior; and Fig. 3 a schematic flow chart of a method according to the invention.

[0041] Fig. Figure 1 shows a top view of a vehicle 7 according to the invention, comprising a vehicle interior 1, a camera 3 capturing at least sections of the vehicle interior 1, and a processing unit 8 for evaluating camera images generated by the camera 3. The field of view of the camera 3 is blocked by the vehicle seats, resulting in a substantially arrow-shaped or Christmas tree-shaped detection area 9. The camera 3 is arranged parallel to a longitudinal axis 10 of the vehicle and, in particular, centrally within the vehicle 7. Similarly, an optical axis 11 of the camera 3 runs parallel to the longitudinal axis 10 of the vehicle.

[0042] Furthermore, in Fig. 1 so-called “head motion boxes” 12, also referred to as “head motion boxes”, are shown. These head motion boxes 12 can be detected using the camera 3, and the distance between the heads of vehicle occupants (not shown) and the camera 3 can be determined. For this purpose, a method according to the invention is used, which will be described in more detail below.

[0043] To determine the distance according to the inventive method, the vehicle interior 1 is illuminated with two light sources 4.1, 4.2, wherein the camera 3 is positioned in Fig. The first image shown, Image 1, is recorded when the first light source 4.1 emits light, and the second image, Image 2, is recorded when the second light source 4.2 illuminates the vehicle interior 1. The images Image 1 and Image 2 generated by the camera 3 are then at least temporarily stored in the processing unit 8. An arrangement of the camera 3 and the two light sources 4.1 and 4.2 in the vehicle interior 1 is shown in Fig. 2 shown.

[0044] Preferably, the camera 3 and the two light sources 4.1 and 4.2 lie in a common horizontal plane E. H , i.e., they have the same vertical height. Preferably, the first light source 4.1 and the second light source 4.2 are positioned at the same horizontal distance x1 from the camera 3 in this horizontal plane. The central placement of the camera 3 in the vehicle interior 1 thus results in symmetrical illumination of the vehicle interior 1. This simplifies the determination of the required distance values ​​from the recorded brightness values. The two light sources 4.1 and 4.2 can emit light in the visible spectrum or, for example, infrared light. Accordingly, the camera 3 is configured to detect the wavelength components emitted by the light sources 4.1 and 4.2, or the light reflected back from the vehicle interior 1. Furthermore, the camera 3 and the two light sources 4.1 and 4.2 can be arranged as shown in Fig. 2 shown, also in a common vertical plane E V However, this is less significant for simplifying the calculations to be performed than the symmetrical arrangement of the two light sources 4.1, 4.2 relative to the camera 3 in the horizontal plane E. H The compact, symmetrical arrangement allows the two light sources 4.1 and 4.2 and the camera 3 to be installed in a single housing with shared electronics (e.g., power supply). This reduces costs compared to installing separate light sources 4.1 and 4.2.

[0045] Inside the vehicle interior 1 are a first object 2.1 and a second object 2.2. The two objects 2.1 and 2.2 each have a distance a lying in the horizontal plane and perpendicular to the optical axis 11. x towards optical axis 11. The distance a x is different for the two objects 2.1 and 2.2, with only the distance a being different.x for the second object 2.2 in Fig. 2 is shown. Furthermore, the first object 2.1 has a distance d 11 to the first light source 4.1 and a distance d 12 to the second light source 4.2. Similarly, the second object 2.2 has a distance d 21 to the first light source 4.1 and a distance d 22 to the second light source 4.2.

[0046] The vehicle interior 1 is now illuminated successively by light sources 4.1 and 4.2, with camera 3 recording the first and second images (Image 1, Image 2) in each case. The difference between the first and second images (Image 1, Image 2) is then determined, as described in the following section. Fig. 3. In the two images, Image 1 and Image 2, a difference in brightness of the light reflected back from the two objects 2.1 and 2.2 is visible. The difference in brightness depends on both the horizontal distance a xThe brightness difference is determined by the distance of objects 2.1 and 2.2 to the optical axis 11, as well as by the distance of both objects 2.1 and 2.2 to the camera 3 in the direction of the vehicle's longitudinal axis 10 (not shown). The brightness difference is greater the further objects 2.1 and 2.2 are orthogonal to the optical axis 11, and the closer the two objects 2.1 and 2.2 are to the camera 3 in the direction of the vehicle's longitudinal axis 10. The first object 2.1 is closer to the camera 3 than the second object 2.2, which is why the first object 2.1 produces a comparatively large brightness difference and the second object 2.2 produces a comparatively small brightness difference.

[0047] The difference in brightness results from the fact that the light intensity of the light propagating from light sources 4.1 and 4.2 decreases with increasing distance from the respective light source 4.1, 4.2. The closer an object 2.1, 2.2 is to the respective light source 4.1, 4.2, the brighter it appears in the respective image 1, image 2. The first object 2.1 thus appears comparatively dark in the first image 1 and comparatively bright in the second image 2. Accordingly, it produces a comparatively large difference in brightness. The second object 2.2, on the other hand, exhibits a comparatively similar distance d. 21 , d 22 The first and second light sources 4.1 and 4.2 are affected. Accordingly, the second object 2.2 appears with a similar brightness in both images, Image 1 and Image 2. Therefore, the second object 2.2 produces only a comparatively weak difference in brightness.

[0048] Fig. Figure 3 schematically shows a sequence of a method according to the invention. In a process step 301, the first and second images, Image 1 and Image 2, are generated. By way of example, the image content of both images, Image 1 and Image 2, is the vehicle interior 1 with a person driving the vehicle, as seen from the perspective of camera 3.

[0049] In process step 302, the two images, Image1 and Image2, are combined. A first delta image, Delta1, and a second delta image, Delta2, are shown. For example, the first delta image, Delta1, was created by dividing the first image, Image1, by the second image, Image2. The second delta image, Delta2, was created, for example, by dividing the second image, Image2, by the first image, Image1. The two delta images, Delta1 and Delta2, are vertically divided at a horizontal center point into two equal sections, 6.1 and 6.2.

[0050] In process step 303, a total delta image, DeltaGesamt, is generated from the left section 6.1 of the first delta image, Delta1, and the right section 6.2 of the second delta image, Delta2. The left section 6.1 of the first delta image, Delta1, thus forms the left half L, and the right section 6.2 of the second delta image, Delta2, forms the right half R of the total delta image, DeltaGesamt.

[0051] In process step 304, at least one filter is applied to at least one section of the total delta image DeltaTotal. This might be, for example, a blur filter. This reduces reflections and / or artifacts. The resulting brightness values, or shades, are indicated by hatching in the filtered image.

[0052] In process step 305, at least one interest region 5 is determined in the overall delta image DeltaTotal. For example, only one interest region 5 is shown for the right half R. For the left half L, a multitude of interest regions 5 are shown, arranged adjacent to each other in a two-dimensional matrix. This enables computationally efficient distance determination for the entire detection range 9 of camera 3.

[0053] In process step 306, a brightness distribution of interest region 5 is evaluated. For example, a histogram 13 can be generated for the respective interest region 5, with a brightness value H plotted on the abscissa and an intensity I on the ordinate. To determine a brightness representative of interest region 5, the brightness or brightness difference for interest region 5 can then be averaged. A corresponding mean value is available for the in Fig. Histogram 13, shown in Figure 3, is plotted at a brightness value of 130. This brightness value corresponds to the brightness difference of an object 2.1, 2.2 between images 1 and 2, and thus to the desired distance of the respective object 2.1, 2.2 to the camera 3.

[0054] A possible assignment of brightness values ​​to distances can be determined, for example, by calibration, and corresponding calibration data can be stored in the computing unit 8.

[0055] The method enables the use of relatively simple hardware and the application of relatively simple computational algorithms. The method according to the invention can therefore be applied particularly cost-effectively and implemented with technical ease.

Claims

[1] A method for determining the distance of at least one area of ​​an object (2.1, 2.2) located in a vehicle interior (1) to a camera (3) capturing at least a section of the vehicle interior (1), wherein the camera (3) records at least a first (image 1) and a second image (image 2), at least the section of the vehicle interior (1) is illuminated by two light sources (4.1, 4.2) during the recording of the images (image 1, image 2), and the first (image 1) and second (image 2) are combined to obtain distance information, comprising at least the following method steps: - Recording the first (Image1) and second image (Image2), wherein at least the section of the vehicle interior (1) is illuminated by a first light source (4.1) during the recording of the first image (Image1) and by a second light source (4.2) spaced apart from the first light source (4.1) during the recording of the second image (Image2); - Combining the first (Image1) and second (Image2) images to generate at least one delta image (Delta1, Delta2); - Determine at least one region of interest (5) in at least one delta image (Delta1, Delta2); - Determining a quantitative brightness value for the region of interest (5), wherein the brightness level correlates with the distance sought, wherein - to generate a delta image (Delta1, Delta2) a difference or quotient is generated from the first (image1) and second image (image2), and wherein - to generate a first delta image (Delta1) the calculation operation: Image1 / Image2 is performed and to generate a second delta image (Delta2) the calculation operation: Image2 / Image1 is performed, or to generate the first delta image (Delta1) the calculation operation: Image1 - Image2 and to generate the second delta image (Delta2) the calculation operation: Image2 - Image1 is performed; and the first (Delta1) and second delta images (Delta2) are each divided vertically in a horizontal image center into two equal sections (6.1, 6.2) and to generate a total delta image (DeltaTotal) a left section (6.1) of the first delta image (Delta1) is used as the left half (L) of the total delta image (DeltaTotal) and a right section (6.2) of the second delta image (Delta2) is used as the right half (R) of the total delta image (DeltaTotal). [2] Method according to claim 1, characterized by, that to illuminate at least the section of the vehicle interior (1) two in a common horizontal plane (E H ) lying light sources (4.1, 4.2) are used, which are preferably in a common vertical (E V ) and / or horizontal plane (E H ) with the camera (3), wherein in particular the two light sources (4.1, 4.2) have the same horizontal distance (x1) to the camera (3). [3] Method according to claim 1 or 2, characterized by , that in the calculation of the first image (Image1) and the second image (Image2) an offset is added to or subtracted from at least one of the images (Image1, Image2). [4] Method according to any one of claims 1 to 3, characterized by, that to determine the quantitative brightness value, brightness values ​​above a specified brightness maximum and brightness values ​​below a specified brightness minimum are ignored and, in particular, an average brightness value for the region of interest (5) is determined. [5] Method according to any one of claims 1 to 4, characterized by , that at least one filter is applied to at least one section of at least one delta image (Delta1, Delta2), in particular a blur filter, Fourier filter or bilateral filter. [6] Method according to any one of claims 1 to 5, characterized by , that artificial intelligence is used to evaluate a delta image (Delta1, Delta2). [7] Method according to any one of claims 1 to 6, characterized by , that object recognition is applied to the first (Image1) and / or the second image (Image2). [8] Vehicle (7) with a camera (3), two light sources (4.1, 4.2) and a computing unit (8), characterized by , that the camera (3), the two light sources (4.1, 4.2) and the computing unit (8) are configured to carry out a method according to one of claims 1 to 7.

Images