Method for monitoring a taillight, monitoring system and vehicle

Abstract

The invention relates to a method for monitoring a taillight (5) of a vehicle (1) by means of a rearwardly directed camera (8), wherein the taillight (5) has a lighting unit (7a, 7b, 7c). The taillight (5) and / or a lamp environment (U5) illuminated by the taillight (5) is / are detected by the at least one camera (8) and camera signals (S8) are output. A camera image containing pixels is provided depending on the output camera signals (S8), wherein the taillight (5) of the vehicle (1) or the lamp environment (U5) is represented in taillight / lamp-related pixels of the respective camera image. Brightness values at least for the taillight / lamp-related pixels of the respective camera image are determined and evaluated. A functional status of the taillight (5) is determined and output depending on the determined brightness values.

Description

[0001] Method for monitoring a rear lamp, monitoring system and vehicle

[0002] The invention relates to a method for monitoring a rear light of a vehicle, in particular a commercial vehicle, with a camera system, a monitoring system and a vehicle with a monitoring system.

[0003] Functional and visible taillights on vehicles are a key safety aspect for road traffic and are therefore subject to numerous legal regulations. For example, the German Road Traffic Licensing Regulations (StVZO) regulate which taillights may be installed on a motor vehicle or commercial vehicle and which taillights must be installed at a minimum. Examples of taillights include taillights, brake lights, indicators, reflectors, rear fog lights, reversing lights, and license plate lights.

[0004] From the prior art, luminaire monitoring systems for such rear lights are known, which are based on the measured power consumption or on the monitoring of the supply line, in particular on monitoring the line interruption (filament burnt out) of the individual bulbs or LEDs of the rear lights.

[0005] In commercial vehicles, especially trailers, the lights are monitored by a trailer light control unit on the motor vehicle or towing vehicle, so that the monitoring in the trailer depends on whether and to what extent the motor vehicle carries out light monitoring.

[0006] Even a self-test of rear lights, for example by means of integrated photodiodes next to the individual bulbs / LEDs of the respective lighting units of the rear light, leads to problems, since, for example, an external covering, e.g. due to dirt or snow covering the entire rear light, cannot be reliably detected and rear lights must also report a detected error to a control unit, i.e. must be able to communicate, which in turn is complex and expensive.

[0007] Ultimately, however, the light monitoring systems described in the prior art fail to achieve the actual goal, namely ensuring that the taillights are visible or recognizable to following traffic when they are illuminated, i.e., that they are actually functioning. For example, the brake light may not be visible due to a snow-covered taillight, even though the taillight itself is functioning and therefore cannot be detected as faulty by the line monitoring system.

[0008] The invention is based on the object of providing a method for monitoring the rear lights of a vehicle, with which the functionality of the rear lights can be monitored simply and reliably. Furthermore, the object of the invention is to provide a monitoring system and a vehicle with which the method can be implemented.

[0009] This object is achieved by a method, a monitoring system, and a vehicle according to the independent claims. The subclaims specify preferred developments.

[0010] The inventive method for monitoring a rear light of a vehicle with a camera system, in particular as a component of a reversing assistance system, wherein the camera system has a rearward-facing camera. The camera can thus be directed toward a rear space behind the vehicle or behind the respective vehicle part on which the camera is located. The rear light has a lighting unit. The method comprises the following steps:

[0011] The rear light of the vehicle and / or a lighting area illuminated by the rear light is detected by the camera on the vehicle and camera signals are output.

[0012] A camera image with pixels is displayed depending on the output camera signals, with the vehicle's taillight and / or a lighting area illuminated by the taillight being displayed in lighting pixels of the respective camera image. The pixels of the camera image in which the taillight and / or the lighting area are displayed are therefore referred to as lighting pixels.

[0013] Brightness values ​​are determined and recorded for at least the luminous pixels of the respective camera image. This may mean that a value characterizing the brightness is determined and processed for at least the luminous pixels or some of the luminous pixels.

[0014] The functional status of the rear light is determined and output based on the measured brightness values, or at least based on the brightness values ​​assigned to the respective rear light displayed in the respective camera image. The functional status can be output, for example, on a user interface in the form of a graphic and / or acoustic signal.

[0015] It is possible for the camera system to have more than one camera and also for the camera system to have more than one rear-facing camera. It is also possible for the vehicle to have more than one taillight. In this case, several taillights or all of the vehicle's taillights can be captured by the same or different cameras. The method can thus be carried out using one camera or multiple cameras on one taillight or multiple taillights simultaneously, or using multiple cameras in parallel for one or more taillights.

[0016] It can be provided that before one of the above-mentioned steps for monitoring the rear lights, i.e. at least before determining and outputting a functional status of the at least one rear light as a function of the determined brightness values, a check is first carried out to determine whether the at least one rear light is activated. For this purpose, an activation signal transmitted via a CAN bus can be evaluated, which transmits the information as to whether the respective rear light is activated or not. The method is therefore only carried out when the respective rear light to be monitored is also switched on or activated, whereby effort and computing power can be saved and the output of the functional status is limited to the relevant period. According to the invention, a monitoring system and a vehicle with such a monitoring system are also provided.

[0017] The monitoring system therefore automatically detects whether the vehicle's taillights are functional or visible, and can therefore be seen by following traffic. This can either directly image the lights themselves or the surroundings illuminated by the lights when they are activated, such as the roadway, buildings, the driver's own vehicle, other vehicles, etc. The functional status of the taillights therefore includes not only their technical functionality, which can be impaired by a defective power cable and / or a defective LED, but also their actual perceptibility, i.e., whether the taillights can fulfill their function of warning and / or increasing visibility for other road users.

[0018] In particular, it overcomes the disadvantage of conventional light monitoring, whereby covered taillights, such as those covered by dirt or snow, are recognized as functional, but are no longer (fully) visible or perceptible to other road users, thus preventing them from fulfilling their intended function. This is achieved by directly observing the taillights or by observing the area surrounding the lights.

[0019] Another advantage is that the camera, which supports the driver as part of a reversing assistance system when maneuvering, is also used to monitor the rear lights. This allows this camera to fulfill a dual function. Such cameras, especially reversing cameras, are already widespread and will be mandatory in commercial vehicles in the future, so they will already be present. Therefore, no additional hardware is required; an existing camera can be used.

[0020] The inventive method for monitoring lights can thus simplify / automate and accelerate operational processes, such as the departure check, which involves a functional check of the rear lights. Furthermore, it increases the safety of vehicles, especially autonomous vehicles that operate without a driver controlling the vehicle and / or monitoring the journey.

[0021] According to one embodiment of the method, the brightness values ​​are evaluated or the functional status of at least one rear light is determined by creating a histogram and / or a pattern representation (in a higher-dimensional feature space) from the provided camera image, wherein the histogram and / or the feature representation depicts a distribution of the brightness values ​​of at least the light pixels of the respective camera image. In this way, a brightness distribution characteristic of the current situation can be used to determine the functionality of the rear lights. If a pattern representation is used, not only brightness values ​​but also, for example, color values, saturations, and contrasts are available as additional features for evaluation, for example using machine learning tools.

[0022] According to one embodiment, it can be provided that, in order to determine the functional status of the at least one rear light, the histogram created from the camera image is compared with a reference histogram, wherein the reference histogram is created from a reference camera image. Alternatively or cumulatively, it can be provided that, in order to determine the functional status of the at least one rear light, the pattern representation created from the camera image is compared with a reference pattern representation, wherein the reference pattern representation is created from the reference camera image. Such a comparison therefore includes a comparison of historical, previously recorded (reference) camera images with a current camera image or the resulting brightness distributions.

[0023] In this case, it can be provided that the reference camera image shows fully functional rear lights of the vehicle and / or the area around the lights illuminated by the fully functional rear lights of the vehicle. Functionality is therefore easily checked by whether the brightness distribution at the time the camera image is recorded matches a brightness distribution with functioning rear lights. If, for example, the light pixels in the reference camera images are shown brighter, i.e. with a higher intensity, when switched on, than in the current camera image when the rear lights are switched on, it can be easily concluded that the rear light currently being recorded is defective or covered, e.g. dirty or covered in snow, and action can be taken accordingly.

[0024] The functional status can then be determined and output as a faulty or covered rear light or taillight if the histogram created from the camera image deviates from the reference histogram and / or the pattern representation created from the camera image deviates from the reference pattern representation. This enables simple and reliable determination and output of functional capability.

[0025] On the other hand, it can also be provided that the reference camera image shows deactivated taillights of the vehicle and / or the surrounding light when the vehicle's taillights are deactivated. Functionality is therefore simply checked by whether the brightness distribution at the time the camera image is recorded (with activated taillights) differs from a brightness distribution with deactivated taillights (reference camera image), which is to be assumed since the recorded brightness should change. It can then be determined and output as the functional status that the taillight or taillights are defective or covered if the histogram created from the camera image corresponds to the reference histogram and / or the pattern created from the camera image corresponds to the reference pattern, since in this case the brightness has not changed.This also enables simple and reliable determination and output of functionality.

[0026] According to a further embodiment, the histogram is subjected to a brightness adjustment before being compared with the reference histogram and / or the pattern representation is subjected to a brightness adjustment before being compared with the reference pattern representation, wherein the brightness values ​​displayed in the histogram or the pattern representation are adjusted depending on the current ambient brightness. This takes into account that the ambient brightness can vary greatly, for example depending on the time of day or the environment, which also has an impact on the histogram and / or the pattern representation or the brightness values ​​displayed therein. In order to take this effect into account, the ambient brightness is subtracted or the brightness of the histogram and / or the pattern representation is standardized accordingly in order to provide a histogram or a pattern representation for comparison with the reference, regardless of the ambient brightness.The ambient brightness can be determined, for example, by averaging the brightness values ​​of all pixels of the camera image or the brightness values ​​of all pixels except the luminaire pixels.

[0027] According to a further embodiment, it can be provided that, in order to determine the functional status of the rear light, it is determined whether the brightness values ​​of the histogram and / or the pattern representation lie in a target brightness range assigned to the respective rear light. Each rear light can be assigned several target brightness ranges. It can therefore also be checked whether specific brightness values ​​for the respective rear light are recorded in the histogram and / or in the pattern representation, whereby this can, for example, depend on the color of the rear light or the lighting units contained therein. Each lighting unit of the respective rear light can therefore also be assigned a target brightness range, which can then be read off for each light from the histogram and / or the pattern representation.

[0028] It can also be provided that a supplementary determination is made as to whether at least a specified target pixel proportion of brightness values ​​lies within the target brightness range assigned to the respective rear light. This means that not only is a check carried out as to whether a specific brightness value occurs in the histogram and / or in the pattern display, but also the frequency of the brightness values ​​that lie within the brightness range. In this way, a statement can also be made as to the degree of functionality, e.g. partially defective or partially covered. It can then be determined and output as the functional status that the or a rear light is defective or covered if, for example, the brightness values ​​of the histogram and / or pattern display created from the camera image do not lie within the assigned target brightness range or do not lie within the at least one assigned target brightness range for the specified target pixel proportion.This makes it easy to check and display the functionality of the rear lights.

[0029] According to another embodiment of the method, luminaire monitoring is further improved or made more stable by a calibration step, which further limits the number of pixels used to evaluate the brightness values. This can be achieved, for example, by determining and evaluating brightness values ​​only for static pixels of the provided camera image, with the luminaire pixels being included as a subset of the static pixels.

[0030] While the vehicle is moving, all non-static, i.e. dynamic pixels, which represent the continuously changing environment around the vehicle, change. By detecting these dynamic pixels, the static pixels in the camera image can be clearly identified, which in turn represent objects that are fixed to the camera and therefore fixed to the vehicle. By detecting the static pixels, the area in which the taillights are located in the camera image can be narrowed down, since the taillights are fixed to the vehicle and the light pixels are therefore part or a subset of these static pixels. The contours of the vehicle visible in the camera image, including the taillights, can therefore be identified (in advance) while the vehicle is moving through this moving image comparison or this differentiation between dynamic and static pixels.

[0031] The identification of the rear lights in the camera image can be further refined using a histogram, so that it is possible to determine and evaluate brightness values ​​only for the light pixels of the provided camera image, wherein the light pixels are identified by temporally evaluating the brightness values ​​of the pixels, in particular only the static pixels, of the provided camera image upon activation and subsequent deactivation of the respective rear light, wherein the pixels whose brightness values ​​change over time due to the activation and subsequent deactivation of the respective rear light are identified as light pixels.

[0032] To more precisely pinpoint the position of the taillights in the camera image, an initial calibration in a fully functional state can be performed by controlled activation and deactivation of the taillights while observing the brightness values. This allows the light pixels in the camera image to be located easily and reliably, and the subsequent evaluation of the brightness values ​​during vehicle operation can be limited to these light pixels. This reduces the analysis effort and makes detection more stable, as the histogram contains no "background" caused by brightness values ​​not assigned to the taillights, or this "background" can be minimized.

[0033] In a further embodiment of the method, the light monitoring is carried out for a motor vehicle rear light on a motor vehicle of the vehicle and / or for a trailer rear light on a trailer of the vehicle, wherein the respective rear light has at least one lighting unit selected from the group consisting of: reflector, brake light, tail light, rear fog light, license plate light, indicator, clearance light.

[0034] This allows the functional status of the taillights or taillights to be determined and displayed individually, and if necessary, for each lamp unit, allowing individual taillights or lamp units to be replaced, repaired, and / or cleaned. The process is therefore not limited to a specific vehicle type.

[0035] In a further embodiment of the method, a heating element on the respective rear light and / or a cleaning system of the respective rear light is activated depending on the determined and evaluated brightness values ​​at least for the light pixels of the respective camera image and / or depending on the determined and output functional status of the at least one rear light.

[0036] New LED lights can be equipped with heating elements to melt snow cover and thus ensure the visibility of the lights during snowfall. It is advantageous to switch these heating elements on only when needed to save energy, which in this case can be advantageously done in coordination with the previously determined functional status. A cleaning system, particularly for the camera or rear lights, such as a wiper or spray device, can be used to remove dirt and / or snow from the camera or rear lights. Therefore, if the functional status initially indicates that a rear light is defective or covered, cleaning can be used to attempt to restore a reliable camera image or enable the rear lights to illuminate.

[0037] The procedure can then be repeated after the heating element and / or cleaning system have been activated, and a functional status can be checked again to indicate a definitive defect. This prevents an incorrect functional status of the taillights from being output due to dirt or snow covering the taillights or the camera.

[0038] In one embodiment of the monitoring system, it is provided that the at least one rear-facing camera has a detection range with an aperture angle of > 180°, in particular is a fisheye camera, so that the camera can directly detect the at least one rear light and the area surrounding the light. This makes it possible to detect the rear lights with just one camera on the rear of the respective vehicle or vehicle part. Such cameras can also be used efficiently in reversing assistance, as they detect a large part of the rear space behind the vehicle or the respective vehicle part. The vehicle's rear lights are then normally also located in this detection range. However, it can also be provided that the camera can only detect the area surrounding the lights, for example if a fisheye camera is not yet used or the field of view is obscured.In this case, luminaire monitoring can also be carried out as described by only recording the luminaire environment.

[0039] The invention is explained in more detail below with reference to the accompanying drawings. They show:

[0040] Fig. 1 is a schematic plan view of a vehicle with a camera system;

[0041] Fig. 2 shows a histogram with an exemplary distribution of brightness values ​​of a rear light detected by the camera system;

[0042] Fig. 3a a camera image of a camera of the camera system;

[0043] Fig. 3b a camera image of the camera, showing only the dynamic pixels;

[0044] Fig. 3c a camera image of the camera, showing only the static pixels; and

[0045] Fig. 4 is a flow chart of the method according to the invention.

[0046] Figure 1 shows a schematic top view of a trailer 2b or semi-trailer as part of a vehicle 1, which, as indicated, also has a motor vehicle 2a as a towing vehicle. The vehicle 1 has a camera system 4 with cameras 8, in particular with a rearward-facing trailer camera 8b arranged on the trailer 2b, which can, for example, be part of a reversing assistance system 3, and a control unit 6. The control unit 6 can also be arranged in the motor vehicle 2a.

[0047] The trailer camera 8b is equipped with a fisheye lens, i.e., it is designed as a fisheye camera 9 (also known as a fisheye camera), and therefore has a detection range E with an aperture angle of > 180°. The towing vehicle 2a can also have a rearward-facing towing vehicle camera 8a, designed as a fisheye camera 9, as part of the camera system 6 and which can be part of the reversing assistance system 3. In this way, the towing vehicle 2a can be assisted in approaching the trailer 2b, e.g., during a coupling process.

[0048] The control unit 6 is configured to carry out a method for monitoring a rear light 5 of the vehicle 1, in particular trailer rear lights 5b on the trailer 2b. In the illustrated embodiment, each trailer rear light 5b consists of three lighting units 7a, 7b, 7c, which, when the respective rear light 5 is activated, emit light into a lighting environment U5, in particular onto a surface, a building, the vehicle's own vehicle 1, other vehicles, etc. However, motor vehicle rear lights 5a on the motor vehicle 2a can also be monitored using the method, in particular when no trailer 2b is attached. Furthermore, the method can also be used to monitor rear lights 5 of a one-piece vehicle 1, which then only has the motor vehicle 2a, for example a panel van, box truck, flatbed truck, etc.

[0049] To perform such monitoring, the control unit 6 is connected via a corresponding interface 6a to a bus system 10 of the vehicle 1, for example, a CAN bus, in order to first determine whether the respective rear light 5 of the vehicle 1 is currently activated or not. A corresponding message or activation signal S5 containing information about the activation or deactivation of the rear light 5 is transmitted via the bus system 10. Monitoring is only meaningful if a respective rear light 5 is activated.

[0050] With a monitoring system 20, which is formed by the control unit 6 and the camera system 4 or the cameras 8; 8a, 8b, the method for monitoring the rear light(s) 5; 5a, 5b of the vehicle 1 can be carried out, for example, in the following steps shown in Fig. 4: Detecting at least one rear light 5; 5a, 5b of the vehicle 1 and / or its lighting surroundings U5 by the camera 8; 8a, 8b on the vehicle 1 and outputting camera signals S8; S8a, S8b (ST1);

[0051] Providing a camera image B; Ba, Bb with pixels P or image points depending on the output camera signals S8; S8a, S8b, wherein the at least one rear light 5; 5a, 5b of the vehicle 1 and / or the lighting environment U5 is represented in lighting pixels PL of the respective camera image B; Ba, Bb (ST2);

[0052] Determining and evaluating brightness values ​​HW at least for the luminaire pixels PL of the respective camera image B; Ba, Bb (ST3) and determining and outputting a functional status F of the at least one rear light 5; 5a, 5b depending on the determined and evaluated brightness values ​​HW, in particular depending on the brightness values ​​HW that are assigned to the respective rear light 5; 5a, 5b displayed and / or the luminaire environment U5 in the respective camera image B; Ba, Bb (ST4).

[0053] As indicated by the dotted semicircle (detection area E) in Fig. 1, the respective camera 8; 8a, 8b with a fisheye lens can capture all rear lights 5; 5a, 5b of the two-part vehicle 1 in this case, as well as the lighting environment U5 into which the light of the respective rear light 5; 5a, 5b falls, in the first step ST1. Subsequently, the generated camera signals S8; S8a, S8b and / or the camera image B; Ba, Bb formed therefrom are sent to the control unit 6 in the second step ST2, which then performs a light / dark detection in the provided camera image B; Ba, Bb based on the brightness values ​​HW determined in the third step ST3, particularly for the light pixels PL.

[0054] When processing the respective camera image B; Ba, Bb in this way, it is possible that brightness values ​​HW of pixels P are also determined and evaluated that are not assigned to a rear light 5; 5a, 5b or the lighting environment U5 or cannot be clearly assigned, for example because a clear demarcation of image areas with and without a rear light 5; 5a, 5b or with and without an illuminated lighting environment U5 is not possible or intended. This must be taken into account in the following consideration. During light / dark detection, the determined brightness values ​​HW can, for example, be subjected to a histogram comparison in a first evaluation step ST3.1. As part of the histogram comparison, the distribution of the brightness values ​​HW of the relevant pixels P is displayed in a histogram H, as shown by way of example in Fig. 2. This created histogram H can then be compared with a reference histogram HR (dashed in Fig.2), whereby the reference histogram HR also shows a distribution of brightness values ​​HW of the respective pixels P.

[0055] The reference histogram HR is derived from a historical, previously recorded reference camera image BR, which, for example, depicts fully functional taillights 5; 5a, 5b or the lighting environment U5 illuminated by them. In this exemplary embodiment, the histogram comparison thus includes a comparison of historical brightness values ​​HW in the fully functional state of the taillights 5; 5a, 5b with brightness values ​​HW for the current state of the taillights 5; 5a, 5b.

[0056] In addition, it can be provided that the histogram H is subjected to a brightness adjustment before being compared with the reference histogram HR, whereby the brightness values ​​HW shown in the histogram H are adjusted depending on the current ambient brightness UH. This takes into account that the ambient brightness UH can vary greatly, for example depending on the time of day or the environment, which also has an impact on the histogram H or the brightness values ​​HW shown therein. In order to take this effect into account, the ambient brightness UH is subtracted or the brightness of the histogram H is normalized accordingly in order to provide a histogram H for comparison with the reference histogram HR, independent of the ambient brightness UH.The ambient brightness UH can be determined, for example, by averaging the brightness values ​​HW of all pixels P of the camera image B; Ba, Bb or the brightness values ​​HW of all pixels P except the light pixels PL. If the reference histogram HR then shows that the fully functional rear lights 5; 5a, 5b have a different or higher intensity than the current histogram H, it can be concluded that the rear light(s) 5; 5a, 5b shown in the current camera image B; Ba, Bb or the rear light(s) 5; 5a, 5b illuminating the displayed light environment U5 is / are defective or covered (dirt or snow). Both a defect and a covering cause a change, in particular a reduction, in the respective brightness value HW (compared to the expected brightness value). The fourth step ST4 can then output the function status F accordingly.

[0057] According to a further embodiment, the reference histogram HR, in coordination with the activation signal S5, can also be derived from a historical, previously recorded reference camera image BR in a situation with deactivated rear lights 5; 5a, 5b, and the histogram H after activation of the rear lights 5; 5a, 5b can be derived from the then current camera image B; Ba, Bb. In this exemplary embodiment, the histogram comparison includes a comparison of historical brightness values ​​HW in the deactivated state of the rear lights 5; 5a, 5b with brightness values ​​HW for the current state of the rear lights 5; 5a, 5b in the activated state. This embodiment is then suitable for a direct representation of the rear lights 5; 5a, 5b in the light pixels PL as well as for a representation of the light environment U5 illuminated by the rear lights 5; 5a, 5b.

[0058] In a further embodiment, the functionality of the rear lights 5; 5a, 5b can be determined from the histogram H even without recourse to a reference histogram HR. For this purpose, in a second evaluation step ST3.2, it can be determined, for example, whether the majority of brightness values ​​HW lie in one or more target brightness range(s) HSoll of the histogram H that are to be expected for the respective rear light 5; 5a, 5b shown or used, if necessary also taking into account the ambient brightness UH as described above. For this purpose, for example, a target pixel proportion PSoll (relative or absolute) of pixels P that lie in the target brightness range(s) HSoll can be defined. The target pixel proportion PSoll can, for example, be based on the size of the image section being viewed (number of pixels P) and / or the resolution of the camera image B; Ba, Bb orientate and / or what type (gravel, asphalt, buildings, own vehicle 1 , etc.) of the luminaire environment U5 is illuminated.

[0059] Furthermore, each rear light 5; 5a, 5b is / are assigned one or more target brightness range(s) HSoll depending on the type and extent of the lighting units 7a, 7b, 7c located therein, since, for example, a rear light 5; 5a, 5b that glows purely red causes a different distribution of the brightness values ​​HW in the histogram H than a rear light 5; 5a, 5b that glows purely orange or a rear light 5; 5a, 5b that glows purely white or a rear light 5; 5a, 5b with lighting units 7a, 7b, 7c that glow in different colors.

[0060] The functional status F of the respective rear light 5; 5a, 5b can then be determined in the fourth step ST4 from whether the target pixel proportion PSoll for the respective applicable target brightness range HSoll is reached (or exceeded) or not. If rear lights 5; 5a, 5b with differently colored light units 7a, 7b, 7c are located next to each other, several target brightness ranges HSoll (separately for each color) can also be subjected to such an evaluation simultaneously, and thus the functional status F for different light units 7a, 7b, 7c can be determined in parallel.

[0061] Therefore, by means of the camera system 4, which is already present in the vehicle 1 as part of the reversing assistance system 3, light monitoring can also be carried out by means of a corresponding extension in the signal evaluation in the control unit 6.

[0062] According to a further embodiment, it can be provided that for the evaluation of the brightness values ​​HW, a feature representation M is used instead of a histogram H, in which not only the brightness is considered as a dimension, but also, for example, a color channel (RGB), a saturation, a contrast, etc. Therefore, a feature representation M is used in a higher-dimensional feature space, which enables a more detailed evaluation. The luminaire monitoring can then be carried out in a comparable manner as for the histogram H by comparing the feature representation M with a reference feature representation MR, for example using machine learning tools. The reference feature representation MR is generated for each additional feature of the feature space under the same conditions as described for the reference histogram.

[0063] In addition, in a calibration step ST2.1, the position of the luminaire pixels PL in the camera image B; Ba, Bb can be calibrated in advance, which further improves the reliability of the method and makes the detection more stable, since for the creation of the histogram H or the feature representation M and, if applicable, the reference histogram HR or the reference feature representation MR, mainly brightness values ​​HW or features from the luminaire pixels PL in the camera image B; Ba, Bb or in the reference camera image BR can be used.

[0064] Fig. 3a shows a black-and-white camera image Bb recorded by the trailer camera 8b, which consists of several pixels P. This camera image Bb from the trailer camera 8b also shows the fixed position of the trailer taillights 5b or the associated light pixels PL. The position of these light pixels PL does not change, since both the trailer camera 8b and the trailer taillights 5b are permanently connected to the vehicle 1 or the trailer 2b, respectively, and thus their relative position to each other is fixed.

[0065] This can be used to calibrate the position of the light pixels PL. For this purpose, the camera image Bb of the trailer camera 8b is divided into areas with dynamic pixels PD (Fig. 3b, hatched) and areas with static pixels PS (Fig. 3c, hatched), as shown in Figures 3b and 3c. It is assumed that the dynamic pixels PD change as the vehicle 1 travels due to the changing environment, while the trailer 2b is shown consistently in the static pixels PS. Accordingly, the trailer rear lights 5b and the light pixels PL assigned to them fall as a subset into the area of ​​the static pixels PS, whereby the static pixels PS can be recognized by appropriate image processing. The number of pixels P that are taken from the camera image Bb of the trailer 2b to create the histogram H orThe error rate can be significantly reduced by using only the static pixels PS for the feature representation M. The same applies to the creation of the reference histogram HR or the reference feature representation MR in the first evaluation step ST3.1, whereby only the static pixels PS in the reference camera image BR are used for this purpose.

[0066] This can be further refined by, for example, creating histograms H or feature representations M for these static pixels PS during activation and subsequent deactivation of the trailer rear lights 5b during the initial installation and evaluating these histograms over time. From a change in the brightness values ​​HW or features during such an initial activation / deactivation, it can be determined in advance for each individual static pixel PS whether it is assigned to a trailer rear light 5b or not, or whether this static pixel PS is a light pixel PL or not. This knowledge can then be used for the subsequent monitoring of the trailer rear lights 5b by only using the light pixels PL that have already been identified to create the current histogram H or the current feature representation M. The histogram H orThe feature representation M, which is created after the initial activation of the trailer tail lights 5b, can then also be saved simultaneously as a reference histogram HR or reference feature representation MR.

[0067] If such a refined selection of static pixels PS cannot be made, then at least when evaluating the histogram H or the feature representation M in the second evaluation step ST3.2, a type of "background" (components next to the rear lights 5; 5a, 5b) must be taken into account, which leads to a correspondingly modified distribution of the brightness values ​​HW or features. In the first evaluation step ST3.1, however, this background can also be found in the reference histogram HR or in the reference feature representation MR, so that it is already taken into account in the histogram comparison or feature representation comparison. If the respective camera 8; 8a, 8b is further designed in such a way that it can only reliably capture the light environment U5, i.e. the environment irradiated by the respective rear light 5; 5a, 5b or lighting unit 7a, 7b, 7c, and the respective rear light 5; 5a, 5b orIf the lighting unit 7a, 7b, 7c itself is not available or not fully available, such a refined selection of static pixels PS can also only be made to a limited extent. In this case, the histogram comparison or feature representation comparison between activated and deactivated rear lights 5; 5a, 5b or lighting units 7a, 7b, 7c described above, or the evaluation of the target pixel proportions PSoll for the respective applicable target brightness range HSoll, must be used.

[0068] The described selection of pixels P by subdividing them into static and dynamic pixels PS, PD is also carried out in an identical manner for the motor vehicle rear light 5a or any other rear light 5 on the vehicle 1.

[0069] Depending on the determined functional status F, a heating element 14 can then be switched on, for example, depending on the situation, which specifically heats the respective rear light 5; 5a, 5b or lighting unit 7a, 7b, 7c, for example, to defrost snow cover, which can lead to reduced brightness values ​​HW. The heating element 14 can therefore only be used in an energy-saving manner if the functional status F indicates a likely snow cover. A cleaning system 15 can also be specifically controlled, for example, to clean dirty rear lights 5; 5a, 5b, which can also lead to reduced brightness values ​​HW.

[0070] After such activation of the heating element 14 and / or the cleaning system 15, monitoring can be performed again in the steps described. If the brightness values ​​HW still indicate a defective or covered rear light 5; 5a, 5b, this can be definitively output as function status F, to which the driver can respond accordingly. List of reference symbols (part of the description)

[0071] 1 vehicle

[0072] 2a Motor vehicle

[0073] 2b trailer

[0074] 3 Reversing assistance

[0075] 4 camera system

[0076] 5a Taillights of the motor vehicle 2a

[0077] 5b Rear lights of trailer 2b

[0078] 6 Control unit

[0079] 6a Interface

[0080] 7a, 7b, 7c lighting unit

[0081] 8 Camera

[0082] 8a Motor vehicle camera

[0083] 8b Trailer Camera

[0084] 9 Fisheye camera

[0085] 10 Bus system

[0086] 14 Heating element

[0087] 15 Cleaning system

[0088] 20 Surveillance system

[0089] B Camera image of camera 8

[0090] Ba camera image from the motor vehicle camera 8a

[0091] Bb camera image of trailer camera 8b

[0092] BR reference camera image

[0093] E Detection range

[0094] F Functional status

[0095] H Histogram

[0096] HR reference H istogram m

[0097] HW brightness value

[0098] HU ambient brightness

[0099] M feature representation

[0100] MR reference feature representation

[0101] P pixels

[0102] PL light pixel S5 activation signal

[0103] S8 camera signal of camera 8

[0104] S8a camera signal of the motor vehicle camera 8a

[0105] S8b Camera signal of trailer camera 8b

[0106] U5 Lighting environment

Claims

Patent claims 1 . Method for monitoring a rear light (5) of a vehicle (1) with a camera system (4), wherein the camera system (4) has a rearward-facing camera (8), wherein the rear light (5) has a lighting unit (7a, 7b, 7c), comprising the following steps: Detecting the rear light (5) of the vehicle (1) and / or a lighting environment (U5) illuminated by the rear light (5) by the camera (8) on the vehicle (1) and outputting camera signals (S8) (ST1); Providing a camera image (B) with pixels (P) depending on the output camera signals (S8), wherein the rear light (5) of the vehicle (1) and / or the lighting environment (U5) illuminated by the rear light (5) is represented in lighting pixels (PL) of the respective camera image (B) (ST2); Determining and evaluating brightness values ​​(HW) for the luminous pixels (PL) of the respective camera image (B) (ST3); and Determining and outputting a functional status (F) of the rear light (5) depending on the determined brightness values ​​(HW) (ST4).

2. Method according to claim 1, characterized in that in order to determine the functional status (F) of the rear light (5) a histogram (H) and / or a pattern representation (M) is created from the provided camera image (B), wherein in the histogram (H) and / or in the pattern representation (M) a distribution of the brightness values ​​(HW) of at least the light pixels (PL) of the respective camera image (B) is shown.

3. Method according to claim 2, characterized in that for determining the functional status (F) of the rear light (5) a reference camera image (BR) is provided and the histogram (H) created from the camera image (B) is compared with a reference histogram (HR) (ST3.1), wherein the reference histogram (HR) is created from the reference camera image (BR), and / or the pattern representation (M) created from the camera image (B) is compared with a reference pattern representation (MR) (ST3.1), wherein the Reference pattern representation (MR) is created from the reference camera image (BR).

4. Method according to claim 3, characterized in that in the reference camera image (BR) the rear lights (5) of the vehicle (1) are shown in a fully functional state and / or in the reference camera image (BR) the lighting environment (U5) is shown in a state irradiated by the rear lights (5) of the vehicle (1) in the fully functional state, wherein the functional status (F) is determined and output (ST4) that the rear light (5) is defective or covered if the histogram (H) created from the camera image (B) deviates from the reference histogram (HR) and / or the pattern representation (M) created from the camera image (B) deviates from the reference pattern representation (MR).

5. The method according to claim 3, characterized in that in the reference camera image (BR) the rear lights (5) of the vehicle (1) are shown in a deactivated state and / or in the reference camera image (BR) the lighting environment (U5) is shown in a state in which the rear lights (5) of the vehicle (1) are in the deactivated state, wherein the functional status (F) is determined and output (ST4) that the rear light (5) is defective or covered if the histogram (H) created from the camera image (B) corresponds to the reference histogram (HR) and / or the pattern representation (M) created from the camera image (B) corresponds to the reference pattern representation (MR).

6. Method according to one of claims 3 to 5, characterized in that the histogram (H) before comparison with the reference histogram (HR) and / or the pattern representation (M) is subjected to a brightness adjustment before being compared with the reference pattern representation (MR), wherein the brightness values ​​(HW) shown in the histogram (H) and / or the pattern representation (M) are adjusted depending on a current ambient brightness (HU).

7. Method according to one of claims 2 to 6, characterized in that in order to determine the functional status (F) of the rear light (5) it is determined whether the brightness values ​​(HW) of the histogram (H) and / or the pattern display (M) lie in at least one target brightness range (HSoll) assigned to the respective rear light (5).

8. Method according to claim 7, characterized in that it is additionally determined whether at least one specified target pixel proportion (PSoll) of brightness values ​​(HW) lies in the at least one target brightness range (HSoll) assigned to the respective rear light (5).

9. Method according to claim 8, characterized in that it is determined and output (ST4) as the functional status (F) that the rear light (5) is defective or covered if the brightness values ​​(HW) of the histogram (H) created from the camera image (B) and / or the pattern representation (M) created from the camera image (B) are not within the at least one assigned target brightness range (HSoll) or are not within the at least one assigned target brightness range (Hsoll) for the specified target pixel proportion (PSoll).

10. Method according to one of claims 7 to 9, characterized in that each lighting unit (7a, 7b, 7c) of the respective rear light (5) is assigned a desired brightness range (Hsoll).

11. Method according to one of the preceding claims, characterized in that the determination and evaluation of brightness values ​​(HW) is carried out only for static pixels (PS) of the provided camera image (B), wherein the luminous pixels (PL) are contained in the static pixels (PS).

12. The method according to claim 11, characterized in that the static pixels (PS) of the provided camera image (B) are identified in advance in a calibration step (ST2.1) during a journey of the vehicle (1).

13. Method according to one of the preceding claims, characterized in that the determination and evaluation of brightness values ​​(HW) is carried out only for the light pixels (PL) of the provided camera image (B), wherein the light pixels (PL) are identified by temporally evaluating the brightness values ​​(HW) of the pixels (P) of the provided camera image (B) upon activation and subsequent deactivation of the respective rear light (5), wherein the pixels (P) whose brightness values ​​(HW) change over time due to the activation and subsequent deactivation of the respective rear light (5) are identified as light pixels (PL).

14. Method according to one of the preceding claims, characterized in that the monitoring of the rear light (5) is carried out for a motor vehicle rear light (5a) on a motor vehicle (2a) of the vehicle (1) and / or a trailer rear light (5b) on a trailer (2b) of the vehicle (1), wherein the respective rear light (5) has a lighting unit (7a, 7b, 7c) which is selected from the group consisting of: reflector, brake light, tail light, rear fog light, license plate light, indicator, clearance light.

15. Method according to one of the preceding claims, characterized in that depending on the determined and evaluated brightness values ​​(HW) at least for the light pixels (PL) of the respective camera image (B) and / or depending on the determined and output functional status (F) of the rear light (5), a heating element (14) on the respective rear light (5) and / or a cleaning system (15) of the respective rear light (5) is activated.

16. The method according to claim 15, characterized in that the method is carried out again after the heating element (14) and / or the cleaning system (15) has been activated.

17. Method according to one of the preceding claims, characterized in that before determining and outputting a functional status (F) of the rear light (5) as a function of the determined brightness values ​​(HW) (ST4), it is checked whether the rear light (5) is activated.

18. A monitoring system (100) for a vehicle (1), comprising: a camera system (4) with a rearward-facing camera (8) for detecting a rear light (5) of the vehicle (1) and / or a lighting environment (U5) illuminated by the rear light (5), wherein the camera (8) is designed to output camera signals (S8), and a control unit (6), in particular for carrying out a method according to one of the preceding claims, wherein the control unit (6) is designed to process a camera image (B) provided as a function of the output camera signals (S8) with pixels (P), wherein the rear light (5) of the vehicle (1) and / or the lighting environment (U5) illuminated by the rear light (5) is represented or depicted in lighting pixels (PL) of the respective camera image (B); To determine and output brightness values ​​(HW) at least for the light pixels (PL) of the respective camera image (B); and to determine and output a functional status (F) of the rear light (5) depending on the determined brightness values ​​(HW).

19. Surveillance system (20) according to claim 18, characterized in that the rearward-facing camera (8) has a detection range (E) such that the at least one rear light (5) and the light surroundings (U5) can be directly detected by the camera (8), in particular in that the rearward-facing camera (8) is designed as a fisheye camera (9), or that only the light surroundings (U5) can be detected.

20. Monitoring system (20) according to claim 18 or 19, characterized in that the camera system (4) is part of a reversing assistance system (3).

21. Monitoring system (20) according to one of claims 18 to 20, characterized in that the control unit (6) has an interface (6a) for receiving an activation signal (S5), wherein the activation signal (S5) transmits whether the rear light (5) is activated or not, and the control unit (6) is designed to determine whether the rear light (5) is activated based on the activation signal (S5) before determining the functional status (F) of the rear light (5).

22. Vehicle (1), in particular commercial vehicle, with a monitoring system (20) according to one of claims 18 to 21, wherein the camera (8) is a motor vehicle camera (8a) on the motor vehicle (2a) of the vehicle (1) and / or a trailer camera (8b) on a trailer (2b) of the vehicle (1), which are each directed backwards.

Images