Determining a luminance map for implementing a motor vehicle function by means of a model

Abstract

The invention relates to a method comprising a preliminary step of storing (204) a model that is capable of determining at least one luminance value according to a colour or grayscale pixel of an image. The method comprises the following consecutive steps: - obtaining (211) a first image comprising first colour or grayscale pixels, the first image being obtained from data from a camera of the vehicle; - obtaining (212), from the stored model and the obtained first image, a first luminance map comprising second pixels, each second pixel being associated with a luminance value; - transmitting (213) the luminance map to a module of the motor vehicle capable of implementing at least one function using at least the first luminance map.

Description

Description Title of the invention: Determination by a model of a luminance map for the implementation of a motor vehicle function

[0001] The present invention relates to the field of functions of a motor vehicle using parameters of the vehicle's environment. It relates in particular to a vehicle system fulfilling at least one function based on light data from the scene facing the vehicle.

[0002] It is particularly advantageous in the case of vehicle functions requiring real-time adaptation.

[0003] It is known to use one or more vehicle sensors to characterize an environment of the vehicle and to implement functions adapting the operation of the vehicle to the environment thus characterized, such as driving assistance functions, adaptive lighting functions or functions for analyzing the scene facing the vehicle, which may indicate the presence of a sign or a pedestrian facing the motor vehicle.

[0004] Several types of sensors can thus be used, alone or in combination, to characterize the vehicle's environment: radar, lidar, camera acquiring images of the scene facing the vehicle, etc.

[0005] RGB cameras, for "red-green-blue", are capable of acquiring color images, which can be processed in real time, and are thus capable of measuring light parameters relating to the vehicle's environment. Such data can thus be used by vehicle functions that take light parameters into account, such as adaptive lighting functions for example.

[0006] However, images from an RGB camera do not allow certain light parameters to be directly reflected, particularly in night driving situations.

[0007] , Cameras capable of acquiring luminance images, or luminance maps, are known but they are on the one hand very expensive, and on the other hand, they require long processing times to acquire a single luminance map, thus making their use incompatible with real-time functions of a motor vehicle.

[0008] There is thus a need to enable the enrichment of light parameter data, or light data, supplied to functions of a motor vehicle, while enabling real-time operation of such functions, and without incurring prohibitive costs.

[0009] . To this end, a first aspect of the invention relates to a method for transmitting light data for at least one function of a motor vehicle, the method comprising a prior step of storing a model in the motor vehicle, the model being capable of determining at least one luminance value as a function of at least one color or grayscale pixel of an image. The method further comprises the following common steps: - obtaining at least a first image comprising first pixels in color or in grayscale, the first image being obtained from data from a camera of the vehicle; - obtaining, from the stored model and the first acquired image, a first luminance map comprising second pixels, each second pixel being associated with a luminance value; - transmission of the luminance map to a module of the motor vehicle capable of implementing at least one function from at least said first luminance map.

[0010] . Thus, it is possible to determine a first luminance map from a first image from a color or grayscale camera. The use of such a model makes it possible to obtain a map real-time luminance, the acquisition of a grayscale or color image being generally fast, and based on cameras much less expensive than luminance cameras. The first luminance map can thus be used by any vehicle function, including functions that must operate in real time.

[0011] . According to embodiments, the stored model may be capable of determining from a color or grayscale pixel, a corresponding luminance value, and obtaining the luminance map may comprise determining, for each first pixel, a luminance value given by the model, and associating the given luminance value with a second pixel of the luminance map corresponding to the first pixel.

[0012] , Thus, a simple model to design and execute is used, the model determining a single output value. The model can for example be a simple function associating an output value with an input value. The luminance map can thus be obtained by successively applying the model to all the first pixels, in order to determine the luminance value of each of the corresponding second pixels.

[0013] . Alternatively, the stored model may be capable of determining a luminance map directly from an image comprising grayscale or color pixels, received as input to the model, and obtaining the first luminance map may comprise the determination by the stored model of the first luminance map, directly from the first image received as input to the stored model.

[0014] , Thus, the model is able to take an entire image as input. The accuracy of the model can thus be improved, in that each luminance value of a second pixel can depend on several values ​​of first pixels of the first image. The consistency of the first luminance map obtained can in particular be improved.

[0015] . According to embodiments, the method may comprise, during a prior phase, obtaining a training data set, training a model by machine learning from the training data set so that the model is capable of determining at least one luminance value as a function of at least one color or grayscale pixel of an image, and storing said model in the vehicle.

[0016] . Thus, the model can be derived from machine learning based on a training database. This makes it possible to obtain an accurate model trained on real or simulated training data.

[0017] , Additionally, the training dataset may include associations between grayscale or color training images and training luminance maps, and the training of said model may be training by supervised learning.

[0018] . It is thus made possible to obtain an accurate model to determine the first luminance map.

[0019] . According to embodiments, the module may be capable of implementing at least one of the following functions: - an object detection function in said at least one first image; - a segmentation function of said at least one first image; - a function for adapting the lighting of the motor vehicle; and / or - a function to alert or inform the driver of the motor vehicle.

[0020] . Thus, the first luminance map can be used for various functions of the motor vehicle.

[0021] , According to embodiments, the method may further comprise transmitting said at least one first image to the module, said module being capable of implementing at least one function from said first luminance map and from said at least one first image.

[0022] , Thus, the function can be implemented based on complementary light data, including the first image and the first corresponding luminance map.

[0023] . According to embodiments, the method may further comprise, after storing the model in the motor vehicle and before obtaining the first image, obtaining calibration images from the vehicle camera and calibrating the model based on said calibration images.

[0024] , Thus, the trained model can be a generic model, which can then be calibrated according to a type of camera with which the motor vehicle is equipped.

[0025] A second aspect of the invention relates to a computer program comprising instructions for implementing the method according to the first aspect of the invention, when these instructions are executed by a processor.

[0026] A third aspect of the invention relates to a device for transmitting light data for at least one function of a motor vehicle, the device comprising a memory storing a model or being able to access said model stored in the vehicle, the model being able to determine at least one luminance value as a function of at least one pixel in color or gray levels of an image, the device further comprising: - a first interface capable of obtaining at least a first image comprising first pixels in color or in grayscale, the first image being obtained from data from a camera of the vehicle; - a processor capable of obtaining, from the stored model and the first acquired image, a first luminance map comprising second pixels, each second pixel being associated with a luminance value; - a second interface capable of transmitting the luminance map to a module of the motor vehicle capable of implementing at least one function from at least said first luminance map.

[0027] , Other characteristics and advantages of the invention will appear on examining the detailed description below, and the appended drawings in which:

[0028] . [Fig 1] illustrates a motor vehicle according to embodiments of the invention;

[0029] . [Fig 2] illustrates the steps of a method according to embodiments of the invention;

[0030] . [Fig 3] illustrates a device for transmitting light data according to embodiments of the invention;

[0031] The description focuses on the characteristics that distinguish the method and device from those known in the state of the art.

[0032] , Figure 1 illustrates a vehicle 100 according to embodiments of the invention.

[0033] The vehicle 100 comprises a device 101 according to embodiments of the invention, the device 101 being capable of determining a luminance map for at least one image received from a camera 102 of the vehicle, by means of a model that the device 101 stores or to which the device 101 accesses, when the model is stored in an entity of the vehicle 100 other than the device 101.

[0034] The camera 102 is capable of acquiring images, or frames, of the environment of the vehicle 100, preferably of the scene of the vehicle located in front of the vehicle 100, that is to say in front of the vehicle 100 in a direction of travel of the vehicle 100. In particular, the camera 102 may be located at the front of the vehicle, for example at the top of the windshield of the vehicle 100 or at the bumper of the vehicle 100.

[0035] . The images acquired by the camera 102 may be grayscale images, that is to say that each image comprises first pixels, arranged in a matrix, each first pixel being coded on strictly more than two levels, with at least one intermediate grayscale between white and black. Each first pixel may thus be coded on several bits, for example on a byte coding 256 grayscale levels.

[0036] . Alternatively, the images acquired by the camera 102 are color images, that is to say that each image comprises first pixels arranged in a matrix, and that each first pixel is coded so as to correspond to one color among strictly more than two colors. For example, each first pixel can be coded in red-green-blue, or RGB for “Red Green Blue”, that is to say that the set of colors is represented by three coordinates respectively red, green and blue. Each first pixel can thus be coded on several bits, for example on a byte coding 256 colors.

[0037] The camera 102 may be capable of acquiring a still image regularly, at a given frequency, for example every second. However, preferably in order to allow the use of data from the camera by real-time functions of the vehicle, the camera 102 may be capable of obtaining images constituting video frames at a frequency of several frames per second, or fps, for “frame per second” in English, for example at a frequency greater than 10 fps, in particular equal to 30 fps.

[0038] According to the invention, the device 101 is capable of receiving a first image from the camera 102 and of determining a first luminance map by applying the model that it stores or which it accesses.

[0039] Luminance is a quantity that designates the luminous flux coming from an illuminated surface and which is reflected in the eye, which is expressed in candela per square meter, or cd / m 2Luminance thus translates a visual sensation of brightness of a surface.

[0040] More precisely, luminance is the power of visible light passing through or being emitted by a surface element in a given direction, per unit area and per unit solid angle.

[0041] In outdoor lighting, the measurement of luminance is normalized according to the distance from the surface, the position and height of the observer, the angle of observation, etc.

[0042] , Thus, the luminance map is distinct from a grayscale image, in which the grayscale of a pixel illustrates the brightness perceived by the camera sensor 102 at a given location, but which does not correspond to luminance as defined above.

[0043] The device 101 is thus capable of transforming at least one grayscale or color image into a luminance map comprising second pixels, each second pixel being associated with a luminance value.

[0044] , Device 101 is particularly capable of: - obtaining a first luminance map for each image captured by the camera, each image captured by the camera thus being a first image processed by the model according to the invention. Such an embodiment is advantageous when the camera 102 captures photos at a low frequency, for example less than a few fps, for example less than 5 fps; - obtaining a first luminance map for an image among several images captured by the camera 102, called the first image. For example, one image per second can be used as the first image to obtain a luminance map when the acquisition frequency of the camera 102 is greater than several fps. Such an embodiment is advantageous when the acquisition frequency of the camera is greater than 10 fps, for example equal to 30 fps; - first obtain a luminance map from an average image obtained from several images captured by the camera 102, the average image being the first image described with reference to FIG. 2. The average image can for example be obtained by averaging the images captured by the camera 102 over a given duration, for example for one second, when the acquisition frequency is greater than several fps, for example greater than 10 fps, in particular equal to 3 fps.

[0045] The luminance map thus determined can be used as light data by one or more functions of the vehicle 100, such as: - a detection function implemented by a detection module 104 of the vehicle 100, capable of detecting an element of the scene represented in the images acquired by the camera 102; - a function for segmenting the black and white or color images acquired by the camera 102, implemented by a segmentation module 105 of the vehicle 100; - a driver alert or information function implemented by a human-machine interface, HMI, 103. The alert function can take as inputs outputs from the detection function implemented by the detection module 104; - a vehicle lighting adaptation function implemented by at least one lighting module 107 of a pair of vehicle lighting modules 100. In particular, each lighting module may comprise a matrix light source, which comprises individual light elements that can be individually controlled. Depending on objects detected in the scene, the lighting may be adapted by modifying the light intensity of one or more of the individual light elements. The individual light elements may be electroluminescent elements, such as LEDs in particular, the light intensities of which can be controlled by the electrical currents that are respectively applied to them.

[0046] The detection module 104 is thus able to identify one or more objects in the scene captured by the camera 102, on the basis at least of the first luminance map received from the device 101. Advantageously, the detection module 104 can receive both the first luminance map obtained and the first image from the camera data, or all of the images captured by the camera. 102. In a complementary manner, the detection module can also receive data from other sensors of the vehicle 100, not shown in FIG. 1.

[0047] , No restriction is attached to the HMI 103, which comprises any element capable of receiving a command from the user, whether this command relates to the piloting of the vehicle or to the control of interior or exterior equipment, and / or to transmit information to the user. For this purpose, the HMI 103 may comprise a screen, such as a touch screen, a set of one or more buttons, a speaker, a microphone, a dashboard capable of displaying one and / or more luminous pictograms of predefined shapes, a steering wheel vibration system, etc.

[0048] . Preferably, the HMI 103 is capable of transmitting visual and / or audio data to the driver of the vehicle. The HMI 103 can in particular indicate objects detected by the detection module 104.

[0049] The segmentation module 105 is capable of segmenting the images captured by the camera 102, or some of these images. Such segmentation may consist of dividing the scene represented by each image into several zones, each zone being associated with a distance interval relative to the vehicle 100. For example, the segmentation module 105 may be capable of identifying a first zone corresponding to elements of the scene between 0 and 10 meters from the vehicle 100, a second zone corresponding to elements of the scene between 10 and 100 meters, and a third zone corresponding to elements of the scene beyond 100 meters from the vehicle 100.

[0050] . Each zone corresponds to a set of pixels which can be identified in an image captured in an image perceived by the camera 102 or in a determined luminance map. Thus, from the segmentation determined by the segmentation module 105, it is possible to determine at least one luminance value in each area resulting from the segmentation, for example an average luminance in each area resulting from the segmentation, from the luminance map obtained by the device 101.

[0051] . According to embodiments of the invention, the segmentation module 105 can segment the scene according to the luminance map from the device 101 and / or according to at least one image perceived by the camera 102.

[0052] The vehicle 100 may further comprise a regulatory constraints verification module 106, capable of comparing luminance values ​​with regulatory thresholds. In the event of a deviation of at least one luminance value in at least one given zone of the captured images compared to regulatory values, the module 106 may transmit a lighting modification command to the lighting module 107, in order to modify the light intensity of at least one light element.

[0053] The module 106 can receive as input the luminance map, but also the segmentation of a captured image from the module 105, in order to compare the luminance values, or average luminance values, in areas of the scene resulting from the segmentation, with regulatory values.

[0054] , Modules 104, 105, 106, 107 have been shown in FIG. 1. However, the aforementioned modules, or some of these modules, may be grouped into a single module.

[0055] . Figure 2 illustrates the steps of a method for transmitting light data according to embodiments of the invention.

[0056] The method may comprise a preliminary phase 200, during which the model according to the invention is obtained then stored in the device 101, and optionally calibrated. Thus, some of the steps of the preliminary phase 200 may be implemented outside the device 101.

[0057] , At a step 201, a training database can be created.

[0058] The training database may comprise grayscale or color training images from one camera, or from several separate cameras. In particular, the training images in the training database may come from different cameras, in particular from cameras of different types, two different types having at least one technical characteristic that differs or having different manufacturers. The creation of a database with training images from cameras of different types allows the training of a generic model, which can then be integrated into any vehicle, regardless of the camera 102 of the vehicle 100.

[0059] . Alternatively, all images are from cameras of the same type as camera 102, in which case the model trained from the training data is specific to the camera 102 previously described.

[0060] . Preferably, the training data of the base constituted in step 201 comprise associations between the aforementioned training images and respective training luminance maps. Each training luminance map associated with a training image can be acquired by a luminance camera, placed next to the camera acquiring the training image, so as to acquire visual data relating to the same scene as that of the camera.

[0061] . At a step 202, a model can be trained based on the training data of the base trained at step 201. No restriction is attached to the model resulting from the training of step 202 which can be: - a curve or function that maps a value of a pixel in a grayscale or color image to a luminance value of a corresponding pixel in a luminance map. Such a curve can be obtained by a curve fitting analysis technique, or "curve fitting" in English, from the associations stored in the training database. Indeed, from these associations in the training database, each association associating a training image with a training luminance map, pixel associations can be determined between a given pixel in the training image and a corresponding pixel in the training luminance map. These associations between pixels feed the curve fitting analysis to obtain a model in the form of a curve; - a support vector machine, or SVM model, for "Support Vector Machine" in English, capable of determining, by classification, a luminance value of a pixel in a luminance map, from a value of a corresponding pixel in a grayscale or color image. An SVM model can be built by supervised learning from the associations of the training database. Indeed, from these associations of the training database, each association associating a training image with a training luminance map, pixel associations can be determined between a given pixel of the training image and a corresponding pixel in the training luminance map. These associations between pixels feed the supervised learning of the SVM model. The advantage of an SVM model is that it can be trained from a training database with few data; - a neural network, for example a convolutional neural network, capable of determining an entire luminance map from a grayscale or color image received as input. The neural network can be obtained by machine learning from the data in the training database constituted in step 201. In the case of supervised learning, the neural network can be configured by modifying parameters of certain neurons, as a function of a measured difference between the luminance map determined by the neural network when a given training image is submitted to it and the training luminance map associated in the training database with the given training image; - any other model resulting from machine learning or by a given analysis technique, and capable of determining a luminance value from a value of a pixel in grayscale or in color, or capable of determining an entire luminance map from a grayscale or color image received as input

[0062] , Step 202 can be implemented by a training module not shown in FIG. 1, which can be external to the vehicle 100, and which is able to access the training database.

[0063] . In a step 203, at the end of the training of step 202, which is an analysis or machine learning as detailed previously, a model according to the invention is obtained, the model being capable of determining a luminance value from a value of a pixel in grayscale or in color, or capable of determining an entire luminance map from a grayscale or color image received as input.

[0064] , In a step 204, the model obtained in step 203 is stored in the motor vehicle 100, either in a memory of the device 101, or in a memory external to the device 101 but accessible to the device 101.

[0065] . In a step 205, the device 101 calibrates the stored model based on calibration images received from the camera 102 of the vehicle. Step 205 is optional. It is notably not necessary when the training database comprises training images from a type of camera identical to the camera 102 of the motor vehicle 100. Step 205 is however preferable, but not necessary, when a generic model is obtained from varied training data, from cameras of several types.

[0066] . Preferably, the calibration images received from the camera 102 are images which vary according to the situations represented and / or according to the colors or grayscales that they contain, so that most situations can be represented by interpolation of these calibration images. Calibrating the model on the basis of these calibration images thus makes it possible to improve the accuracy associated with the calibrated model, which is then a model adapted to the type of camera 102 with which the vehicle 100 is equipped. In the event of a change of camera of the vehicle, or a change of certain parameters of the camera, the calibration step 205 can be repeated in order to adapt the model to the change made.

[0067] The method according to the invention further comprises a current phase 210. By current phase is meant a phase comprising steps implemented while the vehicle is in working order, for example when it is traveling on a road, or stationary but in contact. The current phase 210 may comprise the steps described below.

[0068] . At a step 211, the device 101 obtains a first image from the camera 102. As indicated previously, the first image can be: - each image captured by the camera 102. In this case, the method is iterated for each image acquired by the camera. The image most recently acquired by the camera 102 is then considered as the first image of steps 211 to 213; - one image from among N images captured by the camera 102, N being an integer greater than or equal to 2. In this case, the method is iterated for each set of N images acquired by the camera 102. Thus, for each set of N images from the camera 102, one of the N images is selected as the first image of steps 211 to 213; - an average image calculated from N images captured by the camera 102, N being an integer greater than or equal to 2. In this case, the method is iterated for each set of N images acquired by the camera 102. Thus, for each set of N images from the camera 102, the average image is determined by the device 101 or by the camera 102, then used by the device 101 as the first image of steps 211 to 213.

[0069] The first image consists of a set of first pixels, each pixel being associated with a value encoding its gray level or color. There are no restrictions on the format of the first image or the number of pixels it contains.

[0070] . In a step 212, the device 101 obtains a first luminance map, from the stored model and the first image obtained, of a first luminance map comprising second pixels, each second pixel being associated with a luminance value. The adjective “second” does not indicate an order, but refers to a pixel belonging to a luminance map, as opposed to a pixel of a grayscale or color image, which is called the first pixel.

[0071] , Each first pixel is thus associated with a value coding a color or a gray level, while each second pixel is associated with a luminance value.

[0072] , As previously indicated, the first luminance map can be determined by the device 101 in step 212: - directly by the model when the model is capable of receiving a first entire image in color or in grayscale, and of determining a first entire luminance map; - second pixel by second pixel, when the model is capable of determining a luminance value of a second pixel of the first luminance map from a value of a corresponding first pixel of the first image.

[0073] . In a step 213, the device 101 transmits at least the first luminance map obtained to one of the modules 103, 104, 105 and 107 previously described, for the implementation of at least one function using the first luminance map. In addition, the device 101 can transmit to the module the first image in addition to the first luminance map. The current steps 211 to 213 are iterated on acquisition of new images by the camera 102, as indicated previously.

[0074] , It should be noted that according to embodiments, the number of second pixels of the first luminance map is equal to the number of first pixels of the first image. Alternatively, the first luminance map may comprise fewer pixels, for example N times fewer pixels, N being an integer greater than or equal to 2, and each luminance value of a second pixel is determined from the value of N first pixels of the first image.

[0075] . Figure 3 shows the structure of the device 101 according to embodiments of the invention.

[0076] . The device 101 comprises a processor 301 configured to communicate unidirectionally or bidirectionally, via one or more buses or via a direct wired connection, with a memory 302 such as a “Random Access Memory” type memory, RAM, or a “Read Only Memory” type memory, ROM, or any other type of memory (Flash, EEPROM, etc.). Alternatively, the memory 302 comprises several memories of the aforementioned types.

[0077] , The memory 302 is capable of storing, permanently or temporarily, at least some of the data used and / or resulting from the implementation of the steps of the method according to the invention illustrated with reference to figure 2.

[0078] In particular, the memory 302 may be capable of storing the aforementioned model, during step 204, and the calibration images during the optional step 205.

[0079] The processor 301 is capable of executing instructions, stored in the memory 302, for the implementation of steps 211 to 213, and optionally of step 205, of the method according to the invention, described with reference to FIG. 2. Alternatively, the processor 301 can be replaced by a microcontroller designed and configured to carry out steps 211 to 213, and optionally step 205, of the method according to the invention, described with reference to FIG. 2.

[0080] The device 101 comprises a first interface 303 capable of receiving grayscale or color images from the camera 102.

[0081] The device 101 further comprises a second interface 304 capable of transmitting light data to one or more of the modules 103, 104, 105 and 107 previously described. The second interface 304 may in particular be capable of transmitting the first luminance map, and optionally the first image in color or in grayscale.

[0082] The present invention is not limited to the embodiments described above as examples; it extends to other variants.

Claims

Claims

1. Method for transmitting light data for at least one function of a motor vehicle (100), the method comprising a prior step of storing (204) a model in the motor vehicle, the model being capable of determining at least one luminance value as a function of at least one pixel in color or grayscale of an image, the method further comprising the following common steps: - obtaining (211) a first image comprising first pixels in color or in grayscale, the first image being obtained from data from a camera (102) of the vehicle; - obtaining (212), from the stored model and the first image obtained, a first luminance map comprising second pixels, each second pixel being associated with a luminance value; - transmission (213) of the first luminance map to a module (104; 105; 106; 107) of the motor vehicle capable of implementing at least one function from at least said first luminance map.

2. Method according to claim 1, in which the stored model is capable of determining, from a pixel in color or in gray level, a corresponding luminance value, and in which, obtaining (212) the first luminance map comprises determining, for each first pixel, a luminance value given by the model, and associating the given luminance value with a second pixel of the luminance map corresponding to the first pixel.

3. The method of claim 1, wherein the stored model is capable of determining a luminance map directly from an image comprising grayscale or color pixels received as input to the model, and wherein obtaining (212) the first luminance map comprises determining by the stored model the first luminance map directly from the first image received as input to the stored model.

4. Method according to one of the preceding claims, comprising, during a prior phase (200), obtaining (201) a training data set, training (202) a model by machine learning from the training data set so that the model is capable of determining at least one luminance value as a function of at least one color or grayscale pixel of an image, and storing (204) said model in the vehicle.

5. A method according to claim 4 and claim 3, wherein the training dataset comprises associations between grayscale or color training images and training luminance maps, and wherein the training (202) of said model is training by supervised learning.

6. Method according to one of the preceding claims, in which the module (104; 105; 106; 107) is capable of implementing at least one of the following functions: - an object detection function in said at least one first image; - a segmentation function of said at least one first image; - a function for adapting the lighting of the motor vehicle; and / or - a function to alert or inform the driver of the motor vehicle.

7. Method according to one of the preceding claims, further comprising the transmission (213) of said at least one first image to the module (104; 105; 106; 107), said module being capable of implementing at least one function from said first luminance map and from said at least one first image.

8. Claim according to one of the preceding claims, further comprising, after storing (204) the model in the motor vehicle and before obtaining (211) the first image, obtaining calibration images from the camera (102) of the vehicle and calibrating (205) the model according to said calibration images.

9. Computer program comprising instructions for implementing the method according to one of the preceding claims, when these instructions are executed by a processor (301).

10. Device (101) for transmitting light data for at least one function of a motor vehicle (100), the device comprising a memory (302) storing a model or being able to access said model stored in the vehicle, the model being able to determine at least one luminance value as a function of at least one pixel in color or gray levels of an image, the device further comprising: - a first interface (303) capable of obtaining at least a first image comprising first pixels in color or in grayscale, the first image being obtained from data from a camera of the vehicle; - a processor (301) capable of obtaining, from the stored model and the first image obtained, a first luminance map comprising second pixels, each second pixel being associated with a luminance value; - a second interface (304) capable of transmitting the first luminance map to a module (104; 105; 106; 107) of the motor vehicle capable of implementing at least one function from at least said first luminance map.

Images