Vehicle light beam compatibility inspection

Abstract

The present invention relates to a method for evaluating the suitability of a light beam of a lighting module of an automobile, the method comprising the step of storing a model in the automobile (203). The model is capable of determining information regarding the suitability of the light beam based on an image received as input to the model. The method includes confirming that the lighting module of the automobile is activated (221), acquiring a first image based on an item of data from an image acquisition device of the automobile (223), applying the model to the first image to determine first information regarding the suitability of the light beam (225), and evaluating the suitability of the light beam of the at least one lighting module according to the at least one piece of first information regarding the suitability of the light beam (226).

Description

Technical Field

[0001] The present invention relates to the field of evaluating the conformity of the light beam projected by an automotive lighting module. In particular, it relates to a method and an apparatus for evaluating the light beam of an automotive lighting module.

[0002] The present invention is particularly useful when correcting the focus of an automotive lighting module or when correcting the luminous intensity distribution within the projected light beam.

Background Art

[0003] Automobiles are equipped with lighting modules on the right and left sides of the front of the vehicle to perform a plurality of lighting functions such as the "low beam" or LB function and the "high beam" or HB function.

[0004] The lighting module can perform the above two functions from the same light source. Alternatively, a plurality of separate modules are provided. A first module pair is dedicated to the LB function, and another module pair is dedicated to the HB function.

[0005] The term "focus" is understood to mean the direction and shape of the light beam projected by an automotive lighting module. Such shape and direction may be subject to regulatory constraints, particularly to provide good visibility of the scene in front of the driver while not dazzling other road users.

[0006] Furthermore, due to regulatory constraints, the distribution of the luminous intensity projected within the beam can be defined. In particular, for a specific function, it can be stipulated that the projection of the left lighting module onto the ground has a symmetrical shape and the luminous intensity is evenly distributed within each light beam.

[0007] The focal point of a lighting module can be observed or characterized by its projection onto the ground. In particular, the focal points of two identical lighting modules can be evaluated by their respective projections within a reference zone. Thus, during a technical inspection of a vehicle, experts can verify whether the focal point of each lighting module conforms to regulatory constraints, or whether the focal points of identical lighting modules are reliably symmetrical. Furthermore, experts can verify the correct distribution of luminous intensity within the beam of each lighting module and across multiple beams of a pair of lighting modules.

[0008] However, technical inspections of such vehicles are not frequent enough, and improper changes in the focus or light distribution of the lighting modules may go undetected for months or even years. During this time, such improper changes in focus can reduce the driver's visibility of the scene and / or dazzle other road users, which can lead to accidents.

[0009] Thus, the inability to periodically evaluate the suitability of the light beam characteristics of a lighting module to predetermined characteristics (e.g., regulatory or technical features) can lead to safety problems for vehicle occupants and other road users. [Overview of the project] [Problems that the invention aims to solve]

[0010] Therefore, there is a need to periodically evaluate the suitability of the light beam of automotive lighting modules without relying on inspections by experts. [Means for solving the problem]

[0011] For this purpose, a first aspect of the present invention relates to a method for evaluating the suitability of a light beam of at least one lighting module of an automobile, the method comprising a preliminary step of storing a model in the memory of the automobile, the model being capable of determining at least one piece of information relating to the suitability of the light beam from at least one image received as input to the model, Verify that at least one lighting module in the vehicle is activated, To acquire at least one first image from data transmitted from an image acquisition device of a vehicle, which is positioned to acquire data representing the scene facing the vehicle, To determine at least one piece of first information regarding the suitability of the light beam, the model is applied to the at least one first image, Based on the said at least one first conformity information, the conformity of the light beam of the said at least one lighting module is evaluated, Includes.

[0012] Therefore, the present invention makes it possible to automatically evaluate the suitability of the light beam of an automobile's lighting module by applying a model stored in the vehicle and a first image acquired from data from an image acquisition device. Since such image acquisition devices are generally integrated into most modern vehicles, the present invention does not require any additional sensors specifically for this invention and can therefore be implemented at low cost.

[0013] According to some embodiments, the method may further include a preliminary step of obtaining the model from a training dataset by machine learning.

[0014] Therefore, the model may be derived from machine learning, which may provide reliable predictions of first suitability information.

[0015] In addition, the model may be acquired by supervised learning, and the training dataset may include associations between at least one training image and at least one piece of information regarding the suitability of the light beam.

[0016] Therefore, by forming a training database that includes both training images and the corresponding relevance information, an accurate model for predicting relevance information may be obtained, stored in the vehicle, and then used in the current phase to obtain (or predict) relevance information.

[0017] According to some embodiments, the model may be capable of determining a binary compatibility category from among the compatible and non-compatible categories.

[0018] Therefore, the model may be a simple classifier with two possible outputs, which reduces the cost associated with training the model and the execution time required to apply it.

[0019] Alternatively, the model may be capable of determining the type of non-conformity from at least two different types.

[0020] Therefore, by distinguishing between multiple non-conformity types, richer conformance information than binary information may be obtained. Such distinctions improve the evaluation provided upon completion of the method and facilitate corrective actions by the operator if the light beam is non-conformity.

[0021] In addition, when applying the model to the at least one first image, a first non-conformity type may be determined, and the evaluation may include instructions for modifications to be made to the lighting module, the modifications depending on the determined first non-conformity type.

[0022] Therefore, if the light beam is non-compliant, it may be easier for the operator to take corrective action.

[0023] According to one embodiment, the image acquisition device may be a camera capable of acquiring color images or grayscale images.

[0024] Therefore, it is possible to use the cameras present in most automobiles, and the cameras are used, for example, for driving assistance functions related to the lateral or longitudinal trajectories of the vehicle or for parking assistance functions of the vehicle. Therefore, the costs associated with the implementation of the method according to the present invention are reduced.

[0025] In addition, the model may be able to determine at least one piece of information regarding the suitability of the light beam from at least one color image or grayscale image received as an input, and the at least one acquired first image may be in color or grayscale and may be transmitted from an image acquisition device.

[0026] Therefore, the image of the camera can be directly used as an input to the model, thereby accelerating the execution of the method according to the present invention.

[0027] Alternatively, the model may be a first model capable of determining at least one piece of information regarding the suitability of the light beam from at least one luminance image, and the method may further include a preliminary step of storing a second model in the automobile, and the second model can determine at least one luminance value of pixels of the luminance image from at least one value of pixels of the color image or grayscale image. The at least one acquired first image may be a first luminance image. Acquiring the at least one first luminance image may include applying the second model to the pixels of at least one first color image or first grayscale image acquired by an image acquisition device. The first model can be applied to the at least one first luminance image to determine the at least one piece of information regarding the suitability of the light beam.

[0028] The prediction of suitability information for light beams is improved by considering luminance images as input to the model. In fact, luminance images provide better contrast and better identification not only for the geometric characteristics of the light beam but also for the luminance distribution within the beam. Furthermore, the luminance images are determined from a second model, thereby avoiding the use of expensive and slow-processing luminance cameras.

[0029] According to another alternative, the image acquisition device is an infrared camera capable of acquiring infrared images, and the model may be capable of determining at least one piece of information relating to the suitability of a light beam from at least one infrared image received as input.

[0030] By considering infrared images as input to the model, better predictions regarding the suitability of the light beam become possible compared to the use of grayscale or color images.

[0031] According to some embodiments, the at least one activated lighting module may include a right-side low-beam module and a left-side low-beam module of the vehicle.

[0032] Therefore, the suitability of the low beams can be evaluated, thereby enabling the detection of non-suitability, in particular, which may dazzle other users of the road on which the vehicle is traveling and / or may insufficiently illuminate the scene in front of the vehicle, impairing the visibility of the vehicle's driver.

[0033] According to some embodiments, a first model can determine at least one piece of information relating to the suitability of a light beam from a series of images received as input, the at least one first image may be a series of first images, and the model can be applied to a series of first images to obtain the at least one piece of first information.

[0034] By considering a series of images as input to the model, better accuracy is provided for predicting information regarding the suitability of the light beam.

[0035] A second aspect of the present invention relates to a computer program that includes instructions for performing a method according to the first aspect of the present invention when executed by a processor.

[0036] A third aspect of the present invention relates to a vehicle device, wherein the device is A memory for storing a model capable of determining at least one piece of information regarding the suitability of a light beam from at least one image received as input to the model, or an access interface for accessing the first model, It is a processor, Verify that at least one of the vehicle's lighting modules is activated, To acquire at least one first image from data transmitted from an image acquisition device of a vehicle, which is positioned to acquire data representing the scene facing the vehicle, To determine at least one piece of first information regarding the suitability of the light beam, the model is applied to the at least one first image, Based on the said at least one first conformity information, the conformity of the light beam of the said at least one lighting module is evaluated, Includes. [Brief explanation of the drawing]

[0037] Further features and advantages of the present invention will become apparent upon reference to the following detailed description and accompanying drawings.

[0038] [Figure 1] This document shows an automobile according to several embodiments of the present invention. [Figure 2] The steps of a method according to several embodiments of the present invention are shown below. [Figure 3a] The light beams of automotive lighting modules according to some embodiments of the present invention are shown when the light beams are suitable. [Figure 3b] The following shows the light beam of an automotive lighting module according to some embodiments of the present invention when the light beam is incompatible. [Figure 4] The present invention illustrates an apparatus for calibrating an automotive lighting module according to several embodiments of the present invention. [Modes for carrying out the invention]

[0039] This description focuses on features that distinguish the methods and apparatus from those known in the prior art.

[0040] Figure 1 shows a vehicle 100 according to several embodiments of the present invention.

[0041] Vehicle 100 is equipped with a device 101 according to several embodiments of the present invention, the device 101 capable of evaluating the suitability of the light beam of at least one lighting module of vehicle 100. It should be noted that the suitability is evaluated against a predetermined set of criteria, which may be regulatory standards applicable to all vehicles traveling in a given area, and / or technical standards derived from the manufacturer of vehicle 100.

[0042] The vehicle 100 actually comprises a pair of lighting devices 103, each including at least one lighting module capable of performing at least one lighting function of the vehicle 100.

[0043] In the example in Figure 1, the vehicle 100 is shown as a side view, and therefore a single lighting device 103, such as a left-side lighting device, is shown in the figure. However, it will be understood that the vehicle 100 may also have a further lighting device located on the right side of the vehicle, positioned symmetrically with respect to the longitudinal axle of the vehicle with respect to the lighting device 103.

[0044] In the example shown in Figure 1, the lighting device 103 includes a first lighting module 104 dedicated to the first lighting function and a second lighting module 105 dedicated to a second lighting function distinct from the first lighting function. Alternatively, a single lighting module can implement both the first and second lighting functions.

[0045] The first lighting function may be a low beam function called LB, while the second lighting function may be a high beam function called HB.

[0046] Vehicle 100 further comprises a control module 102 for the lighting device 103, which can control the lighting device 103, and the control module 102 can start / stop the first and / or second functions of the vehicle. If the light sources of lighting modules 104 and 105 are of a matrix type with multiple individually controllable light-emitting elements, the control module 102 can also transmit pixelated illumination photometry, from which the lighting device 103 can selectively start / stop the light-emitting elements of any of the light sources of the lighting modules.

[0047] The control module 102 may be dedicated solely to controlling the lighting system, or it may also control other functions of the vehicle. In particular, the control module 102 may be the central control module of the vehicle 100.

[0048] The vehicle 100 according to the present invention may further include an image acquisition device 106.

[0049] The image acquisition device 106 is capable of acquiring data representing the scene of a vehicle located on the opposite side of vehicle 100, that is, the area in front of vehicle 100 in the direction of vehicle 100's travel. In particular, the image acquisition device 106 may be positioned, for example, in front of the vehicle, on the top of the windshield of vehicle 100 or on the bumper of vehicle 100.

[0050] The image acquisition device 106 may be an RGB type camera. An RGB (red-green-blue) type camera can acquire color image data, process it in real time, and therefore measure light parameters related to the vehicle environment.

[0051] Therefore, the data acquired by the image acquisition device 106 may be a color image, that is, each image includes pixels arranged in a matrix, and each pixel is encoded to correspond to a color from exactly two or more colors. For example, each pixel may be encoded in red-green-blue, i.e., RGB, that is, all colors are represented by three coordinates, red, green, and blue, respectively. Therefore, each pixel may be encoded with multiple bits, for example, one byte encoding 256 colors.

[0052] Alternatively, the data acquired by the image acquisition device 106 may be a grayscale image, that is, each image contains pixels arranged in a matrix, and each pixel is encoded in exactly two or more scales, having at least one intermediate grayscale between black and white. Thus, each pixel may be encoded in multiple bits, for example, one byte encoding 256 grayscales.

[0053] Alternatively, the image acquisition device 106 may be an infrared camera capable of acquiring data that is an infrared image, the infrared image including pixels arranged in a matrix, and each pixel being associated with an infrared value.

[0054] The image acquisition device 106 may be capable of periodically acquiring still images at a predetermined frequency, for example, every second. However, in order to make the data transmitted from the image acquisition device 106 usable for the vehicle's real-time functions, the image acquisition device 106 may preferably be capable of acquiring images that form video frames at multiple frames per second, i.e., at a frequency of fps greater than, for example, 10 fps, particularly at a frequency equal to 30 fps.

[0055] The vehicle 100 may be equipped with a memory 107 capable of storing data for the first model and / or the second model described later. Alternatively, the memory 107 may be integrated into the control module 102 or the device 101.

[0056] Vehicle 100 may be equipped with a human-machine interface 108, or HMI 108. There are no limitations on the HMI 108, and it may include any interface elements capable of receiving commands from the driver or passengers of the vehicle, whether or not such commands relate to the control of the vehicle or the control of interior or exterior equipment, and / or transmit information to the driver and / or passengers of the vehicle. For this purpose, the HMI 108 may include a screen such as a touchscreen, one or more sets of buttons, a speaker, a microphone, a dashboard capable of displaying one and / or more illuminated pictograms of a predetermined shape, a system for vibrating the steering wheel, and the like.

[0057] Figure 2 shows the steps of a method for testing the suitability of the light beam of at least one lighting module, according to some embodiments of the present invention.

[0058] The method may include a pre-phase 200 and a current phase 220. The pre-phase includes steps 201 to 203 described below, and may further include steps 211 to 213 in a particular embodiment.

[0059] The first training database is formed during step 201, and the first training database contains the first training data. The first training data includes the first training images acquired by cameras installed in each vehicle, which are located close to the position of camera 106 described above, i.e., - If camera 106 is positioned in the upper section of the vehicle's windshield, training images are acquired by cameras positioned in the upper section of each of the vehicle's windshields. - When camera 106 is positioned at the bottom of the front body of the vehicle, training images are acquired by cameras positioned at the bottom of each part of the vehicle's body, especially to perform the vehicle's parking assist function.

[0060] The training image represents a scene facing a given vehicle when at least one of the vehicle's lighting modules is turned on.

[0061] The first training image may be any of the following: - Grayscale image, - Color image, - Infrared image, - Also called a luminance image or luminance map.

[0062] There are no restrictions on how the first training data is acquired. In particular, the first training data may be collected, for example, from a group of vehicles in motion during a technical inspection.

[0063] In one embodiment, the first training database includes a time series of training images.

[0064] Each first training image or each time series of training images may be associated with information regarding the suitability of the light beam in the training database. Therefore, the first training data further includes suitability information for each first training image or time series of training images.

[0065] Each piece of relevance information associated with one or more training images is called reference relevance information, and it differs from the relevance information obtained (or predicted) as output from the model described later, in that it is true information, not prediction.

[0066] The compatibility information represents the compatibility or non-compliance of at least one light beam present in the associated first training image (or the associated time series of training images). The compatibility information can also indicate the compatibility or non-compliance of all projected light beams, for example, two light beams projected by a low-beam illumination module.

[0067] Therefore, the relevance information is used to label the first training image (or a time series of training images) in the first database.

[0068] There are no restrictions on the criteria related to conformity, and these may be the manufacturer's technical standards or regulatory standards.

[0069] Furthermore, suitability may relate to the following: - The shape and / or direction of one or more light beams, the suitability of which can be evaluated from the projection onto the ground, and / or - The distribution of luminous intensity within one and / or two light beams.

[0070] According to one embodiment, the conformance information may be a binary conformance category, that is, the conformance information can take two values: a "conformable" category and a "non-conformable" category.

[0071] Alternatively, conformance information can take at least three values ​​from the categories "Conforms," ​​"Non-conforms, First Non-conformity Type," and "Non-conforms, Second Non-conformity Type." More generally, conformance information can identify N non-conformity types using N, which is an integer greater than or equal to 2. The types of non-conformity are: - Indicates the level of non-conformity from among multiple levels, e.g., "mild," "moderate," and "severe," and / or, - Describe the non-conformity. The type of non-conformity may indicate, in particular, that the right or left beam is too high, too low, too far to the right, or too far to the left. Furthermore, the type may indicate a value of deviation from the non-conformity situation, which may be, for example, a distance or an angle value.

[0072] During step 202, the first model is trained using machine learning with the first training data from the training database acquired in step 202. The model thus trained can do the following: - At least one first image is received as input. As described above, depending on the first training image used (grayscale, color, luminance, or infrared scale), the model can be: namely, - Receive a first image of the same type as the first training image. - Determine or predict at least one piece of relevance information from the first image received as input.

[0073] Such a model can be acquired through supervised learning, where reference relevance information associated with the first training image is compared with the prediction of relevance information output by the model. Therefore, the model minimizes the error in determining the at least one piece of relevance information.

[0074] The principles of supervised learning are well known and will not be explained in further detail in this application.

[0075] The first model, thus trained, can predict at least one piece of suitability information based on a first image received as input, or a series of first images received as input.

[0076] Step 202 may be performed by a training module not shown in Figure 1, which may be located outside of vehicle 100 and have access to the first training database.

[0077] During step 203, the first model is stored in the vehicle 100 so that the device 101 can access it. The first model may be stored, for example, in the vehicle's memory 107 or in the internal memory of the device 101.

[0078] There are no restrictions on the type of first model; it may be a support vector machine, i.e., an SVM-type classifier, or a convolutional neural network, or any other model that can be constructed by machine learning and can accept images as input.

[0079] If the first model is trained to receive grayscale or color images, according to some embodiments, the input of the first model may be capable of directly receiving images transmitted from an image acquisition device 106, which is a grayscale or color camera. Such cameras are installed in most vehicles today, particularly for lateral / longitudinal driving assistance functions (cameras at the top of the windshield) or parking assistance functions (cameras at the bottom of the vehicle body). Thus, the costs associated with these embodiments are reduced.

[0080] According to another embodiment, the first model is trained to receive infrared images. In this case, the input to the first model may be capable of directly receiving images transmitted from an image acquisition device 106, which is an infrared camera. The use of infrared images provides better accuracy for predicting the at least one suitability information due to better contrast compared to grayscale or color images.

[0081] In yet another embodiment, the first model is trained to receive luminance images.

[0082] Luminance is a quantity that represents the amount of light emitted from an illuminated surface and reflected back to the eye, and is expressed in candelas per square meter, or cd / m². Therefore, luminance represents the visual perception of the brightness of a surface.

[0083] More specifically, luminance is the output of visible light passing through or emitted by a surface element in a given direction, per unit of surface area and per unit of solid angle.

[0084] In external lighting, luminance measurements are normalized based on factors such as distance from the surface, observer's position and height, and observation angle.

[0085] Therefore, unlike grayscale images, luminance maps show the brightness perceived by the camera's sensor at a given location, where the grayscale of a pixel indicates the brightness, but this does not correspond to the luminance as defined above.

[0086] Predicting suitability information based on luminance images provides greater accuracy in predictions, particularly when the scene facing the vehicle is illuminated by streetlights or interior lighting (parking lot), making it difficult to distinguish ambient lighting from that emanating from the light beams of the vehicle's lighting modules.

[0087] In this embodiment, the input of the first model may be capable of directly receiving images transmitted from an image acquisition device 106, which is a luminance camera. However, such luminance cameras are expensive and require significant processing time to acquire luminance images. Therefore, this is not compatible with real-time applications.

[0088] Preferably, according to this embodiment of the present invention, the second model is trained to predict a luminance image from a grayscale or color image received as input. Thus, the second model can be placed between the image acquisition device 106 and the first model performed by device 101, in which case the image acquisition device 106 can be a color or grayscale camera having the advantages described above.

[0089] The second model can be obtained according to steps 211-213 described later.

[0090] A second training database may be formed during step 211.

[0091] The second training database may include second training data, which may include second grayscale or color training images transmitted from one camera or several different cameras. In particular, the second training images in the second training database may be transmitted from different cameras, especially different types of cameras, where the two different types have at least one different technical feature or are from different manufacturers. By forming a second database that includes second training images from different types of cameras, it becomes possible to train a general-purpose model, which can then be integrated into any vehicle, regardless of whether the vehicle 100 has an image acquisition device 106 (capable of acquiring grayscale or color images).

[0092] Alternatively, all images are transmitted from a camera of the same type as the image acquisition device 106, in which case the second model trained from the second training data is specific to the image acquisition device 106 of the vehicle 100.

[0093] Preferably, the second training data in the second database formed in step 211 includes the association between the aforementioned second training images and their respective training luminance maps. Each training luminance map or image associated with a second training image may also be acquired by a luminance camera positioned next to the camera that acquires the second training images, thereby acquiring visual data about the same scene as that of the camera.

[0094] During step 212, the second model can be trained based on the second training data of the second database formed in step 211. There are no restrictions on the second model that results from the training in step 212, and the second model may be the following: - A curve or function that maps the pixel values ​​of a grayscale or color image to the luminance values ​​of corresponding pixels in a luminance map. Such a curve can be obtained using curve fitting analysis techniques based on associations in a second training database. In fact, based on these associations in the second training database, each association may be associated with a second training image to a training luminance map, and pixel associations between given pixels in the training image and corresponding pixels in the training luminance map may be determined. These pixel associations are fed into curve fitting analysis to obtain a model of the curve's form. - A support vector machine, or SVM model, that can determine the luminance value of a pixel in a luminance map based on the corresponding pixel value in a grayscale or color image through classification. The SVM model can be constructed by supervised learning from associations in a second training database. In fact, based on these associations in the second training database, each association can be associated with a training image in a training luminance map, and the association between a given pixel in the second training image and the corresponding pixel in the training luminance map can be determined. These associations between pixels are fed into the supervised learning of the SVM model. The advantage of the SVM model is that it can be formed based on a second training database containing a limited amount of data. - A neural network, such as 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 acquired by machine learning from the second training data of the second training database formed in step 211. In the case of supervised learning, the neural network can be configured by modifying the parameters of a particular neuron based on the measured deviation between the luminance map determined or predicted by the neural network when a given second training image is submitted thereto, and the training luminance map associated with the given second training image in the second training database. - Any other model derived from machine learning or data analysis techniques that can determine luminance values ​​from the values ​​of grayscale pixels or color pixels, or that can determine an entire luminance map from a grayscale or color image received as input.

[0095] Step 212 can be performed by a training module not shown in Figure 1, which can be installed outside of vehicle 100 and has access to a second training database.

[0096] Upon completion of training in step 212, which is the analysis or machine learning described above, a second model according to the present invention is obtained, which is capable of determining luminance values ​​from the values ​​of grayscale pixels or color pixels, or is capable of determining an entire luminance map from a grayscale image or color image received as input.

[0097] During step 213, the second model derived from the training in step 212 is stored in the vehicle 100 either in the memory of the device 101 or in memory outside of the device 101 but accessible by the device 101.

[0098] The method according to the present invention further includes a current phase 220, which includes steps 221 to 227 described later.

[0099] During step 221, the device 101 verifies that at least one lighting module of the vehicle 100 is turned on. In particular, according to the present invention, the device 101 may verify that the right and left low-beam lighting modules are turned on. Such verification may include confirming the activation status of the low-beam lighting modules by polling the control module 102, or may be based on identifying the light beams in the data received from the image acquisition device received in step 222 (in which case the verification of the lighting module activation in step 221 is performed after step 222).

[0100] If the at least one lighting module of the vehicle 100 is not turned on, the device 101 may request the control module 102, in particular, to turn on the at least one lighting module. Alternatively, the device may periodically poll the activation status of the at least one lighting module, for example, at predetermined intervals, and the following steps of Phase 220 are now performed only when the at least one lighting module is turned on.

[0101] During step 222, the device 101 receives data from the image acquisition device, which includes one or more images from the image acquisition device 106.

[0102] During step 223, the device 101 acquires at least one first image from the data received from the image acquisition device 106 during step 222.

[0103] The at least one first image obtained in step 223 may be a single first image or a time series of multiple first images.

[0104] If the data from the image acquisition device 106 includes a single image, this image forms the first single image acquired in step 223. If the data from the image acquisition device 106 includes multiple images acquired sequentially, the first single image may be any of the following: - An image selected from multiple images continuously acquired by the image acquisition device 106. - An image determined from multiple images acquired sequentially. The image determined in this way may be a stabilized image or an averaged image, thereby obtaining a clearer image than each individual image acquired by the image acquisition device 106.

[0105] Alternatively, if the first model is capable of receiving a series of images as input, then at least one of the first images is a time series of first images (a series of first images). In this case, the time series of first images (a series of first images) can be a time series of images (a series of images) acquired by the image acquisition device 106.

[0106] Each first image 106 may be a grayscale image, a color image, an infrared image, or a luminance image, depending on the embodiment corresponding to the first model acquired during the pre-phase 200.

[0107] If the first model is capable of receiving luminance images as input, the method may include step 224, which is performed by the device 101 between step 222 and step 223, to apply the stored second model to data transmitted from the image acquisition device 106 to acquire at least one first luminance image (a first single luminance image or a time series of luminance images (a series of luminance images)).

[0108] To acquire a time series of the first brightness image (a series of first images), the device 101 sequentially applies the second model to a series of images (grayscale or color) transmitted from the image acquisition device 106.

[0109] During step 225, following step 223 in which the at least one first image is acquired, the device 101 applies a first model stored in the vehicle 100 to the at least one first image in order to acquire at least one suitability piece of information by prediction.

[0110] As described above, the conformance information may be binary, or it may take exactly two or more values ​​to indicate the type of non-conformance in the case of non-conformance.

[0111] During step 226, the apparatus 101 evaluates the conformity of the light beams of at least one lighting module based on the at least one first conformity information. In particular, the evaluation indicates whether one or more light beams emitted from the lighting module conform to technical and / or regulatory standards.

[0112] The evaluation for Step 226 can be obtained from the following: - If the first model is applied to the first image only once, the at least one piece of first conformance information transmitted from step 225. - Multiple suitability information derived from at least two iterations of steps 222-225 described above. At least one first suitability information is obtained for each iteration, and the first suitability information is considered to evaluate the suitability of one or more light beams, thereby improving the accuracy related to the suitability evaluation.

[0113] During step 227, the device 101 transmits the conformity assessment obtained in step 226 to another entity of the vehicle 100. Such a transmission can only be performed if the assessment indicates that at least one light beam is unconformed.

[0114] Therefore, the evaluation may be transmitted to the driver or a third party so that corrective measures can be taken in the event of non-conformity. The evaluation may also indicate corrective measures or modifications to be taken, which are determined by the device 101 from the at least one first conformity information.

[0115] The evaluation may be submitted to, for example, the following: - HMI 108 to notify the driver that at least one light beam is non-compliant or that the light beams are compliant. - Memory 107 for storage and later retrieval. - A wireless communication module accessible via a telecommunications network for sending evaluations to a remote server.

[0116] Figure 3a shows the light beams 301.1 and 301.2 of the lighting modules 103.1 and 103.2 of the automobile 100 when the light beams are fitted, according to some embodiment of the present invention.

[0117] Figure 3a illustrates, as an example, the ground projection of the light beams projected by lighting modules 103.1 and 103.2, which are the low beam, right-side, and left-side lighting modules. The projections of light beams 301.1 and 301.2 are shown as an overhead view from above the road onto which the light beams 301.1 and 301.2 are projected. Such an overhead view differs from the viewpoint of the image acquisition device 106 mounted on the vehicle. However, this overhead view is shown to make the criteria provided as an example easier to understand.

[0118] The criteria for the conformity of light beams 301.1 and 301.2 may be that the light beams 301.1 and 301.2 do not intersect within the reference zone 302, which may be a range of a predetermined distance from the front of the vehicle. More specifically, the criteria may specify that the two beams intersect at the end of the reference zone 302.

[0119] Therefore, in the example shown in Figure 1, light beams 301.1 and 301.2 meet the above criteria. If the first model is trained to identify conformity information for this criterion only, training images acquired for light beams having projections similar to those in Figure 3a can be associated with conformity information indicating that the light beams are conforming.

[0120] The criteria defined from the reference zone are provided as examples. Other criteria may be evaluated additionally or as alternatives to characterize conformity.

[0121] Figure 3b shows the light beams 311.1 and 311.2 of the lighting modules 103.1 and 103.2 of vehicle 100 in the case of non-conformity of the light beams to the above-described criteria, according to some embodiments of the present invention.

[0122] In the case of Figure 3b, light beams 311.1 and 311.2 intersect within the reference zone and do not meet the above criteria.

[0123] If the first model is trained to identify conformity information for this criterion only, training images acquired for a light beam with a projection similar to that in Figure 3b may be associated with conformity information indicating that the light beam is non-conforming.

[0124] As described above, conformity information can be more accurate than simply indicating non-conformity and can also indicate the type of non-conformity. In the example in Figure 3b, conformity information may indicate that the non-conformity is due to the beams intersecting in reference zone 302. More specifically, conformity information may indicate that the right-hand beam is positioned excessively low.

[0125] Figure 4 shows the structure of the apparatus 101 according to several embodiments of the present invention.

[0126] The device 101 includes a processor 401 configured to communicate unidirectionally or bidirectionally with memory 402, such as random access memory (RAM) or read-only memory (ROM), or any other type of memory (flash, EEPROM, etc.), via one or more buses or via a direct wired connection. Alternatively, memory 402 may comprise multiple memories of the aforementioned types.

[0127] Memory 402 is capable of permanently or temporarily storing at least a portion of the data used and / or transmitted from the execution of the steps of the method according to the present invention, as shown with reference to Figure 2.

[0128] In particular, memory 402 may be capable of storing the first model during step 203 and optionally the second model during step 213.

[0129] Furthermore, the at least one first image obtained in step 223 may be temporarily stored in memory 402.

[0130] The processor 401 is capable of executing instructions stored in memory 402 to perform steps 221-227 of the current phase 220, as described with reference to Figure 2. Alternatively, the processor 401 can be replaced with a microcontroller designed and configured to perform steps 221-227 of the current phase 220, as described with reference to Figure 2.

[0131] Device 101 is equipped with a first interface 403 capable of receiving data transmitted from image acquisition device 106.

[0132] During step 227 described above, the device 101 includes a second interface 404 that can transmit the evaluation to an entity outside the device 101.

[0133] Device 101 includes a third interface 405 that allows access to the first model and / or the second model when the first model and / or the second model are stored outside of device 101 within the vehicle 100.

[0134] The present invention is not limited to the embodiments described above as examples, but includes other modifications.

Claims

1. A method for evaluating the suitability of a light beam (301.1; 301.2; 311.1; 311.2) of at least one lighting module (105; 104; 104.1; 104.2) of an automobile (100), comprising a preliminary step (203) of storing a model in the automobile's memory (107; 402), wherein the model is capable of determining at least one piece of information relating to the suitability of the light beam from at least one image received as input to the model, and the method further, Confirm that at least one lighting module of the vehicle is activated (221), The acquisition of at least one first image (223) from data transmitted from an image acquisition device (106) of the vehicle, which is arranged to acquire data representing a scene facing the vehicle, To determine at least one piece of first information relating to the suitability of the light beam, the model is applied to the at least one first image (225), (226) Evaluating the suitability of the light beam of the at least one lighting module based on the at least one first suitability information, A method that includes [a certain feature].

2. The method according to claim 1, further comprising a preliminary step (202) of obtaining the model from a training dataset by machine learning.

3. The method according to claim 2, wherein the model is acquired by supervised learning (202), and the training dataset comprises an association between at least one training image and at least one piece of information relating to the suitability of the light beam.

4. The method according to any one of claims 1 to 3, wherein the model can determine a binary compatibility category from among compatible and non-compatible categories.

5. The method according to any one of claims 1 to 3, wherein the model can determine the type of non-conformity from at least two different types of non-conformities.

6. The method according to claim 5, wherein when the model is applied to the at least one first image (225), a first non-conformity type is determined, and the evaluation includes instructions for modifications to be made to the lighting module, the modifications depending on the determined first non-conformity type.

7. The method according to any one of claims 1 to 6, wherein the image acquisition device (106) is a camera capable of acquiring a color image or a grayscale image.

8. The method according to claim 7, wherein the model is capable of determining at least one piece of information relating to the suitability of the light beam from at least one color or grayscale image received as input, the at least one acquired first image being color or grayscale and transmitted from the image acquisition device (106).

9. The model is a first model capable of determining at least one piece of information relating to the suitability of the light beam from at least one luminance image, and the method further includes a preliminary step (213) of storing a second model in the automobile (100), wherein the second model is capable of determining at least one luminance value of a pixel in a luminance image from at least one value of a pixel in a color image or a grayscale image, The acquired at least one first image is a first brightness image, Acquiring the at least one first luminance image includes applying the second model (224) to the pixels of the at least one first color image or first grayscale image acquired by the image acquisition device (106), The first model is applied to the at least one first brightness image in order to determine the at least one piece of information relating to the suitability of the light beam. The method according to claim 7.

10. The method according to any one of claims 1 to 6, wherein the image acquisition device (106) is an infrared camera capable of acquiring infrared images, and the model is capable of determining at least one piece of information relating to the suitability of the light beam from at least one infrared image received as input.

11. The method according to any one of claims 1 to 10, wherein the at least one activated lighting module (104; 105; 104.1; 104.2) comprises a right low beam module (104.2) and a left low beam module (104.1) of the automobile (100).

12. The method according to any one of claims 1 to 11, wherein the first model is capable of determining at least one piece of information relating to the suitability of the light beam from a series of images received as input, the at least one first image being a series of first images, and the model is applied to the series of first images to obtain the at least one first suitability information (225).

13. A computer program comprising instructions for performing the method described in any one of claims 1 to 12 when executed by a processor (401).

14. A vehicle (100) device (101), A memory (402) for storing a model or an access interface (405) for accessing the model, wherein the model is capable of determining at least one piece of information relating to the suitability of a light beam from at least one image received as input to the model, Processor (401), Confirm that at least one lighting module (104; 105; 104.1; 104.2) of the vehicle is activated, To acquire at least one first image from data transmitted from an image acquisition device (106) of the vehicle, which is arranged to acquire data representing a scene facing the vehicle, To determine at least one piece of first information relating to the suitability of the light beam, the model is applied to the at least one first image, Based on the aforementioned at least one first suitability information, the suitability of the light beam of the at least one lighting module is evaluated, A processor that executes, A device equipped with the following features.

Images