Method for displaying neighboring vehicles, associated device and vehicle

Abstract

A method for displaying information using a screen (140) in a self-contained vehicle (100) characterized in that it comprises the following steps: Detection of a plurality of vehicles (200, 210, 220, 240) around the self-contained vehicle (100); Determination, by a first neural network, of a model identifier for each of the vehicles in the plurality of vehicles (200, 210, 220, 240); Display, on the screen (140), of the plurality of vehicles (200, 210, 220, 240), and of model information (201, 211, 221, 241) obtained from the identifier. Figure for the abstract: Figure 2

Description

Title of the invention: Method for displaying neighboring vehicles, associated device and vehicle

[0001] The invention relates to the display of information on a screen of a motor vehicle.

[0002] There is a need to display on the screen of a motor vehicle information of increasing interest to the driver and / or passengers of the motor vehicle, in particular when the vehicle is driven automatically.

[0003] For this purpose, the invention relates to a method of displaying by a screen in an ego motor vehicle, also called ego vehicle, characterized in that it comprises the following steps: - Detection of a plurality of vehicles around the ego vehicle, - Determination, by a first neural network, of an identifier of a model for each of the vehicles of the plurality of vehicles, - Display, on the screen, possibly of the vehicle ego (in other words: a representation of the vehicle ego), of the plurality of vehicles (in other words: a representation of the plurality of vehicles), and of model information obtained from the model identifier of one of the vehicles of the plurality of vehicles.

[0004] Thus, the passengers or the driver of the ego vehicle can view on the screen (and know) the model of the vehicles surrounding the ego vehicle.

[0005] The model information may include the model identifier.

[0006] A vehicle comprising a body, by model is meant the physical characteristics, including for example the (external) shape of the body, and possibly the materials making up the body and / or inscriptions on the body. The physical characteristics may also include the noise emitted by a propulsion engine of the vehicle.

[0007] According to one embodiment, the detection of the plurality of vehicles is implemented (or carried out) from video images (at least) obtained by a camera of the ego vehicle, and / or radar images obtained by a radar of the ego vehicle. The detection can also be carried out from radio signals received for example from each vehicle of the plurality of vehicles.

[0008] The detection may comprise isolating each vehicle from the plurality of vehicles in the video or radar images (and recognizing that they are vehicles). The detection step may comprise implementing a third neural network to perform this isolation (and recognition).

[0009] The first neural network may be a classifier neural network. It may be a convolutional neural network.

[0010] For example: - The input layer of the first neural network comprises a number of neurons equal to the number of pixels in the video images, and / or - The output layer of the first neural network includes a number of neurons equal to the number of vehicle models that the network can classify.

[0011] During the detection step, the position on the road of each vehicle of the plurality of vehicles may also be determined, for example relative to the vehicle ego (for example from the video and / or radar images). The position on the screen of each vehicle in the display may be determined from the position on the road thus determined.

[0012] According to one embodiment, the ego vehicle traveling on a road in one direction of travel, during the step of detecting the plurality of vehicles, the plurality of vehicles comprises all the vehicles which: - Are located around the ego vehicle, and - Are located at a distance less than a threshold, and - Travel in the direction of traffic.

[0013] Alternatively, vehicles parked at the roadside may also be detected.

[0014] According to one embodiment, the method comprises a step of determining a color identifier of each of the vehicles of the plurality of vehicles and the display step comprises displaying on the screen color information from the color identifier.

[0015] The color determination step includes the implementation of a second neural network.

[0016] For example, the color of each vehicle of the plurality of vehicles is displayed on the screen.

[0017] The determination of the color may comprise a step of establishing a histogram of the colors (of the pixels) of each vehicle of the plurality of vehicles. The color of the vehicle being that most often encountered in the histogram.

[0018] The first neural network, the second neural network and the third neural network are artificial neural networks, preferably electronic, that is to say implemented by an electronic circuit, for example a microcircuit, comprising for example a processor with a memory or a dedicated electronic circuit.

[0019] According to one embodiment, the method further comprises the following steps: - Transmission (for example, by a telecommunications network) to a server of a request including the model identifier, - Receiving a response to the request including the model information.

[0020] The model information received from the server may include a range of vehicle manufacturing dates, the different engine(s) available, and anecdotes relating to the vehicles.

[0021] Alternatively, the model information is in the memory of the vehicle ego.

[0022] The speed of each vehicle of the plurality of vehicles may also be determined during the detection step and optionally be displayed by the screen.

[0023] According to one embodiment, the method further comprises a step where the ego vehicle is driven automatically.

[0024] Thus, the method can be implemented during a step where the ego vehicle is driven automatically. The driver's attention can then be focused on the vehicles around the ego vehicle.

[0025] According to one embodiment, the method further comprises the following steps: - Invitation to enter a model name of a first vehicle (the invitation comprises for example an invitation message on the screen) of the plurality of vehicles, - Entering a model name (e.g. via a touchscreen keyboard or voice command), - Checking for a match between, on the one hand, the model name, and on the other hand, the model information or the model identifier, - Display model information on screen from verification result.

[0026] Alternatively, for each vehicle of the plurality of vehicles, the model information and possibly the color may be displayed (directly) on the screen.

[0027] The invention also relates to a computer program comprising instructions, executable by a microprocessor or a microcontroller, for implementing the method according to the invention.

[0028] The method according to the invention is for example implemented by an electronic driving assistance device. The invention therefore also relates to an electronic device configured to implement the steps of the method according to the invention, and a motor vehicle comprising the electronic device. The electronic device may comprise the first and / or the second neural network and / or the third neural network.

[0029] The characteristics and advantages of the electronic device, the motor vehicle and the computer program are identical to those of the method, which is why they are not repeated here.

[0030] It is understood that an element such as the electronic device, the screen, or another element is “configured to” carry out a step or an operation, by the fact that the element comprises means for (in other words “is shaped to” or “is adapted to”) carrying out the step or the operation. These are preferably electronic means, for example a computer program, data in memory and / or specialized electronic circuits.

[0031] When a step or operation is performed by such an element, this generally implies that the element comprises means for (in other words "is shaped for" or "is adapted for" or "is configured for") performing the step or operation. This also includes, for example, electronic means, for example, a computer program, data in memory and / or specialized electronic circuits.

[0032] Other characteristics and advantages of the present invention will appear more clearly on reading the detailed description which follows, comprising embodiments of the invention given as non-limiting examples and illustrated by the appended drawings, in which.

[0033] [Fig-1] represents an electronic device of a motor vehicle and a vehicle according to one embodiment of the invention.

[0034] [Fig.2] and [Fig.3] represent a display according to embodiments of the invention.

[0035] [Fig.4] represents the method of the invention, according to an exemplary embodiment, implemented in particular by the electronic device and the vehicle of [Fig.l].

[0036] Detailed description of an exemplary embodiment of the invention

[0037] [Fig.l], an ego vehicle travels on a road 300 in the direction Tl. The ego vehicle comprises a camera (or a plurality of cameras) 120, and / or a radar (or a plurality of radars) 130 capable of obtaining images of the vehicles all around the ego vehicle 100. The ego vehicle 100 comprises an electronic driving assistance device 110 comprising a first neural network 111, a second neural network 112, a third neural network 113, and a screen 140 in the passenger compartment 150 of the ego vehicle 100. The road further comprises the vehicles 200, 210, 220 and 240 also traveling in the direction Tl at a maximum distance D from the ego vehicle 100.

[0038] During the steps below, the ego vehicle 100 is driven by the electronic device 110 automatically.

[0039] [Fig.4], in step S10, the electronic device 110 detects all the vehicles 200, 210, 220 and 240 around the ego vehicle 100. The detection of the plurality of vehicles 200, 210, 220 and 240 may be implemented (or carried out) from video images (at least) obtained by the camera 120 of the ego vehicle 100, and / or from images obtained by the radar 130 of the ego vehicle 100. The third neural network 113 can receive these video and / or radar images as input and isolate the vehicles 200, 210, 220 and 240 (and recognize that they are vehicles) in the video and / or radar images output by the neural network. The device 100 can verify that the vehicles 200, 210, 220 and 240 are located at a distance less than a threshold from the ego vehicle 100, and are traveling in the direction of travel T1 (also from the video and / or radar images). The position of each vehicle of the plurality of vehicles 200, 210, 220, 240 is also determined, for example relative to the ego vehicle 200.

[0040] The third neural network may be as described in one of the following articles: - Joseph Redmon, Ali Farhadi, YOLOv3: An Incremental Improvement (https: / / arxiv.org / abs / 1804.02767), - Marius Cordts, Mohamed Omran, et al., The Cityscapes Dataset for Semantic Urban Scene Understanding (https: / / arxiv.org / abs / 1604.01685).

[0041] In step S20, the device 110 determines a color identifier for each of the vehicles 200, 210, 220 and 240.

[0042] The color determination step is for example carried out using the second neural network 112.

[0043] The determination of the color may comprise a step of establishing a histogram of the colors (of the pixels) of each vehicle of the plurality of vehicles. The color of the vehicle being that most often encountered in the histogram. The determination of the color is for example implemented by the method described in the following article: "Multi-Class and Single-Class Classification Approaches to Vehicle Model Recognition from Images", Daniel Munroe and Michael G. Madden. Proceedings of AICS-05: Irish Conference on Artificial Intelligence and Cognitive Science, Portstewart, Sept 2005.

[0044] In step S30, the neural network 111 determines a model identifier for each of the vehicles 200, 210, 220 and 240 detected in step S10 from the video images obtained by the camera 120.

[0045] The first neural network 111 may be a classifier neural network. It may be a convolutional neural network trained from an image base as described in the article: - Jonathan Krause et al., 3D Object Representations for Fine-Grained Categorization, (http: / / ai.stanford.edu / ~jkrause / papers / 3drrl3.pdf).

[0046] For example: - The input layer of the first neural network 111 comprises a number of neurons equal to the number of pixels of the video images, and / or - The output layer of the first neural network 111 comprises a number of neurons equal to the number of vehicle models that the network can classify.

[0047] In step S40, the device 110 controls the display by the screen 140, shown [Fig. 2], of the vehicles (or the representation of the vehicles) 200, 210, 220, and 240, for each vehicle, in the color determined in step S20, and at a position on the screen determined from the position on the road obtained in step S10. The device 110 also controls the display, on the screen 140, of a text 201 (for the vehicle 200), 211 (for the vehicle 211), 221 (for the vehicle 220) and 241 (for the vehicle 240), obtained from the model identifier for each vehicle. The text is for example constituted by the brand and model name of the vehicle.

[0048] For example, for vehicle 200, the model identifier and text 201 (i.e., the model information) is “Maserati Grecale.”

[0049] The text may further include the year of production of each vehicle of the plurality of vehicles received from a server in response to a request including the identifier transmitted to the server (the request and the response may be exchanged via a mobile telecommunications network).

[0050] Alternatively, in step S40, the user or driver of the ego vehicle 100 may be prompted to enter the model name of the vehicle 200. [Fig. 3], a flashing halo 400 may surround the vehicle 200 to indicate that the displayed question 500 (e.g., in 500, a text displays “give the model of the designated vehicle”) relates to the vehicle 200. If the user or driver enters (i.e., typed in), for example by voice command, a model name identical to or corresponding to the model identifier of the vehicle 200 determined in step S30, for example, “Maserati Grecale”, a message (i.e., model information) “Bravo, vehicle identified” or “Bravo, Maserati Grecal identified” is displayed on the screen 140.

Claims

Claims

1. A method of displaying by a screen (140) in an ego vehicle (100) automobile characterized in that it comprises the following steps: - Detection (S 10) of a plurality of vehicles (200, 210, 220, 240) around the ego vehicle (100), - Determination (S30), by a first neural network (111), of an identifier of a model for each of the vehicles of the plurality of vehicles (200, 210, 220, 240), - Display (S40), on the screen (140), of the plurality of vehicles (200, 210, 220, 240), and of model information (201, 211, 221, 241) obtained from the model identifier of one of the vehicles of the plurality of vehicles (200, 210, 220, 240), further comprising the following steps: - Inviting to enter a model name of a first vehicle (200) of the plurality of vehicles (200, 210, 220, 240), - Entering a model name, - Checking for a match between the model name and the model identifier,- Displaying the model information on the screen based on the result of the verification. The method being implemented by an electronic driving assistance device (100).,

2. Display method according to the preceding claim, the ego vehicle traveling on a road (300) in a direction of travel (Tl), wherein, during the step of detecting the plurality of vehicles (200, 210, 220, 240), the plurality of vehicles (200, 210, 220, 240) comprises all the vehicles which: - Are located around the ego vehicle (100), and - Are located at a distance (D) less than a threshold, and - Travel in the direction of travel (Tl).

3. A display method according to any preceding claim further comprising a step of determining a color identifier of each of the plurality of vehicles (200, 210, 220, 240) and the step of displaying (S40), on the screen (140), the plurality of vehicles (200, 210, 220, 240), and model information (201, 211, 221, 241) comprises displaying on the screen (140) color information from the color identifier.

4. A display method according to any preceding claim further comprising the following steps: - Transmitting to a server a request comprising the model identifier, - Receiving a response to the request comprising the model information.

5. A display method according to any preceding claim further comprising a step wherein the ego vehicle (100) is driven automatically.

6. Computer program comprising instructions, executable by a microprocessor or a microcontroller, for implementing the steps of the method according to any one of claims 1 to 5.

7. An electronic device (110) comprising a first neural network (111), a second neural network (112), a third neural network (113) and a display (140), configured to implement the steps of the method according to any one of claims 1 to 5.

8. Motor vehicle (100) comprising the electronic device (110) according to the preceding claim.

Images