Vehicle and method for anticipating a situation of driving in the wrong direction

Abstract

Vehicle and method for anticipating a one-way street situation. The present invention relates to a vehicle (1) comprising: - an image-capturing device (11), capable of providing an image of a traffic sign, - a positioning device (12) for the vehicle (1), capable of providing a position of the vehicle (1), - map data, - a control unit (10), configured to associate the position of the vehicle (1) with a traffic lane from the map data, to detect a one-way street in the image and to determine, based on the position of the vehicle (1) in the traffic lane and the direction of the vehicle (1), whether the vehicle (1) is traveling in the wrong direction or not. (Figure 1)

Description

Title of the invention: Vehicle and method for anticipating a situation of driving in the wrong direction

[0001] The present invention relates to the fields of the automotive industry and object detection and classification systems, and finds particular application in driver assistance systems or autonomous vehicles.

[0002] Despite the development of driver assistance systems in motor vehicles, fatal road accidents have decreased only slightly. The majority of these accidents are caused by driver negligence or misjudgments, which further underscores the importance of equipping motor vehicles with preventative devices to mitigate such negligence or errors. Examples of such errors include driving the wrong way down a one-way street, whether unintentionally or intentionally, such as a kamikaze-style reckless driving.

[0003] Implementing a device to prevent such a situation from occurring requires accurate recognition of road signs. However, these signs vary from one country to another. Road infrastructure can easily detect a vehicle traveling the wrong way, but cannot act on the vehicle in question except by mobilizing external resources, the intervention of which cannot be immediate.

[0004] There therefore remains a need for an effective device for preventing a vehicle from driving in the wrong direction, which makes it possible to avoid an accident, despite the difficulties mentioned above, in particular those which are linked to the differences in signage from one country to another.

[0005] The present invention aims to remedy at least in part the aforementioned drawbacks by providing a vehicle equipped with means for recognizing a wrong-way driving situation, and a method for preventing an accident-prone situation, which are robust to differences in signage from one country to another and allow direct intervention on the vehicle, using its driver assistance equipment.

[0006] To this end, the invention proposes a vehicle comprising: - an image capture device capable of producing an image containing a road sign, - a vehicle positioning device capable of providing a vehicle position and direction, - cartographic data, - a control unit, configured to associate the vehicle's position with a traffic lane from map data, to detect a one-way street the image and to determine, based on the vehicle's position on the traffic lane and the vehicle's direction, an authorized or unauthorized direction of vehicle movement, the control unit being further configured to take a preventive measure against an accident-prone situation when it determines an unauthorized direction of vehicle movement.

[0007] The image capture device is for example a monocular camera used by other driver assistance systems, for example by a level 2 autonomous driving module according to the OICA (International Organization of Motor Vehicle Manufacturers) classification.

[0008] The vehicle positioning system is a GNSS (Global Navigation Satellite System) type positioning system, possibly supplemented, for example, by an RTK (Real Time Kinematics) correction system. The vehicle positioning system allows the control unit to obtain the vehicle's position and direction. In fact, two successive positionings are sufficient to determine this direction.

[0009] The map data is taken from a map database present in the vehicle or is downloaded from a device external to the vehicle, this external device being either a server of the vehicle manufacturer or a phone of a vehicle user.

[0010] The control unit, for example, matches the detected no-entry sign to one or more traffic lanes from the map data, depending on the position on the image of the traffic sign on which this no-entry sign was detected. Alternatively, the control unit uses a neural network to perform this matching.

[0011] If the traffic lane associated with the vehicle is part of the traffic lane or lanes to which the detected no-entry sign corresponds, and if the direction of the vehicle shows that it is moving in this no-entry direction, then the control unit determines an unauthorized direction of travel for the vehicle, and otherwise an authorized direction of travel for the vehicle.

[0012] Thanks to the invention, the vehicle is equipped with a system that detects when it is traveling in an unauthorized direction, allowing the control unit of this detection system to take real-time measures to prevent an accident. Since the invention does not rely on road infrastructure, it is much more efficient. The control unit is, for example, part of the vehicle's main computer or communicates with it.

[0013] The control unit preferably comprises software for detecting the "no entry" sign in the image, based on one or more pre-trained neural networks. The control unit uses, for example, a first neural network to detect a A traffic sign is displayed in the image provided by the imaging device, and then a second neural network classifies this sign as a "no entry" sign. These neural networks are pre-trained using a database of images containing traffic signs from various countries. The signs in these images are located on different types of road infrastructure and correspond to different driving situations.

[0014] Alternatively, the control unit uses the first neural network to detect a traffic sign on the image provided by the image capture device, and then a classification algorithm based on an image analysis algorithm.

[0015] In another variant, the control unit uses a single neural network to directly detect a no entry sign on the image provided by the image capture device.

[0016] This use by the control unit of one or more neural networks pre-trained by deep learning algorithms makes it possible to overcome the variety of signs between different countries, and to be able to prevent an accident related to a violation of a no-entry sign in any driving situation, in most countries.

[0017] In one embodiment of the invention, the vehicle according to the invention further comprises a human-machine interface, and the control unit is configured to use the human-machine interface so as to warn a driver of the vehicle of an unauthorized direction of travel when such an unauthorized direction of travel is determined by the control unit. In this embodiment, the preventive measure taken by the control unit therefore consists at least of warning the driver of the vehicle.

[0018] For example, the human-machine interface includes an indicator light and an audible alarm, configured to flash and emit an audible alert every second until the vehicle brakes or changes direction. This implementation is one possible way for the vehicle user to realize their mistake and act accordingly. It also allows for the retrospective characterization of intentional or unintentional behavior on the part of the vehicle driver, for example, if the driver accelerates in the unauthorized direction of travel instead of braking. The control unit preferably records a history of the vehicle's driving parameters and the preventive measures it takes as soon as an unauthorized direction of travel is detected. This characterization is useful to authorities and insurance companies for determining liability after an accident.

[0019] In one embodiment of the invention, the vehicle according to the invention further comprises an emergency alert device, and the control unit is configured to detect, within a predetermined period according to a determination In the event of an unauthorized direction of travel, a decrease in vehicle speed in that direction is detected, with the control unit configured to activate the emergency service alert device if such a decrease in vehicle speed in that direction is not detected within the predetermined period. In this embodiment, the preventive measure taken by the control unit therefore consists, at a minimum, of alerting the emergency service.

[0020] The predetermined period allows, in particular, when the vehicle is traveling on a road with two lanes in opposite directions, for it to overtake without alerting the emergency services. When the road on which the vehicle is traveling has more than two lanes in each direction, the predetermined period can be shorter, for example, thirty seconds. It is therefore understood that the predetermined period depends on the type of road on which the vehicle is traveling according to the invention.

[0021] This embodiment of the invention makes it possible to intercept a suicide vehicle, or to intervene in the event of the driver falling asleep or becoming ill. According to one embodiment, the activation of the emergency service alert device is transparent to the vehicle user, i.e., imperceptible to them, so as to allow for effective interception in the case of a suicide vehicle.

[0022] In one embodiment of the invention, the vehicle according to the invention includes a braking control device, and the control unit is configured to detect, for a predetermined period following the determination of an unauthorized direction of travel for the vehicle, a decrease in the vehicle's speed in said direction, and to activate the braking control device when such a decrease in the vehicle's speed in said direction is not detected for the predetermined period. In this case, when the vehicle also includes a steering control device, the control unit is, for example, configured to perform emergency parking of the vehicle using both the steering control device and the braking control device when a decrease in the vehicle's speed in said direction is not detected for the predetermined period.In this embodiment, the preventive measure taken by the control unit therefore consists at least of braking the vehicle, or even carrying out an emergency stop.

[0023] The predetermined duration is preferably greater than the duration of the predetermined period, in order to avoid unnecessary emergency braking and / or parking in the event that the vehicle driver resumes normal driving conditions between the predetermined period and the predetermined duration. For example, the predetermined duration is set at 90 seconds.

[0024] In one embodiment of the invention, the vehicle according to the invention comprises a V2X (Vehicle-to-Everything) communication device, The control unit is configured, when it has determined that the vehicle is traveling in an unauthorized direction, to use the V2X communication device to alert at least one other road user or road infrastructure. In this embodiment, the preventive measure taken by the control unit therefore consists of at least warning one other road user.

[0025] V2X communication refers to communication between a vehicle and road infrastructure, a road user, or another vehicle. It uses, for example, a wireless connection such as Wi-Fi (according to the IEEE 802.11 standard) or LTE (according to the 3GPP "Long Term Evolution" standard). This embodiment of the invention allows other road users to avoid the vehicle traveling in the wrong direction, and / or allows the road infrastructure used by the vehicle to signal this vehicle traveling in the wrong direction to other road users.

[0026] In one embodiment of the invention, the map data includes an indication of the direction of travel on the traffic lane, and the control unit is configured to correlate said indication with a no-entry sign detected on the image and to determine an unauthorized direction of travel only if the direction of travel of said indication is compatible with the no-entry sign detected on the image.

[0027] In other words, if this correlation is negative—that is, if the no-entry sign detected in the image corresponds to a permitted direction of travel on the road as indicated by the map data—then the control unit takes no preventive measures to alert the driver, brake the vehicle, or alert the emergency services and / or other road users. This avoids taking such measures when they are inappropriate, for example, because the traffic sign is positioned on the road in a way that is misleading to the vehicle's system for detecting a no-entry direction. Other conditions may also inhibit the control unit's preventive measures in the event of a no-entry direction detection, such as the detection of roadworks.Indeed, in such cases, the usual road signs are not always covered, and vehicle navigation maps do not necessarily reflect traffic changes related to roadworks.

[0028] The invention also relates to a method for preventing an accident-prone situation, implemented by a vehicle according to the invention and comprising the steps of: - capturing an image in which a traffic sign appears, - detecting a no-entry sign in the image, - obtaining a positioning and direction of the vehicle, - associating a vehicle position provided by the positioning step with a traffic lane, using map data, - Determination of an authorized or unauthorized direction of travel based on the vehicle's direction and position on the traffic lane, and if an unauthorized direction of travel is determined, - taking a preventive measure against an accident-prone situation.

[0029] When the determination step results in the determination of an unauthorized direction of travel for the vehicle, the step of taking a measure to prevent an accident-prone situation consists of at least one of the following steps: - a visual and / or audible warning step by a vehicle driver, - an emergency service alert step when a vehicle's speed does not decrease for a predetermined period after an unauthorized direction of travel has been determined, - a vehicle braking phase when the vehicle's speed does not decrease for a predetermined period after an unauthorized direction of vehicle movement has been determined, - an emergency vehicle parking maneuver when the vehicle's speed does not decrease for a predetermined period after an unauthorized direction of travel has been determined, - an alert step for at least one other user of the traffic lane or road infrastructure.

[0030] The step of determining an authorized or unauthorized direction of traffic is preferably conditioned on: - compatibility between the no-entry sign detected in the image and a direction of travel indication on the lane derived from the vehicle's mapping data, and - an association of the traffic sign with the traffic lane.

[0031] The method of preventing an accident-prone situation according to the invention has advantages similar to those of the vehicle according to the invention.

[0032] Other features and advantages of the invention will become apparent from the following description on the one hand, and from several illustrative and non-limiting examples of embodiments given with reference to the accompanying schematic drawings on the other hand, in which:

[0033] [Fig-1] represents equipment of a vehicle according to the invention, in a mode of the realization of the invention, enabling the implementation of a method for preventing an accident-prone situation according to the invention,

[0034] [Fig.2] represents the first steps of a process for preventing a situation accident-prone according to the invention, in this embodiment of the invention, and

[0035] [Fig.3] represents other steps in the process of preventing an accident-prone situation from [Fig.2].

[0036] According to an embodiment shown in [Fig.1], a vehicle 1 according to the invention comprises a control unit 10, an image capture device 11, a positioning device 12, a mapping database 13, a human-machine interface 14, an emergency service alert device 15, a braking control device 16, a steering control device 17 and a V2X communication device 18.

[0037] The control unit 10 is for example the main computer of the vehicle, or a computer dedicated to the detection of objects outside the vehicle, comprising a graphics processor (or GPU) and capable of communicating with the main computer of the vehicle, for example via a CAN bus (Controller Area Network).

[0038] The control unit 10 is connected, for example via the CAN bus, to the various other equipment mentioned above, namely to the image capture device 11, the positioning device 12, the map database 13, the human-machine interface 14, the alert device 15 of an emergency service, the braking control device 16, the steering control device 17 and the V2X communication device 18.

[0039] The image capture device 11 is, in this embodiment of the invention, a camera mounted in the vehicle, for example a front camera installed on the windshield of the vehicle, and also used by other driver assistance systems, such as a lane keeping system on the motorway.

[0040] In this embodiment of the invention, the positioning device 12 is a GPS positioning device whose positioning is refined by a real-time kinematics technique called RTK (from the English "Real Time Kinematics") and possibly by an inertial measurement unit (IMU) of the vehicle 1. Thus, the control unit 10 is able to position the vehicle 1 to within a few centimeters, in particular with less than ten centimeters of error.

[0041] In this embodiment of the invention, the map database 13 is embedded in the vehicle 1 and is regularly updated by a high-definition map data provider. In other words, the map database 13 contains high-definition maps of the vehicle 1's geographic environment, for example, an entire region in which it regularly travels. Alternatively, the map database 13 contains only high-definition geographic data on the vehicle 1's immediate environment, updated by an external server as the vehicle 1 moves. This immediate environment may extend, for example, to a radius of one kilometer around the vehicle 1.

[0042] The human-machine interface 14 includes, for example, an indicator light on the dashboard of the vehicle 1, and a loudspeaker.

[0043] The 15 alert device of an emergency service uses, for example, a communication module of vehicle 1, equipped with a USIM card (from the English "Universal Subscriber Identity Module") from the manufacturer of vehicle 1. This 15 alert device is configured to call, via this communication module, a number of an emergency service, for example a gendarmerie, when it is activated.

[0044] The braking control device 16 is for example a VDC (for "Vehicle Dynamic Control") system or an AEBS (for "Advanced Emergency Braking Systems") automatic emergency braking system configured to brake the vehicle on command from the vehicle's main computer.

[0045] The steering control device 17 is for example an EPAS (Power-Assisted Steering System) controlled by the vehicle's main computer.

[0046] The V2X 18 communication device includes, for example, the communication module of the vehicle equipped with the USIM card, and is configured, for example, to send a broadcast message to other road users on which the vehicle 1 is traveling according to the invention, using this communication module.

[0047] Figures 2 and 3 illustrate a method 100 for preventing an accident-prone situation according to the invention, in this embodiment of the invention. This method 100 for preventing an accident-prone situation can be implemented by the vehicle 1 according to the invention.

[0048] A first step 102 of the method 100 for preventing an accident-prone situation is the capture of an image by the image capture device 11. This image is sent by the image capture device 11 via the CAN bus, to the control unit 10.

[0049] A second step 104 of the method 100 for preventing an accident-prone situation is the detection of a "no entry" sign in the image received by the control unit 10 during the first step 102. The control unit 10 implements a first neural network pre-trained on a database of images containing road signs on road infrastructure that may originate from different countries, to detect road signs in these images. During this second step 104, the image received by the control unit 10 is applied as input to this first neural network, which provides as output data representative of a road sign, in this example of use of the invention.

[0050] This output data from the first neural network is then applied to a second neural network, also implemented by the control unit, this second neural network being pre-trained to classify the data representative of a traffic sign at the output of the first neural network, in a type of traffic sign.

[0051] Alternatively, the control unit 10 implements a single neural network allowing direct detection in this second step 104, of a no entry sign on the image received during the first step 102. In yet another variant, the second neural network is replaced by a decision tree implemented by the control unit 10.

[0052] In this example of using method 100 for preventing an accident-prone situation, it is assumed that the second neural network provides as output one or more data points representing a no-entry sign.

[0053] A third step 106 of the method 100 for preventing an accident-prone situation is a step of obtaining a positioning of the vehicle 1. During this third step 106, the control unit 10 receives from the positioning device 12, a very precise position of the vehicle 1. Two successive positions of the vehicle 1 are in fact sent to the control unit 10 during this third step 106, in order to allow the control unit 10 to also obtain a direction of the vehicle 1.

[0054] A fourth step 108 of the method 100 for preventing an accident-prone situation is the association, by the control unit 10, of the position of the vehicle 1 obtained during the previous step, with a traffic lane identified using D-cart map data from the map database 13. In other words, the control unit 10 places the vehicle 1 on the traffic lane that it identifies.

[0055] During this fourth step 108, the control unit 10 also matches the no-entry sign detected during the second step 104 of the method 100 to one or more traffic lanes from the D-cart mapping data. To do this, the control unit uses an analysis of the image provided by the image-capturing device 11, which, for example, locates the traffic sign in relation to the traffic lane on which the vehicle 1 is located. In this example of using the method 100 to prevent an accident-prone situation, it is assumed that the traffic lane on which the control unit 10 locates the vehicle is affected by the traffic sign detected during the second step 104 of the method 100; that is, the traffic sign prohibits one direction of travel on the traffic lane on which the control unit 10 locates the vehicle.

[0056] The next step is then a fifth step 110 of determining an authorized or unauthorized direction of movement of vehicle 1, depending in particular on the direction of vehicle 1 determined during the third step 106.

[0057] In this fifth step 110, the control unit 110 determines an authorized direction of travel if the direction of the vehicle 1 on the traffic lane on which the unit The location of control unit 10 is consistent with the no-entry sign detected during step 104 of procedure 100, meaning that vehicle 1 is not traveling in a direction contrary to that indicated on the traffic sign in the image received by control unit 10. In this case, procedure 100 for preventing an accident-causing situation ends. In this example of using procedure 100 for preventing an accident-causing situation, it is assumed that control unit 10 did not determine an authorized direction of travel for vehicle 1 in this fifth step 110.

[0058] In this embodiment of the invention, the control unit 10 then determines a direction of travel on the traffic lane on which the control unit 10 locates the vehicle 1, using the D-cart map data. This direction of travel is, for example, explicitly coded in this D-cart map data, or deduced from the position of the traffic lane on the road to which it belongs, when this road is a two-way road.

[0059] If the direction of vehicle 1 in the traffic lane on which the control unit 10 locates it is incompatible with the no-entry sign as detected during the second step of the process 100, and if this direction is the opposite of the direction of travel determined by the control unit 10 using the D-cart map data, then the control unit 10 determines, in this fifth step 110, an unauthorized direction of travel for vehicle 1. This determination of an unauthorized direction of travel for vehicle 1 is possibly conditional upon the absence of roadworks detected by the control unit 10. In other cases, the process 100 for preventing an accident-prone situation ends.In other words, determining an unauthorized direction of travel requires that the direction of vehicle 1 contravenes the prohibited direction detected during the second step 104 of process 100, and the direction of travel as indicated by the D-cart map data.

[0060] Alternatively, in this fifth step 110, the control unit 10 determines in all cases a direction of travel on the traffic lane on which the control unit 10 locates the vehicle 1, using the D-cart map data, and determines an authorized or unauthorized direction of travel only if a direction of travel compatible with the prohibited direction on the image corresponds to the direction of travel as indicated by the D-cart map data.

[0061] In this example of use of method 100 for preventing an accident-prone situation, it is assumed that the control unit 10 has determined an unauthorized direction of travel for vehicle 1.

[0062] The next step is then a sixth step 112 of visual and audible warning to a driver of the vehicle, using the human-machine interface 14, and of alerting other road users using the V2X communication device 18.

[0063] In particular, in this sixth step 112, the control unit 10 can control the illumination of the indicator light on the human-machine interface 14 and emit an audible alert through the loudspeaker, the intensity of which increases every second. Furthermore, or alternatively, in this sixth step 112, the control unit 10 can broadcast a message using the V2X communication device 18, the broadcast message containing data representative of a notification that vehicle 1 is driving the wrong way and the position of vehicle 1. The visual and audible warning lasts, for example, 30 seconds, and the broadcast message is repeated regularly as long as vehicle 1 violates the one-way street detected in the second step 104.

[0064] In parallel with the sixth step 112, the control unit 10 can, in a seventh step 114, monitor the vehicle's speed v for a predetermined period, for example, 30 seconds in this embodiment of the invention. If, during this predetermined period, the vehicle's speed v has decreased, then (branch Y) the process 100 terminates. If, on the contrary, at the end of this predetermined period, the vehicle's speed v has not decreased, then (branch N) the process 100 continues (reference A) and the following step is an eighth step 116 of alerting an emergency service, during which the control unit 10 uses the vehicle's communication module to call an emergency service, advantageously in a manner imperceptible to a vehicle user. This alert step 116 is alternatively not implemented when the road on which the vehicle is traveling has only two lanes of traffic in opposite directions.

[0065] In parallel with the eighth step 116, the control unit 10 monitors, during a ninth step 118, the vehicle's speed v for a predetermined period, set for example at 60 seconds in this embodiment of the invention. If, during this predetermined period, i.e., within the ninety seconds following step 104 of determining an unauthorized direction of travel, the vehicle's speed v decreases, then (branch Y) the process 100 terminates. If, on the contrary, at the end of this predetermined period, the vehicle's speed v has not decreased, then (branch N) the control unit 10 uses the braking control device 16 and the steering control device 17 to perform 120 an emergency parking of the vehicle 1.

[0066] Alternatively or in combination, if at the end of this predetermined time the speed v of the vehicle has not decreased, then the control unit 10 limits the use of the accelerator of vehicle 1, for example to prohibit any acceleration of vehicle 1.

[0067] Of course, the invention is not limited to the examples just described, and many modifications can be made to these examples without departing from the scope of the invention. In particular, the characteristics of the different embodiments of the invention envisaged in this application, may be combined to realize the invention, insofar as these variants are not incompatible with each other.

Claims

Demands

1. Vehicle (1) comprising: - an image-taking device (11), capable of providing an image in which a traffic sign appears, - a vehicle (1) positioning device (12), capable of providing a position and direction of the vehicle (1), - map data (D-cart), - a control unit (10), configured to associate the position of the vehicle (1) with a traffic lane from the map data (D-cart), to detect a no entry on the image and to determine, according to the position of the vehicle (1) on the traffic lane and the direction of the vehicle (1), an authorized or an unauthorized direction of movement of the vehicle (1), the control unit (10) being further configured to take a preventive measure of an accident-prone situation when it determines an unauthorized direction of movement of the vehicle (1).

2. Vehicle (1) according to claim 1, wherein the control unit (10) comprises software for detecting the "no entry" sign in the image, based on one or more pre-trained neural networks.

3. Vehicle (1) according to claim 1 or 2, further comprising a human-machine interface (14) and wherein the control unit (10) is configured to use the human-machine interface (14) so ​​as to warn a driver of the vehicle (1) of an unauthorized direction of travel of the vehicle (1) when such an unauthorized direction of travel is determined by the control unit (10).

4. Vehicle (1) according to any one of claims 1 to 3, further comprising an emergency service warning device (15) and wherein the control unit (10) is configured to detect, within a predetermined period following a determination of an unauthorized direction of travel of the vehicle (1), a decrease in speed (v) of the vehicle (1) in said direction, the control unit (10) being configured to activate the emergency service warning device (15) when such a decrease in speed (v) of the vehicle (1) in said direction is not detected within the predetermined period.

5. A vehicle (1) according to any one of claims 1 to 4, comprising a braking control device (16) and in which The control unit (10) is configured to detect, for a predetermined period following a determination of an unauthorized direction of travel of the vehicle (1), a decrease in speed (v) of the vehicle (1) in said direction, and to activate the braking control device (16) when such a decrease in speed (v) of the vehicle (1) in said direction is not detected for the predetermined period.

6. Vehicle (1) according to any one of claims 1 to 5, comprising a V2X communication device (18), and wherein the control unit (10) is configured, when it has determined an unauthorized direction of travel of the vehicle (1), to use the V2X communication device (18) so as to alert at least one other road user or road infrastructure.

7. Vehicle (1) according to any one of claims 1 to 6, wherein the map data (D_cart) includes a direction of travel indication on the traffic lane, and wherein the control unit (10) is configured to correlate said indication with a detected no-entry sign on the image and to determine an unauthorized direction of travel only if the direction of travel of said indication is compatible with the detected no-entry sign on the image.

8. A method for preventing (100) an accident-prone situation, implemented by a vehicle (1) according to any one of claims 1 to 7 and comprising the steps of: - capturing (102) an image in which a traffic sign appears, - detecting (104) a no entry sign on the image, - obtaining a positioning (106) and a direction of the vehicle (1), - associating (108) a position of the vehicle (1) provided by the step of obtaining a positioning (106), with a traffic lane, using map data (D-cart), - determining (110) an authorized or unauthorized direction of traffic as a function of the direction of the vehicle (1) and the position of the vehicle (1) on the traffic lane, and if an unauthorized direction is determined, - taking a measure to prevent an accident-prone situation.

9. A prevention method (100) according to claim 8, wherein when the determination step (110) results in the determination of an unauthorized direction of travel of the vehicle (1), the step of taking a preventive measure against an accident-prone situation consists of at least one of the following steps: - a visual and / or audible warning step (112) to a driver of the vehicle (1), - an alert step (116) to an emergency service when the speed (v) of the vehicle (1) does not decrease for a predetermined period after the determination of an unauthorized direction of travel of the vehicle (1), - a braking step (120) of the vehicle (1) when the speed (r) of the vehicle (1) does not decrease for a predetermined period after the determination of an unauthorized direction of travel of the vehicle (1),- an emergency parking step (120) of the vehicle (1) when the speed (v) of the vehicle (1) does not decrease for the predetermined time after the determination of an unauthorized direction of travel of the vehicle (1), - an alert step (112) of at least one other user of the traffic lane or road infrastructure.

10. Prevention method (100) according to claim 8 or 9, wherein the step of determining an authorized or unauthorized direction of travel is conditioned on: - a compatibility between the prohibited direction detected on the image and an indication of the direction of travel on the traffic lane from the cartographic data (D_cart) of the vehicle (1), and - an association of the traffic sign with the traffic lane.

Images