Improved detection of puddles on a roadway where a motor vehicle is traveling

Abstract

The invention relates to a method for detecting the presence of puddles on a roadway on which a motor vehicle (1) is traveling. The invention also relates to a computer device (100) implementing such a method, as well as a motor vehicle (1) comprising such a device (100). Figure for the abstract: 1

Description

Title of the invention: Improved detection of puddles on a roadway on which a motor vehicle is traveling. Technical field of the invention

[0001] The present invention relates to the field of safety systems for driving motor vehicles. The invention relates in particular to a method for detecting the presence of puddles on a roadway on which a motor vehicle is traveling. The invention also relates to a device implementing such a method, as well as a motor vehicle comprising such a device. The invention is applicable to motor vehicles such as motor land vehicles, in particular cars. Prior art

[0002] The safety of driving a motor vehicle depends in particular on the condition of the road surface on which the vehicle is traveling. Thus, a wet road surface, and especially the presence of puddles, significantly reduces tire grip and can therefore lead to a substantial decrease in driving safety. Therefore, in order to minimize the risk of potentially unfortunate consequences from such recurring situations, manufacturers have developed solutions to alert a driver when a puddle is present on the road ahead. These systems are generally based on the use of a camera mounted in the vehicle to capture images of the driving environment in front of the vehicle.However, these systems are generally unable to distinguish between puddles and potential mirages that can appear when certain conditions are favorable to this type of phenomenon. They are therefore prone to triggering false alerts, which can lead to decreased driver attention and consequently a decline in driving safety. Summary of the invention

[0003] The invention aims to overcome this inconvenience. In particular, it aims to provide a solution for detecting puddles on a roadway where a motor vehicle is traveling more accurately, specifically preventing confusion between a puddle and a mirage. Through this, the invention aims to contribute to improving the safety of driving a motor vehicle.

[0004] In order to achieve these goals, the invention relates, according to a first aspect, to a detection method, by a computer device embedded on board a motor vehicle, due to the presence of puddles on a roadway on which the vehicle is traveling, the procedure includes the following steps: i. obtain data characterizing at least one shot of the vehicle's driving environment; ii. determine data characterizing the road surface temperature and / or the polarization of light at a location in front of the vehicle; and iii. establish the presence of a puddle on the roadway based on the data obtained during step i) and the data determined during step ii).

[0005] According to one variant, step iii) may include the steps of: • feed an artificial neural network of an artificial intelligence module of said device with the data obtained during step i); • process data characterizing the detection of a puddle of water, which are produced by the artificial neural network based on the data determined during step ii) and at least one pre-established threshold value.

[0006] According to another embodiment, step iii) may consist of feeding an artificial neural network of an artificial intelligence module of said device with the data obtained during step i) and the data determined during step ii).

[0007] According to yet another variant, the artificial neural network can be based on the TwinLiteNet model.

[0008] According to yet another variant, the data characterizing the temperature of the pavement and / or the polarization of light at a place in front of the vehicle can be determined by means of a temperature probe and / or a camera equipped with a polarizing filter.

[0009] According to a second aspect, the invention relates to a device for detecting puddles of water on a roadway on which a motor vehicle is traveling, the device comprising at least one information processing unit, including at least one processor, and a data storage medium configured to implement a method as described above.

[0010] According to a third aspect, the invention relates to a computer program comprising program code instructions for the execution of the steps of a process as described above when said program is executed by at least one processor.

[0011] According to a fourth aspect, the invention relates to a medium usable in a computer on which a program as described above is recorded.

[0012] According to a fifth aspect, the invention relates to a motor vehicle which includes a device as described above. Brief description of the figures

[0013] Other features and advantages of the invention will become apparent from an examination of the detailed description below, and the accompanying figures, in which:

[0014] [Fig-1] is a diagram of a motor vehicle according to the invention;

[0015] [Fig.2] is a functional diagram of a device according to the invention; and

[0016] [Fig.3] is a flowchart of the steps of a process according to the invention. Detailed description of the invention

[0017] Figure 1 schematically illustrates a motor vehicle 1 according to the invention. This vehicle is equipped with a first camera 2 arranged in the vehicle to capture images of the driving environment located in front of the vehicle 1. In one embodiment, the vehicle 1 according to the invention also includes a second camera 3 mounted next to the first camera 2 to allow images to be captured in a synchronized manner with the first camera 2, and which is further equipped with a polarizing filter. Thanks to this second camera 3, it is advantageous to determine the polarization of light at a location in front of the vehicle 1. The vehicle 1 according to the invention is also equipped with a temperature sensor 4, which is preferably arranged on the front bumper, or alternatively under the housing of an exterior rearview mirror.Thanks to this probe 4, we can advantageously estimate the temperature of the road surface on which vehicle 1 is traveling.

[0018] Advantageously, the vehicle 1 according to the invention further incorporates a device 100 for detecting the presence of puddles on a roadway on which a motor vehicle within the meaning of the present invention is traveling, as described below, which implements a method for detecting the presence of puddles on a roadway on which a motor vehicle within the meaning of the present invention is traveling, as briefly summarized below and described in detail below.

[0019] During implementation of the method, the device 100 according to the invention continuously obtains images of the area in front of the vehicle 1, taken by the first camera 2. Simultaneously, by interacting with the temperature sensor 4 and / or the second camera 3, it determines the temperature of the road surface on which the vehicle is traveling and / or the polarization of the light in front of the vehicle 1. Finally, according to a first embodiment, it feeds an artificial neural network which it runs with the images, the artificial neural network having been previously trained to detect puddles in the images taken with a camera. It then uses the detection data produced by the artificial neural network and the road surface temperature and / or the polarization of the light to determine the presence of puddles based on temperature threshold values ​​or of pre-established polarizations. This is how it determines whether a detection produced by the neural network actually corresponds to a puddle of water or whether it is, on the contrary, a mirage.

[0020] According to another embodiment, the artificial neural network has been pre-trained to directly determine the presence of puddles based on photographs and road surface temperature values. In this case, the device 100 according to the invention feeds the artificial neural network with the photographs generated by the first camera 2 and the current road surface temperature values, which are determined using the temperature probe 4. The neural network outputs puddle detection data that is considered reliable, in other words, that avoids possible confusion between puddles and mirages.

[0021] This is how the device 100 according to the invention makes it possible to detect puddles of water on a road on which a motor vehicle is traveling more precisely, thus advantageously avoiding any possibility of confusion between puddles of water and mirages, and that it can therefore allow an improvement in driving safety.

[0022] Figure 2 schematically illustrates the device 100 for detecting the presence of puddles on a roadway on which a motor vehicle is traveling, as defined in the present invention. It is essentially a computer device comprising at least one information processing unit 101, including one or more processors, a data storage medium 102, on which is stored, in particular, a program comprising program code instructions for executing the steps of the method according to the invention described below, and an input / output interface 103 for receiving and transmitting data. Advantageously, this hardware is configured to run an artificial intelligence module 104, which uses, for example, an artificial neural network 105, preferably based on the TwinLiteNet model.

[0023] Preferably, the device 100 according to the invention is hosted on an independent computer and interacts via its input and output interface 103 and by means of a wired communication network of the vehicle (e.g. CAN, Ethernet, MOST) - materialized on [Fig. 1] by the double-direction arrows - with the first camera 2, the second camera 3 and the temperature probe 4. As a result, the device 100 according to the invention can, in particular, obtain data characterizing at least one shot of the driving environment of the vehicle 1, and it can determine data characterizing the temperature of the road surface and / or the polarization of light at a place in front of the vehicle.

[0024] According to the invention, all the elements described above contribute to enabling the implementation of a method for detecting the presence of puddles on a roadway on which a motor vehicle is traveling, as described below in relation to [Fig.3].

[0025] Figure 3 illustrates, by means of a flowchart, the steps of the process according to the invention.

[0026] According to a first step 301 of the method according to the invention, the device 100 according to the invention obtains data characterizing at least one view of the vehicle's driving environment. It interacts with the first camera 2 for this purpose. Thus, at the end of this first step, the device 100 according to the invention holds at least one view of the driving environment in front of the vehicle 1, for example, an image.

[0027] Next, according to a second step 302 of the method according to the invention, the device 100 according to the invention determines data characterizing the road surface temperature and / or the polarization of light at a location in front of the vehicle. In a first embodiment, the device 100 according to the invention interacts with the temperature probe 4 to obtain a temperature measurement, which it considers as an estimate of the road surface temperature on which the vehicle is traveling, or to which it applies a calibration factor to arrive at an estimated road surface temperature. In another embodiment, the device 100 according to the invention also interacts with the second camera 3 to determine, based on a measurement or a photograph, an estimated value of the light polarization at a location in front of the vehicle 1.

[0028] Finally, according to a third step 303 of the process according to the invention, the device 100 according to the invention establishes the presence of a puddle on the road on the basis of the data obtained during the first step 301 and the data determined during the second step 302 of the process.

[0029] More specifically, according to a first embodiment, the device 100 according to the invention proceeds as follows. It first feeds the artificial neural network 105 with the data obtained during the first step 301, in other words, with the image of the driving environment located in front of the vehicle 1. This embodiment is applied when the artificial neural network 105 has previously been trained to detect puddles based on a training dataset containing only images. In this case, the artificial neural network 105 outputs detection data that logically identifies the potential presence of a puddle in an image with which it is fed. The device 100 according to the invention then validates this detection produced by the artificial neural network 105 by determining whether the value The road surface temperature determined during the second step is either above or below a pre-established threshold value, for example, 30°C. If the road surface temperature exceeds the threshold value, the system determines that the detection produced by the artificial neural network corresponds to a mirage. Conversely, if the previously determined road surface temperature is below the threshold value, the system determines that the detection produced by the artificial neural network corresponds to a puddle.

[0030] Alternatively or cumulatively, the device 100 according to the invention validates a detection produced by the artificial neural network 105 using the polarization of light in front of the vehicle 1. Indeed, mirages generally form when the sun is high in the sky, which makes strong temperature gradients more likely. At the same time, for the light to be angled—or reflected off a puddle—toward the driver, its angle of incidence must be small, and the skylight will therefore come from a point close to the horizon. This combination of high sun and low-angle light means that the latter will have strong horizontal polarization (due to Rayleigh scattering). The polarization of the light will not change with a mirage, since it only involves refraction, but the light is significantly less polarized after reflection off a puddle.Therefore, in a scenario of strong sunlight and low angle of light, the light from a mirage perceived by the second camera 3 will be strongly filtered by a horizontal polarizing filter placed on the camera, while the light reflected by a puddle of water will be less attenuated by the same filter.

[0031] Thus, as an alternative or complement to the road surface temperature, the device 100 according to the invention validates a detection produced by the artificial neural network 105 using a light polarization value that was determined during the second step 302 of the process. When it combines the two variables, it establishes, for example, the presence of a puddle when the road surface temperature is below a threshold value and when the light polarization value is below a threshold value. Conversely, it establishes that a detection by the artificial neural network 105 corresponds to a mirage when the road surface temperature is above its threshold value or when the light polarization value is above its threshold value.When the device 100 according to the invention uses only the polarization of light, it proceeds in the same way as with the temperature of the pavement, establishing the presence of a puddle when the value of the polarization of light is less than a threshold value and, conversely, that a detection produced by the artificial neural network corresponds to a mirage when the value of the polarization of light exceeds its threshold value.

[0032] According to another embodiment, which is applied when the artificial neural network 105 has previously been trained to detect puddles on the basis of training data containing shots in combination with corresponding road temperature values, the device 100 according to the invention then proceeds by feeding the artificial neural network 105 with, jointly, the data obtained during the first step 301 of the process, i.e. at least one shot of the driving environment in front of the vehicle 1 taken by the first camera 2, and the data determined during the second step 302, in other words the road temperature value determined by means of the temperature probe 4.As output, the artificial neural network 105 produces detection data which the device 100 according to the invention can advantageously use without further processing in order to establish the presence of a puddle of water in front of the vehicle 1.

[0033] Thus, thanks to the method and device according to the invention described above, a solution exists for effectively detecting puddles on a roadway on which a motor vehicle is traveling, in particular by avoiding any possible confusion between puddles and mirages. In this way, the method and device according to the invention contribute to improved driving safety.

Claims

Demands

1. A method for detecting, by a computer device (100) mounted on board a motor vehicle (1), the presence of puddles on a roadway on which the vehicle is traveling, characterized in that the method comprises the steps of: i. obtaining data characterizing at least one image of the vehicle's driving environment; ii. determining data characterizing the temperature of the roadway and / or the polarization of light at a location in front of the vehicle; and iii. establishing the presence of a puddle on the roadway based on the data obtained during step i) and the data determined during step ii).

2. A method according to claim 1, characterized in that step iii) comprises the steps of: • feeding an artificial neural network (105) of an artificial intelligence module (104) of said device with the data obtained during step i); • processing data characterizing the detection of a puddle of water which are produced by the artificial neural network as a function of the data determined during step ii) and at least one pre-established threshold value.

3. A method according to claim 1, characterized in that step iii) consists of feeding an artificial neural network (105) of an artificial intelligence module (104) of said device with the data obtained during step i) and the data determined during step ii).

4. A method according to any one of claims 2 or 3, characterized in that the artificial neural network (105) is based on the TwinLiteNet model.

5. A method according to any one of the preceding claims, characterized in that the data characterizing the temperature of the pavement and / or the polarization of light at a place in front of the vehicle are determined by means of a temperature probe (4) and / or a camera (3) equipped with a polarizing filter.

6. Device (100) for detecting the presence of puddles on a roadway on which a motor vehicle (1) is traveling, characterized in that the device comprises at least one information processing unit (101), comprising at least one processor, and a data storage medium (102) configured to implement a method according to any one of the preceding claims.

7. Computer program comprising program code instructions for carrying out the steps of a process according to any one of claims 1-5 when said program is executed by at least one processor.

8. A medium usable in a computer, characterized in that a program according to claim 7 is stored thereon.

9. Motor vehicle (1), characterized in that the vehicle (1) comprises a device (100) according to claim 6.

Images