Road surface type determination method, device and vehicle

Abstract

The present application discloses a road surface type determination method, device and vehicle, belonging to the field of vehicle technology. It comprises: obtaining an image of the current road surface of the target vehicle, the current driving parameters of the target vehicle and the environmental parameters of the target vehicle's environment; determining the road surface limit adhesion coefficient of the current road surface based on the current driving parameters of the target vehicle; and determining the road surface type of the current road surface based on the image, the road surface limit adhesion coefficient and the environmental parameters. Since the road surface limit adhesion coefficient can indicate the maximum friction between the target vehicle's tires and the ground, the image and the environmental parameters of the target vehicle's environment can indicate the environmental information of the target vehicle's environment. In this way, the current road surface is subsequently determined by combining multiple features, so that the road surface type of the current road surface can be determined more accurately, ensuring that the vehicle's driving parameters can be adjusted in a timely manner, thereby ensuring driving safety.

Description

Technical Field

[0001] The present application relates to the field of vehicle technology, and in particular to a method, device and vehicle for determining a road surface type. Background Art

[0002] Generally speaking, vehicles face varying driving conditions on different road surfaces, resulting in varying driving performance. Therefore, different driving parameters must be controlled to adapt the vehicle to the specific surface and improve driving safety. For example, when driving on snowy surfaces, parameters such as output torque can be controlled to improve driving safety.

[0003] However, in order to ensure that the driving parameters of the vehicle on the corresponding road can be adjusted in time, it is necessary to accurately determine the road type of the road the vehicle is on. Therefore, accurate determination of the road type is very important. In this case, a method for determining the road type is urgently needed. Summary of the Invention

[0004] This application provides a road surface type determination method, device, vehicle, and storage medium that can accurately determine the road surface type of the vehicle, thereby ensuring that the vehicle's driving parameters can be adjusted in a timely manner, thereby ensuring driving safety. The technical solution is as follows:

[0005] In a first aspect, a method for determining a road surface type is provided, the method comprising:

[0006] Acquire an image of a current driving road surface of a target vehicle, current driving parameters of the target vehicle, and environmental parameters of an environment in which the target vehicle is located;

[0007] determining a road surface limit adhesion coefficient of the current driving road surface based on the current driving parameters of the target vehicle;

[0008] The road type of the current driving road is determined based on the image, the road surface limit adhesion coefficient, and the environmental parameters.

[0009] In the present application, an image of the current road surface on which the target vehicle is traveling, current driving parameters, and environmental parameters of the target vehicle's environment are first obtained, that is, the driving parameters of the target vehicle and environmental information of the target vehicle's environment are obtained. Then, based on the current driving parameters of the target vehicle, the road surface limit adhesion coefficient of the current road surface is determined. Subsequently, based on the image, the road surface limit adhesion coefficient, and the environmental parameters, the road surface type of the current road surface is determined. Since the road surface limit adhesion coefficient can indicate the maximum friction between the target vehicle's tires and the ground, the image and the environmental parameters of the target vehicle's environment can indicate environmental information of the target vehicle's environment. In this way, the current road surface is subsequently determined by combining various features, so that the road surface type of the current road surface can be determined more accurately, ensuring that the vehicle's driving parameters can be adjusted in a timely manner, thereby ensuring driving safety.

[0010] Optionally, before acquiring the image of the current driving road of the target vehicle, the current driving parameters of the target vehicle, and the environmental parameters of the target vehicle, the method further includes:

[0011] Obtaining the current speed, current wheel angle, and wheelbase of the target vehicle;

[0012] Determining a current image acquisition range based on a current speed, a current wheel angle, and a wheelbase of the target vehicle;

[0013] The method for obtaining an image of a current road surface on which a target vehicle is traveling comprises:

[0014] According to the current image acquisition range, an image of the current driving road of the target vehicle is acquired.

[0015] Optionally, the current driving parameters include the acceleration of the target vehicle, the speed of the target vehicle, and the wheel speeds of multiple wheels of the target vehicle, and determining the road surface limit adhesion coefficient of the current driving road surface based on the current driving parameters of the target vehicle includes:

[0016] Dividing the acceleration of the target vehicle by the acceleration of gravity to obtain the actual adhesion coefficient of the current driving road surface at the current moment;

[0017] For any one of the plurality of wheels of the target vehicle, determining a slip ratio of the wheel based on a speed of the target vehicle and a wheel speed of the wheel;

[0018] The road surface limit adhesion coefficient is determined based on the slip ratios of the plurality of wheels of the target vehicle and the actual adhesion coefficient.

[0019] Optionally, determining the road surface limit adhesion coefficient based on the slip rates of the multiple wheels of the target vehicle and the actual adhesion coefficient includes:

[0020] When the number of wheels among the plurality of wheels whose slip rates are greater than or equal to a preset slip rate threshold is greater than or equal to a preset number, determining the actual adhesion coefficient as the road surface limit adhesion coefficient;

[0021] When the number of wheels among the multiple wheels whose slip rate is greater than or equal to the preset slip rate threshold is less than the preset number, the maximum value between the actual adhesion coefficient and the target adhesion coefficient is determined as the road surface limit adhesion coefficient, and the target adhesion coefficient is the road surface limit adhesion coefficient at the previous moment.

[0022] Optionally, the determining the road type of the current road surface on which the target vehicle is traveling based on the image, the road surface limit adhesion coefficient, and the environmental parameter includes:

[0023] Classifying the image to obtain a classification result and a confidence level of the image;

[0024] When the classification result is a snowy road surface, the confidence of the classification result is greater than a preset confidence threshold, the road surface limit adhesion coefficient is less than a preset adhesion coefficient threshold, and the environmental parameters meet the target conditions, it is determined that the current driving road surface is a snowy road surface.

[0025] Optionally, the environmental parameter includes at least one of temperature, humidity, and snow amount, and the method for determining whether the environmental parameter meets the target condition includes at least one of the following:

[0026] collecting the temperature of the environment in which the target vehicle is located through a temperature sensor of the target vehicle; if the temperature is less than a preset temperature threshold, determining that the environmental parameter meets the target condition; if the temperature is greater than or equal to the preset temperature threshold, determining that the environmental parameter does not meet the target condition;

[0027] collecting the humidity of the environment in which the target vehicle is located by a humidity sensor of the target vehicle; if the humidity is greater than a preset humidity threshold, determining that the environmental parameter meets the target condition; if the humidity is less than or equal to the preset humidity threshold, determining that the environmental parameter does not meet the target condition;

[0028] The amount of snow in the environment where the target vehicle is located is collected through the snow sensor of the target vehicle; when the amount of snow is greater than a preset snow amount threshold, it is determined that the environmental parameters meet the target conditions; when the amount of snow is less than or equal to the preset snow amount threshold, it is determined that the environmental parameters do not meet the target conditions.

[0029] Optionally, classifying the image to obtain a classification result and a confidence level includes:

[0030] Performing feature extraction on the image to obtain a first image feature of the image;

[0031] activating and pooling the first image features to obtain second image features;

[0032] Performing full connection and normalization on the second image features to obtain confidence levels of multiple predicted classification results, where the confidence levels are used to indicate the credibility of the corresponding predicted classification results;

[0033] The predicted classification result with the greatest confidence among the multiple predicted classification results is determined as the classification result of the image.

[0034] Optionally, after determining the road type of the current driving road based on the image, the road surface limit adhesion coefficient, and the environmental parameters, the method further includes:

[0035] When it is determined that the current driving road surface is a snowy road surface, the target vehicle is controlled to drive in a snow mode.

[0036] In a second aspect, a road surface type determination device is provided, the device comprising:

[0037] A first acquisition module is used to acquire an image of the current road surface on which the target vehicle is traveling, current driving parameters of the target vehicle, and environmental parameters of the environment in which the target vehicle is located;

[0038] A first determining module is configured to determine a road surface limit adhesion coefficient of the current driving road surface based on the current driving parameters of the target vehicle;

[0039] The second determination module is configured to determine the road type of the current driving road based on the image, the road surface limit adhesion coefficient, and the environmental parameters.

[0040] Optionally, the device further comprises:

[0041] A second acquisition module is used to obtain the current speed, current wheel angle and wheelbase of the target vehicle;

[0042] a third determining module, configured to determine a current image acquisition range based on a current speed of the target vehicle, a current wheel angle, and a wheelbase of the target vehicle;

[0043] The first acquisition module is used for:

[0044] According to the current image acquisition range, an image of the current driving road of the target vehicle is acquired.

[0045] Optionally, the current driving parameters include the acceleration of the target vehicle, the speed of the target vehicle, and the wheel speeds of multiple wheels of the target vehicle, and the first determining module is configured to:

[0046] Dividing the acceleration of the target vehicle by the acceleration of gravity to obtain the actual adhesion coefficient of the current driving road surface at the current moment;

[0047] For any one of the plurality of wheels of the target vehicle, determining a slip ratio of the wheel based on a speed of the target vehicle and a wheel speed of the wheel;

[0048] The road surface limit adhesion coefficient is determined based on the slip ratios of the plurality of wheels of the target vehicle and the actual adhesion coefficient.

[0049] Optionally, the first determining module is configured to:

[0050] When the number of wheels among the plurality of wheels whose slip rates are greater than or equal to a preset slip rate threshold is greater than or equal to a preset number, determining the actual adhesion coefficient as the road surface limit adhesion coefficient;

[0051] When the number of wheels among the multiple wheels whose slip rate is greater than or equal to the preset slip rate threshold is less than the preset number, the maximum value between the actual adhesion coefficient and the target road surface limit adhesion coefficient is determined as the road surface limit adhesion coefficient, and the target adhesion coefficient is the road surface limit adhesion coefficient at the previous moment.

[0052] Optionally, the second determining module is configured to:

[0053] Classifying the image to obtain a classification result and a confidence level of the image;

[0054] When the classification result is a snowy road surface, the confidence of the classification result is greater than a preset confidence threshold, the road surface limit adhesion coefficient is less than a preset adhesion coefficient threshold, and the environmental parameters meet the target conditions, it is determined that the current driving road surface is a snowy road surface.

[0055] Optionally, the environmental parameter includes at least one of temperature, humidity, and snow amount, and the second determination module is configured to:

[0056] collecting the temperature of the environment in which the target vehicle is located through a temperature sensor of the target vehicle; if the temperature is less than a preset temperature threshold, determining that the environmental parameter meets the target condition; if the temperature is greater than or equal to the preset temperature threshold, determining that the environmental parameter does not meet the target condition;

[0057] collecting the humidity of the environment in which the target vehicle is located by a humidity sensor of the target vehicle; if the humidity is greater than a preset humidity threshold, determining that the environmental parameter meets the target condition; if the humidity is less than or equal to the preset humidity threshold, determining that the environmental parameter does not meet the target condition;

[0058] The amount of snow in the environment where the target vehicle is located is collected through the snow sensor of the target vehicle; when the amount of snow is greater than a preset snow amount threshold, it is determined that the environmental parameters meet the target conditions; when the amount of snow is less than or equal to the preset snow amount threshold, it is determined that the environmental parameters do not meet the target conditions.

[0059] Optionally, the second determining module is configured to:

[0060] Performing feature extraction on the image to obtain a first image feature of the image;

[0061] activating and pooling the first image features to obtain second image features;

[0062] Performing full connection and normalization on the second image features to obtain confidence levels of multiple predicted classification results, where the confidence levels are used to indicate the credibility of the corresponding predicted classification results;

[0063] The predicted classification result with the greatest confidence among the multiple predicted classification results is determined as the classification result of the image.

[0064] Optionally, the device further comprises:

[0065] The control module is used to control the target vehicle to travel in a snow mode when it is determined that the current driving road surface is a snow road surface.

[0066] In a third aspect, a vehicle is provided, comprising:

[0067] a memory for storing executable program code;

[0068] The processor is used to call and run the executable program code from the memory, so that the vehicle executes the above-mentioned road surface type determination method.

[0069] In a fourth aspect, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the above-mentioned road surface type determination method is implemented.

[0070] In a fifth aspect, a computer program product comprising instructions is provided, which, when executed on a computer, enables the computer to execute the steps of the above-mentioned road surface type determination method.

[0071] It can be understood that the beneficial effects of the second, third, fourth and fifth aspects mentioned above can be found in the relevant description of the first aspect mentioned above, and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0073] Figure 1 This is a schematic diagram of a scenario of a road surface type determination method provided in an embodiment of the present application;

[0074] Figure 2 This is a flow chart of a method for determining a road surface type provided in an embodiment of the present application;

[0075] Figure 3 This is a schematic diagram of determining the current image acquisition range provided by an embodiment of the present application;

[0076] Figure 4 This is a schematic diagram of a maximum pooling process provided by an embodiment of the present application;

[0077] Figure 5 This is a schematic diagram of the structure of a residual block provided in an embodiment of the present application;

[0078] Figure 6 This is a schematic diagram of the structure of a road surface classification model provided in an embodiment of the present application;

[0079] Figure 7 This is a schematic diagram of the structure of another residual block provided in an embodiment of the present application;

[0080] Figure 8 is a schematic structural diagram of a road surface type determination device provided in an embodiment of the present application;

[0081] Figure 9 It is a structural schematic diagram of a vehicle provided in an embodiment of the present application. DETAILED DESCRIPTION

[0082] In order to make the objectives, technical solutions and advantages of this application clearer, the implementation methods of this application will be further described in detail below with reference to the accompanying drawings.

[0083] It should be understood that the “multiple” mentioned in this application refers to two or more. In the description of this application, unless otherwise specified, “ / ” means or, for example, A / B can mean A or B; “and / or” in this article is merely a description of the association relationship of associated objects, indicating that there can be three relationships, for example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, in order to facilitate the clear description of the technical solution of this application, words such as “first” and “second” are used to distinguish between identical or similar items with basically the same functions and effects. Those skilled in the art can understand that words such as “first” and “second” do not limit the quantity and execution order, and words such as “first” and “second” do not necessarily limit them to be different.

[0084] Before explaining the road surface type determination method provided in the embodiment of the present application, the application scenario of the embodiment of the present application is first explained.

[0085] Vehicles face different driving conditions when driving on different road surfaces. For example, when driving on snowy roads, the grip of the vehicle tires will be reduced, causing the vehicle to slip. In this case, the vehicle will lose control, which will put the vehicle driving in danger.

[0086] Therefore, when the vehicle is on a snowy road, the vehicle's driving parameters can be adjusted, such as reducing output torque, keeping the electronic stability program on, etc., so that the vehicle can smoothly pass through the snowy road and ensure safe driving. In order to ensure that the vehicle's driving parameters can be adjusted in a timely manner on snowy roads, it is necessary to promptly and accurately determine that the vehicle is currently driving on a snowy road.

[0087] The following introduces the relevant technologies by taking the vehicle driving on a snowy road as an example.

[0088] One related technology is to determine whether the vehicle is currently driving on a snowy road by observing the surrounding environment. If the vehicle is currently driving on a snowy road, the driver can make corresponding vehicle controls, such as reducing the speed, to ensure safe driving on the snowy road.

[0089] The above method relies on the driver's observation to determine whether the vehicle is currently driving on a snowy road, and actively controls the vehicle to reduce speed when the driver determines that the vehicle is driving on a snowy road. However, this method only allows for manual control of the vehicle, and if the driver is not paying attention, it may cause the vehicle to be dangerous.

[0090] Another method is to determine whether the vehicle is traveling on a snowy road using a single feature. For example, the friction coefficient between the vehicle and the current road surface can be obtained first. If the friction coefficient is determined to be less than a friction coefficient threshold, it can be determined that the vehicle is currently traveling on a snowy road.

[0091] In the above method, whether the current road surface is snowy is determined by the friction coefficient between the vehicle and the current road surface. However, if other substances exist on the current road surface and affect the determination of the friction coefficient between the vehicle and the current road surface, the final determination result will be inaccurate.

[0092] To this end, an embodiment of the present application provides a road surface type determination method, which can be applied to a scenario of determining the road surface type of a vehicle's driving road.

[0093] For example, the road type determination method can be applied to a scenario in which it is determined whether the road on which the vehicle is currently traveling is a snowy road.

[0094] For example, Figure 1 This is a scenario diagram of a road surface type determination method, see Figure 1 , Figure 1 The vehicle 101 and the road surface 102 are included. The vehicle 101 is traveling on the road surface 102.

[0095] Specifically, an image of vehicle 101 traveling on road 102, the current acceleration of vehicle 101, and the ambient temperature of vehicle 101's current environment are first acquired. The ultimate adhesion coefficient of road surface 102 is then determined based on vehicle 101's current acceleration. The image of vehicle 101 traveling on road surface 102 is then recognized to determine whether road surface 102 is snowy. Based on the recognition result of the image, the ultimate adhesion coefficient of road surface 102, and the ambient temperature of the current environment, it is determined whether road surface 102 currently traveled by vehicle 101 is snowy. By subsequently combining multiple features to determine the current road surface, the determination of whether the current road surface is snowy can be made more accurately, ensuring that the vehicle's driving parameters can be adjusted promptly, thereby ensuring driving safety.

[0096] The road surface type determination method provided in the embodiment of the present application is explained in detail below.

[0097] Figure 2 This is a flow chart of a method for determining a road surface type provided by an embodiment of the present application. This method can be applied to a vehicle controller, for example, to a vehicle ECU (Electronic Control Unit). Figure 2 , the method includes the following steps.

[0098] Step 201: Acquire an image of the current road surface on which the target vehicle is traveling, current driving parameters of the target vehicle, and environmental parameters of the environment in which the target vehicle is located.

[0099] The current driving road surface refers to the road surface where the target vehicle is currently located. The image of the current driving road surface can be collected by the camera of the target vehicle.

[0100] The current driving parameters of the target vehicle refer to the driving parameters of the target vehicle at the current moment. Optionally, the current driving parameters of the target vehicle may include at least one of the target vehicle's speed, acceleration, and wheel speeds of multiple wheels of the target vehicle, etc., which is not limited in this embodiment of the present application.

[0101] The environmental parameters of the target vehicle's environment are used to indicate the environmental state of the target vehicle's environment. Optionally, the environmental parameters of the target vehicle's environment may include at least one of ambient temperature, ambient humidity, and snowfall, etc., which is not limited in this embodiment of the present application. Optionally, the environmental parameters of the target vehicle's environment may be measured by an environmental sensor, and the controller may then obtain the environmental parameters collected by the environmental sensor.

[0102] Since the image and the environmental parameters of the target vehicle's environment can represent the environmental information of the target vehicle's environment, in this case, it is equivalent to obtaining the target vehicle's current driving parameters and the environmental information of its environment, that is, obtaining multiple parameter information, which can be used to subsequently determine the road type of the current driving road.

[0103] Optionally, before step 201 , the current speed, current wheel angle and wheelbase of the target vehicle may be obtained first; based on the current speed, current wheel angle and wheelbase of the target vehicle, the current image acquisition range may be determined.

[0104] The current wheel angle refers to the angle of the front wheels of the target vehicle at the current moment. Optionally, the current wheel angle can be obtained by a wheel angle sensor. That is, the wheel angle sensor can detect the angle of the front wheels of the target vehicle. The controller then obtains the front wheel angle detected by the angle sensor and uses it as the current wheel angle.

[0105] The wheelbase refers to the distance between the front and rear axles of the target vehicle, that is, the distance between the center of the front wheels and the center of the rear wheels of the target vehicle. The current image acquisition range refers to the range of the current road surface that the target vehicle is currently traveling on.

[0106] Because the target vehicle is constantly in motion, determining the road surface type takes time. Therefore, by defining a current image acquisition range, we can ensure that the current road surface determination result can be obtained before the target vehicle reaches the road area indicated by the captured image, thereby ensuring that the current road surface determination result is valid.

[0107] Among them, the operation of determining the current image acquisition range based on the current speed, current wheel angle and wheelbase of the target vehicle can be: based on the current speed of the target vehicle, determining the horizontal minimum value and the horizontal maximum value in the current image acquisition range; based on the current wheel angle and the wheelbase, determining the vertical minimum value and the vertical maximum value in the current image acquisition range.

[0108] The horizontal minimum value refers to the horizontal distance between the left boundary of the road area to be collected and the camera of the target vehicle. The horizontal maximum value refers to the horizontal distance between the right boundary of the road area to be collected and the camera of the target vehicle.

[0109] The vertical minimum value refers to the vertical distance between the upper boundary of the road surface area to be collected and the center axis of the target vehicle. The vertical maximum value refers to the vertical distance between the lower boundary of the road surface area to be collected and the center axis of the target vehicle.

[0110] Among them, based on the current speed of the target vehicle, the operation of determining the horizontal minimum value and the horizontal maximum value in the current image acquisition range can be: based on the current speed of the target vehicle, determine the corresponding horizontal minimum value from the corresponding relationship between the speed and the horizontal minimum value; add the horizontal minimum value to the target value to obtain the horizontal maximum value.

[0111] The target value can be pre-set and can be adjusted by the technician according to actual needs. For example, the target value can be set to 10m (meters), the horizontal minimum value is 20m, and the horizontal maximum value can be 30m.

[0112] The correspondence between the vehicle speed and the horizontal minimum value may include multiple vehicle speed ranges and multiple horizontal minimum values. The multiple vehicle speed ranges correspond one-to-one with the multiple horizontal minimum values, that is, one vehicle speed range corresponds to one horizontal minimum value. In this case, based on the current speed of the target vehicle, the speed range to which the current speed belongs can be determined, and the horizontal minimum value corresponding to this speed range can be obtained.

[0113] For example, Table 1 shows the correspondence between the vehicle speed and the horizontal minimum value. As shown in Table 1, Table 1 includes three speed ranges and three horizontal minimum values. For example, if the current speed of the target vehicle is 100 km / h (kilometers per hour), then Table 1 shows that the current speed of the target vehicle belongs to the third speed range [80, 120), and thus the horizontal minimum value corresponding to the third speed range can be obtained as 20 meters.

[0114] Table 1

[0115] Speed range Horizontal minimum [0，40) 5 [40，80) 10 [80，120) 20 …… ……

[0116] The embodiment of the present application only uses the above Table 1 as an example to illustrate the corresponding relationship between the vehicle speed and the horizontal minimum value, and does not constitute a limitation to the embodiment of the present application.

[0117] Among them, the operation of determining the vertical minimum value and the vertical maximum value in the current image acquisition range based on the current wheel angle and the wheelbase can be: based on the current wheel angle and the wheelbase, determine the vertical minimum value in the current image acquisition range by the following formula (1), and determine the vertical maximum value in the current image acquisition range by the following formula (2).

[0118] Y min =1.75+50δ / L (I)

[0119] Y mm =-1.75+50δ / L (2)

[0120] Among them, Y min is the vertical minimum, Y max is the vertical maximum value, δ is the current wheel angle, and L is the wheelbase. The negative sign in formula (2) is used to indicate the direction, that is, to indicate that the center axis of the target vehicle points to the direction of the lower boundary.

[0121] For example, Figure 3 It is a schematic diagram to determine the current image acquisition range, see Figure 3 , Figure 3 The target vehicle 301 and the road area 302 determined according to the current image acquisition range are included, and the target vehicle 301 includes a camera 303. Figure 3 As shown, the current image acquisition range is [(X min , X max ), (Y min , Y max )].

[0122] Among them, X min is the horizontal minimum value of the current image acquisition range, that is, the distance between the left boundary of the road area 302 and the camera 303, X maxis the horizontal maximum value of the current image acquisition range, that is, the distance between the right boundary of the road area 302 and the camera 303. The difference between the horizontal minimum value and the horizontal maximum value is the target value. Figure 3 The target value is 10. min Y is the distance between the upper boundary of the road area 302 and the center axis of the target vehicle 301. max It refers to the distance between the lower boundary of the road area 302 and the center axis of the target vehicle 301. Figure 3 The current wheel angle of the target vehicle 301 is 0, so the Y min and Y max Both are 1.75.

[0123] It is worth noting that after the current image acquisition range is determined, an image of the current road surface on which the target vehicle is traveling may be acquired according to the current image acquisition range.

[0124] Specifically, the camera's acquisition angle can be adjusted according to the current image acquisition range, and then the camera can be controlled to acquire road images at the adjusted acquisition angle, that is, images of the road area determined by the current image acquisition range can be acquired.

[0125] In this case, by acquiring an image of the current driving road surface of the target vehicle according to the current image acquisition range, it can be ensured that the current driving road surface can be determined before the target vehicle drives to the road surface area indicated by the acquired image, thereby ensuring that the driving parameters of the target vehicle can be adjusted in time when the target vehicle drives to the corresponding road surface, thereby ensuring the safety of the target vehicle's driving.

[0126] Step 202: Based on the current driving parameters of the target vehicle, determine the road surface limit adhesion coefficient of the current driving road surface.

[0127] The road's maximum adhesion coefficient indicates the maximum friction between the target vehicle's wheels and the road surface. For example, a target vehicle has less grip on snowy roads, so the friction coefficient is also lower. Therefore, a smaller road's maximum adhesion coefficient indicates less maximum friction between the target vehicle's wheels and the road surface, and a higher likelihood that the road surface is snowy. A larger road's maximum adhesion coefficient indicates greater maximum friction between the target vehicle's wheels and the road surface, and a lower likelihood that the road surface is snowy.

[0128] In this case, determining the road surface limit adhesion coefficient of the current driving road surface can provide an effective reference for subsequently determining the road surface type of the current driving road surface, so that the road surface type of the current driving road surface can be determined more accurately by referring to the road surface limit adhesion coefficient.

[0129] In the case that the current driving parameters may include the acceleration, speed, and wheel speeds of multiple wheels of the target vehicle, the operation of step 202 may include the following steps (1) to (3).

[0130] (1) Divide the acceleration of the target vehicle by the acceleration due to gravity to obtain the actual adhesion coefficient of the current road surface at the current moment.

[0131] The target vehicle's acceleration is the target vehicle's current longitudinal acceleration, which can be acquired by an acceleration sensor mounted on the target vehicle. Of course, it can also be calculated based on the target vehicle's speed, though this is not a limitation in the present embodiment. The actual adhesion coefficient represents the road surface adhesion coefficient of the current road surface at the current moment, and serves as a reference for determining the road surface's limit adhesion coefficient.

[0132] When the target vehicle is traveling on the current driving road surface, the current driving road surface can provide the target vehicle with a driving force for moving forward. Specifically, the calculation of the driving force provided by the current driving road surface to the target vehicle can be implemented by the following formula (3).

[0133] F 驱 =u 实际 ×M×g (3)

[0134] Among them, F 驱 is the driving force provided by the current road surface to the target vehicle, u 实际 is the actual adhesion coefficient of the current road surface. M is the weight of the target vehicle, and g is the acceleration due to gravity. Generally, the value of g is 9.8.

[0135] In addition, the driving force of the target vehicle can also be determined by the vehicle longitudinal travel equation. The process of determining the driving force of the target vehicle by the vehicle longitudinal travel equation is shown in the following formula (4).

[0136]

[0137] Among them, A x is the acceleration of the target vehicle, ρ is the air density, C d is the drag coefficient, S 面 is the frontal area, V 车 is the speed of the target vehicle, and θ is the slope of the current driving road.

[0138] Because the transmission ratio of the target vehicle's drive system is smaller when traveling at high speeds than when traveling at low speeds, and the torque output by the target vehicle's engine or motor is also lower than when traveling at low speeds, the torque output by the engine or motor will not cause the wheels to slip and become unstable. Therefore, the road surface type determination method provided in the embodiments of this application is primarily used in scenarios where the target vehicle is traveling at low speeds.

[0139] In addition, when the target vehicle is traveling at a low speed, the air resistance it encounters is very small, so the effect of air resistance on the longitudinal travel of the vehicle can be ignored. Then the drag coefficient C in the above formula (4) is d can be 0. Furthermore, the acceleration of the target vehicle is collected by an acceleration sensor installed on the target vehicle. Since the acceleration sensor is usually fixed to the vehicle body, the acceleration detected by it actually takes into account the slope resistance information. Therefore, the calculation of the slope resistance in the above formula (4) can be ignored.

[0140] Therefore, the operation of step (1) can be achieved by the following formula (5).

[0141]

[0142] For example, if the acceleration of the target vehicle is 9.8, then the actual adhesion coefficient of the current road surface is calculated to be u 实际 =9.8÷9.8=1.

[0143] (2) For any one wheel among the plurality of wheels of the target vehicle, the slip ratio of the wheel is determined based on the speed of the target vehicle and the wheel speed of the wheel.

[0144] The speed of the target vehicle refers to the longitudinal speed of the target vehicle at the current moment, and the wheel speed of this wheel refers to the rotation speed of this wheel at the current moment.

[0145] The slip rate of a wheel refers to the proportion of wheel slip during the movement of the wheel. The slip rate is used to indicate the slip generated between the tire tread and the road surface when the wheel brakes or accelerates while moving straight.

[0146] By performing the above step (2) on each of the multiple wheels of the target vehicle, the slip ratio of each of the multiple wheels of the target vehicle can be obtained. In this case, by determining the slip ratios of the multiple wheels, it can be known whether the multiple wheels of the target vehicle will slip when braking or accelerating, thereby determining whether the wheels of the target vehicle will slip during driving, and further based on this, a more accurate limit adhesion coefficient of the road surface can be determined.

[0147] Specifically, the operation of step (2) may be: for any one wheel among the multiple wheels of the target vehicle, based on the speed of the target vehicle and the wheel speed of the wheel, determine the slip rate of the wheel by the following formula (6).

[0148]

[0149] Among them, s is the slip rate of this wheel, V 车 is the speed of the target vehicle, V 轮 is the wheel speed of this wheel.

[0150] Taking the calculation of the slip rate of the left front wheel of the target vehicle as an example, the target vehicle speed is 80, and the wheel speed of the left front wheel of the target vehicle is 120, then the above formula can be obtained The slip rate of the left front wheel of the target vehicle is -0.5.

[0151] Optionally, wheel speed sensors are installed on multiple wheels of the target vehicle. In this case, the wheel speeds of the multiple wheels of the target vehicle can be detected by the corresponding wheel speed sensors.

[0152] Optionally, the vehicle speed may be determined in any of the following ways.

[0153] Example 1: A speed sensor may be installed on the target vehicle, and the speed of the target vehicle may be acquired by the speed sensor.

[0154] Example 2: The target vehicle's speed can also be calculated by wheel speed conversion. Optionally, methods for calculating vehicle speed based on wheel speed include the average wheel speed method and the maximum wheel speed method. The average wheel speed method uses the average wheel speed (wheel speed) of the target vehicle's two rear wheels as the vehicle speed; the maximum wheel speed method uses the maximum wheel speed of the target vehicle's four wheels as the vehicle speed.

[0155] In Example 3, the target vehicle's speed can also be obtained using another wheel speed conversion method. The specific calculation process is: Vehicle Speed = Wheel Circumference * Wheel Speed. Wheel speed can be obtained using a wheel speed sensor, and wheel circumference is an inherent parameter of the tire.

[0156] It should be understood that the above methods are only illustrative examples, and the calculated speeds are not very different and can all be used as the actual speed of the target vehicle. Any method for calculating the speed falls within the scope of protection of this application.

[0157] (3) The road surface limit adhesion coefficient is determined based on the slip ratios of the plurality of wheels of the target vehicle and the actual adhesion coefficient.

[0158] Since the slip ratio can indicate whether the wheels of the target vehicle are slipping, a more accurate limit adhesion coefficient of the road surface can be determined by subsequently combining the slip ratios of multiple wheels of the target vehicle with the actual adhesion coefficient.

[0159] Specifically, the operation of step (3) may be: when the number of wheels among the multiple wheels whose slip rate is greater than or equal to the preset slip rate threshold is greater than or equal to a preset number, the actual adhesion coefficient is determined as the road surface limit adhesion coefficient; when the number of wheels among the multiple wheels whose slip rate is greater than or equal to the preset slip rate threshold is less than a preset number, the maximum value between the actual adhesion coefficient and the target adhesion coefficient is determined as the road surface limit adhesion coefficient.

[0160] The target adhesion coefficient is the road surface limit adhesion coefficient at the previous moment, which is determined based on the driving parameters of the target vehicle at the previous moment.

[0161] The preset number can be set in advance, and the preset number can be set according to actual needs. For example, the preset number can be set to be larger, for example, the preset number can be set to 2.

[0162] The preset slip ratio threshold value can be set in advance, and the preset slip ratio can be set according to the slip ratio when the wheel slips. The preset slip ratio can be set to be relatively large.

[0163] In this case, if the slip rate of a wheel is greater than the preset slip rate threshold, it means that the slip rate of this wheel is large and has exceeded the slip rate when the wheel slips, which means that this wheel is likely to have slipped. If the slip rate of a wheel is less than the preset slip rate threshold, it means that the slip rate of this wheel is not large and has not exceeded the slip rate when the wheel slips, which means that this wheel is likely not to have slipped.

[0164] If the number of wheels among the multiple wheels whose slip rate is greater than or equal to the preset slip rate threshold is greater than or equal to the preset number, it means that a large number of wheels among the multiple wheels are slipping, that is, the target vehicle has a large number of slipping wheels on the current driving road surface, which means that the actual adhesion coefficient calculated at the current moment is the limit adhesion coefficient of the current driving road surface, and the actual adhesion coefficient calculated at the current moment can be determined as the road surface limit adhesion coefficient of the current driving road surface.

[0165] When the number of wheels among the multiple wheels whose slip rate is greater than or equal to the preset slip rate threshold is less than the preset number, it means that the number of wheels that slip among the multiple wheels is not large, that is, the target vehicle has not many wheels that slip on the current driving road surface. In this case, the calculated road surface limit adhesion coefficient will be larger, and the maximum value between the actual adhesion coefficient and the target road surface limit adhesion coefficient can be determined as the road surface limit adhesion coefficient.

[0166] For example, the preset slip rate threshold is 0.2, the preset number is 2, and the actual adhesion coefficient is 0.3. The slip rates of the four wheels of the target vehicle are 0.1, 0.25, 0.3, and 0.05, respectively. Since the slip rate of one wheel (0.1) is less than the preset slip rate threshold (0.2), and the slip rate of another wheel (0.05) is less than the preset slip rate threshold (0.2), the number of slip rates of the four wheels exceeding the preset slip rate threshold exceeds 2. Therefore, the actual adhesion coefficient (0.3) can be determined as the road surface limit adhesion coefficient, and the road surface limit adhesion coefficient is 0.3.

[0167] Step 203: Determine the road type of the current driving road based on the image, the road surface limit adhesion coefficient, and the environmental parameters.

[0168] The road surface type may include snowy roads, including snow-covered roads and icy roads. Based on the image, the road surface's ultimate adhesion coefficient, and the environmental parameters, it can be determined whether the current driving road surface is a snowy road. Of course, the road surface type may also include other roads with lower adhesion coefficients, such as wet and slippery roads, though this embodiment of the present application does not limit this. Wet and slippery roads may include rainy and wet roads.

[0169] Because the road surface's ultimate adhesion coefficient can indicate the friction between the target vehicle's tires and the ground, the image and the environmental parameters of the target vehicle's environment can indicate environmental information about the target vehicle's environment. Therefore, the road surface type of the current driving surface can be determined based on the image, the road surface's ultimate adhesion coefficient, and the environmental parameters.

[0170] In this case, the subsequent process is to determine the current road surface by combining multiple features, so as to more accurately determine the road type of the current road surface, ensuring that the vehicle's driving parameters can be adjusted in a timely manner, thereby ensuring driving safety.

[0171] Specifically, the operation of step 203 may include the following steps (1) to (3).

[0172] (1) Classify the image and obtain the classification result and confidence level of the image.

[0173] The classification result may include road surface types, such as snowy roads, slippery roads, etc. The confidence level is the confidence level of the classification result, which indicates the degree of trustworthiness of the classification result. The greater the confidence level, the more trustworthy the classification result, i.e., the more accurate the classification result. The smaller the confidence level, the less trustworthy the classification result, i.e., the less accurate the classification result.

[0174] In this case, by classifying the image, the current road surface type can be preliminarily determined, thereby providing a reference value for subsequently accurately determining the current road surface type. For example, by classifying the image, it can be preliminarily determined whether the current road surface is snowy, thereby providing a reference value for subsequently accurately determining whether the road surface is snowy.

[0175] Optionally, the image may be input into a road surface classification model, and the image may be processed by the road surface classification model to obtain a classification result and confidence level of the image.

[0176] In this case, the operation of step (1) may include the following steps a to d.

[0177] a. Perform feature extraction on the image to obtain the first image feature of the image.

[0178] Optionally, the pavement classification model may include a feature extraction part. After the image is input into the pavement classification model, the feature extraction part in the pavement classification model may perform feature extraction on the image to extract the image features of the image, that is, to obtain the first image features.

[0179] The feature extraction part may include multiple extraction stages, wherein each of the multiple extraction stages includes a convolution layer, so that the features of the image can be extracted through the multiple extraction stages to obtain the first image feature.

[0180] Optionally, the first extraction stage of the multiple extraction stages may include a convolution layer, a BN (Batch Normalization, batch normalization) layer, an activation layer, a pooling layer, etc. Any extraction stage of the remaining multiple extraction stages except the first extraction stage includes multiple residual blocks. Any residual block of the multiple residual blocks may include multiple convolution layers, BN layers, or activation layers.

[0181] The convolution layer is used to perform convolution operations on the image to obtain a feature matrix. The BN layer is used to normalize this feature matrix so that the features in the feature matrix are mapped to the range [0, 1], thereby obtaining a new feature matrix. In addition, during the training phase of the road surface classification model, the BN layer can increase the learning rate and map the features to the range [0, 1] so that the features obtained each time are different. This is equivalent to adding noise, which can prevent the eigenvalues in the feature matrix from concentrating around a certain value as the network training deepens, thereby causing overfitting. Therefore, the BN layer can also increase the learning rate of the model and avoid the problem of inaccurate model classification due to overfitting.

[0182] The step of normalizing the feature matrix may be: normalizing the feature matrix by using the following formula (7), formula (8), and formula (9) to map the value of the feature matrix to between [0, 1].

[0183]

[0184]

[0185]

[0186] Among them, μ is the mean of the feature matrix, σ is the variance of the feature matrix, is the normalized feature matrix. m is the number of features in the feature matrix, and i represents the i-th feature in the feature matrix. ε is a constant, a small value, used to prevent division by zero.

[0187] Since the BN layer has learning parameters, the feature matrix of the BN layer output can be obtained by the following formula (10).

[0188]

[0189] Among them, y i is the i-th feature in the feature matrix output by the batch normalization layer. γ and β are learning parameters whose purpose is to preserve the original data distribution. During the training phase of the road surface classification model, the initial values of γ and β are 1 and 0, respectively. The values of γ and β are continuously updated as the network is trained.

[0190] Since the BN layer has learning parameters, it cannot guarantee that the feature matrix input to the next layer has a mean of 0, so there may be negative numbers. In this case, the feature matrix output by the BN layer can be input into the activation layer so that the positive part of the feature matrix can be retained after passing through the activation layer, and the negative part of the feature matrix can be changed to 0. That is, the activation operation is performed through the following formula (11).

[0191]

[0192] in, is the feature matrix output by the activation layer.

[0193] Optionally, the activation function used by the activation layer can be a RELU (Rectified Linear Unit) activation function. The RELU activation function has a unilateral inhibition function, and the RELU activation function can output a negative eigenvalue as 0. In addition, during the training phase of the road surface classification model, the gradient is 0 when the eigenvalue is less than 0, and the gradient is always 1 when the eigenvalue is greater than 0. In this case, there will no longer be the problem of gradient vanishing and gradient exploding. Therefore, during the training phase of the road surface classification model, the RELU activation function can also avoid the problem of gradient vanishing during the training process.

[0194] The feature matrix output by the activation layer is then input into the pooling layer to reduce the dimension of the feature matrix through the pooling layer.

[0195] In the embodiment of the present application, the pooling layer can perform pooling operations by using maximum pooling, average pooling, etc., and the embodiment of the present application does not limit this.

[0196] For example, Figure 4 It is a flow chart of pooling operation, see Figure 4 , Figure 4 The feature matrix 401 before the pooling operation, the pooling window 402, and the feature matrix 403 after the pooling operation are included. The feature matrix 401 before the pooling operation is pooled using the maximum pooling method, and the size of the pooling window 402 is 2×2. During the pooling operation, the pooling window 402 is slid across the feature matrix 401. During the sliding process of the pooling window 402, the feature with the largest eigenvalue among the features covered by the pooling window 402 is retained. In this way, the feature with the largest eigenvalue among the features covered by the pooling window 402 in the feature matrix can be obtained each time, and the feature matrix 402 after the pooling operation can be obtained subsequently.

[0197] In addition, the residual block is to extract the image features while retaining the original information of the image features. For example, Figure 5 A schematic diagram of a residual block. Figure 5 , Figure 5The system includes an input feature matrix 501, multiple first convolutional layers 502, a second convolutional layer 503, an activation layer 504, and an output feature matrix 505. The input feature matrix 501 passes through multiple first convolutional layers 502 and second convolutional layers 503, respectively. The second convolutional layer 503 is used to retain the feature matrix with the original image information. The output of the first convolutional layer and the output of the second convolutional layer are then fused and input to the activation layer 504 for activation. Thus, the output feature matrix 505 obtained by the activation layer 504 is a feature matrix including the original information of the image features.

[0198] The number of residual blocks included in each extraction stage may be the same or different, and the structure of each residual block may be the same or different, which is not limited in the embodiment of the present application.

[0199] In this way, by using the residual block for feature extraction, the original information of the image features can be effectively retained, thereby making the extracted image features more accurate, that is, making the first image features more accurate.

[0200] b. Activate and pool the first image features to obtain the second image features.

[0201] Optionally, the activation operation on the first image feature may be: performing an activation operation on the first image feature through a Relu activation function to obtain an activated image feature.

[0202] The activated image features may then be subjected to a pooling operation, optionally by average pooling, to obtain a second image feature.

[0203] In this case, by performing an activation operation on the first image, the eigenvalues of the first image features are mapped to the range [0, 1]. This preserves the positive values in the feature matrix and controls the output of the negative values to 0, thus reducing the computational complexity of the model. Subsequent average pooling reduces the dimensionality while preserving the overall image features and removing redundant information. This reduces the computational effort and yields accurate second image features.

[0204] Optionally, after the feature extraction part of the road surface classification model, an activation layer and a pooling layer may also be included. The activation layer is used to activate the first image feature, and the pooling layer is used to pool the image features output by the activation layer to obtain the second image feature.

[0205] c. Fully connect and normalize the second image features to obtain confidence levels of multiple predicted classification results, where the confidence levels are used to indicate the credibility of the corresponding predicted classification results.

[0206] Optionally, the road surface classification model may also include a fully connected layer and a Softmax normalization layer. The fully connected layer is used to perform a fully connected operation on the second image feature, which can act as a classifier. The Softmax normalization layer is used to determine the probability that the image belongs to the corresponding category, that is, to determine the confidence of multiple predicted classification results. The Softmax normalization layer can use the Softmax function to determine the confidence of multiple predicted classification results, that is, the confidence of the multiple predicted classification results can be determined by the following formula (12).

[0207]

[0208] Among them, Z j It refers to the confidence of the j-th classification result among the multiple predicted classification results, and N is the total number of categories of the multiple predicted classification results.

[0209] d. Determine the predicted classification result with the highest confidence among the multiple predicted classification results as the classification result of the image.

[0210] Since the greater the confidence of a predicted classification result, the greater the probability that the image is the predicted classification result, the predicted classification result is more credible, that is, the classification result of the image is very likely to be the predicted classification result, then the predicted classification result with the highest confidence among the multiple predicted classification results can be determined as the classification result of the image, thereby making the classification of the image more accurate.

[0211] On the contrary, the smaller the confidence of a predicted classification result, the smaller the probability that the image is the predicted classification result, and the predicted classification result is not credible enough, that is, the classification result of the image is most likely not the predicted classification result.

[0212] Now take the feature extraction part including five extraction stages as an example, combined with Figure 6 The model structure of the road surface classification model is illustrated with an example. Figure 6 , Figure 6 It includes an input image 601, multiple extraction stages 602, an activation layer 603, a pooling layer 604, a fully connected layer 605, a Softmax normalization layer 606, and classification results and confidence 607.

[0213] First extraction stage 602: The first extraction stage 602 may include a convolutional layer, a batch normalization layer, a ReLU layer, and a pooling layer. Assume that the convolution kernel size in this convolutional layer is 7*7 and the number of channels is 64. The pooling window size of this pooling layer is 3*3 and the stride is 2.

[0214] If the input image 601 of the first extraction stage 602 has a size of 224*224*3, then after convolution of the input image 601 through the first convolutional layer, a feature matrix of 112*112*64 is obtained. This 112*112*64 feature matrix is then input into the batch normalization layer for normalization, resulting in a new feature matrix with the same dimension, 112*112*64. The normalized feature matrix can then be input into the Relu layer to retain the positive values and output the negative values as 0. The feature matrix output from the Relu layer can then be input into the pooling layer, which further extracts features and reduces the feature dimension, resulting in a feature matrix of 56*56*64. This means that a 56*56*64 feature matrix can be obtained from the first extraction stage 602.

[0215] Second extraction stage 602: The second extraction stage 602 may include one residual block 1 and two residual blocks 2. For example, Figure 7 (a) in the figure is a schematic diagram of residual block 1. Figure 7 (b) in is a schematic diagram of residual block 2. Figure 7 In (a), the input size of residual block 1 is W*H*C, where W is the length of the feature matrix, H is the height of the feature matrix, and C is the number of channels of the feature matrix. After passing through residual block 1, the output is a feature matrix of W*H*4C. Figure 7 The input size of residual block 2 in (b) is W*H*C. After passing through residual block 2, a feature matrix of W*H*C can be output.

[0216] First, the input feature matrix of the second extraction stage 602 is the output feature matrix of the first extraction stage 602, that is, a feature matrix of 56*56*64. This 56*56*64 is then output to residual block 1 to obtain a feature matrix of 56*56*256. This 56*56*256 feature matrix is then input to the residual block formed by stacking two residual blocks 2 to obtain a feature matrix of 56*56*256. In other words, the output feature matrix of the second extraction stage 602 is a feature matrix of 56*56*256.

[0217] The third extraction stage 602: The third extraction stage 602 may include one residual block 3 and three residual blocks 2. For example, see Figure 7 (c) in is a schematic diagram of residual block 3. Figure 7 As shown in (c), the input size of residual block 3 is W*H*C. After processing by residual block 3, a feature matrix of W / 2*H / 2*2C can be obtained.

[0218] First, the input feature matrix of the third extraction stage 602 is the output feature matrix of the second extraction stage 602, that is, a feature matrix of 56*56*256. Then, this 56*56*256 feature matrix is input into the residual block 3 to obtain a feature matrix of 28*28*512. Then, this 28*28*512 feature matrix is input into the residual block stacked by three residual blocks 2. After processing by the three residual blocks 2, a feature matrix of 28*28*512 can be obtained. That is, the output feature matrix of the third extraction stage 602 is a feature matrix of 28*28*512.

[0219] Fourth extraction stage 602: The fourth extraction stage 602 may include one residual block 3 and five residual blocks 2. The input feature matrix of the fourth extraction stage 602 is the output feature matrix of the third extraction stage 602, that is, a 28*28*512 feature matrix. This 28*28*512 feature matrix is then input into residual block 3. After processing by residual block 3, a 14*14*1024 feature matrix can be obtained. This 14*14*1024 feature matrix is then input into a residual block composed of five stacked residual blocks 2. After processing by the five residual blocks 2, a 14*14*1024 feature matrix can be obtained. In other words, the output feature matrix of the fourth extraction stage 602 is a 14*14*1024 feature matrix.

[0220] Fifth extraction stage 602: The fifth extraction stage 602 may include a residual block 3 and two residual blocks 2. The input feature matrix of the fifth extraction stage 602 is the output feature matrix of the fourth extraction stage 602, that is, a 14*14*1024 feature matrix. This 14*14*1024 feature matrix is then input into residual block 3. After processing by residual block 3, a 7*7*2048 feature matrix can be obtained. This 7*7*2048 feature matrix is then input into a residual block formed by stacking two residual blocks 2. After processing by the two residual blocks 2, a 7*7*2048 feature matrix can be obtained. In other words, the output feature matrix of the fifth extraction stage 602 is a 7*7*2048 feature matrix.

[0221] In this case, the fifth extraction stage 602 serves as the last extraction stage 602 of the entire feature extraction part, and the output feature matrix of the fifth extraction stage 602 is also the output feature of the entire feature extraction part, that is, the first image feature obtained after feature extraction of the image.

[0222] Next, the activation layer 603 performs an activation operation on the first image feature to obtain the activated image feature, and then inputs the activated image feature into the pooling layer 604 for performing a pooling operation to obtain the second image feature.

[0223] The second image features are then input into a fully connected layer 605 for full connection, yielding fully connected image features. A softmax normalization layer 606 then calculates the confidence levels of multiple predicted classification results based on the image features output by the fully connected layer 605. The road surface classification model then outputs a classification result and confidence level 607 for the image based on the confidence levels of the multiple predicted classification results.

[0224] In this way, the road surface classification model can extract the deep features of the image, thereby extracting more detailed features. Then, road surface classification is performed based on the deep features of the image, that is, classification operations are performed based on more detailed image features, so that more accurate classification results can be obtained.

[0225] (2) When the classification result is a snowy road surface, the confidence level of the classification result is greater than a preset confidence threshold, the road surface limit adhesion coefficient is less than a preset adhesion coefficient threshold, and the environmental parameters meet the target conditions, it is determined that the current driving road surface is a snowy road surface.

[0226] The preset confidence threshold can be set in advance and can be set relatively high. For example, the preset confidence threshold can be set to 0.7. The preset adhesion coefficient threshold can also be set in advance and can be set based on the road surface's limit adhesion coefficient for snowy roads. For example, the road surface's limit adhesion coefficient can be set to 0.4.

[0227] The target condition is used to determine whether the environmental parameters meet the environmental parameters under snowy conditions. The target condition can also be set in advance, and the target condition can be set according to the environmental parameters of the environment under snowy conditions.

[0228] In this case, when the classification result is a snowy road surface and the confidence level of the classification result is greater than a preset confidence threshold, it indicates that the classification result obtained by classifying the image is relatively credible, that is, the classification result is relatively accurate, indicating that the image may contain features related to a snowy road surface. Therefore, it can be seen from the image that the current driving road surface may be a snowy road surface.

[0229] Furthermore, when the road surface limit adhesion coefficient is less than the preset adhesion coefficient threshold, it means that the road surface limit adhesion coefficient is small and is already within the range of the road surface limit adhesion coefficient of snowy roads, which means that the current driving road surface is likely to be a snowy road surface.

[0230] Furthermore, when the environmental parameters meet the target conditions, it means that the environmental parameters have also reached the environmental parameters of the environment under snowy conditions, which can also mean that the current driving road surface is likely to be a snowy road surface.

[0231] In this way, when all three conditions above are satisfied, that is, when all three conditions above can indicate that the current driving road surface is very likely to be a snowy road surface, it can be determined that the current driving road surface is a snowy road surface.

[0232] It is worth noting that when the classification result is not a snowy road surface, and / or the road surface limit adhesion coefficient is greater than or equal to the preset adhesion coefficient threshold, and / or the environmental parameters do not meet the target conditions, it is determined that the current driving road surface is not a snowy road surface.

[0233] In this case, if the classification result is not a snowy road surface, it means that the image does not contain features related to a snowy road surface. Then, it can be seen from the image that the current driving road surface is not a snowy road surface, and it can be directly determined that the current driving road surface is not a snowy road surface.

[0234] When the road surface limit adhesion coefficient is greater than or equal to the preset adhesion coefficient threshold, it means that the road surface limit adhesion coefficient is large, and the road surface limit adhesion coefficient has exceeded the road surface limit adhesion coefficient of a snowy road surface, which means that the current driving road surface is not a snowy road surface.

[0235] When the environmental parameters do not meet the target conditions, it means that the environmental parameters do not meet the environmental parameters of the environment under snowy conditions, thereby indicating that the current environment of the target vehicle is not a snowy environment, and then it can be determined that the current driving road surface is not a snowy road surface.

[0236] In this way, when at least one of the above three conditions is met, it can be determined that the target vehicle is not currently on a snowy road surface, that is, the current driving road surface is not a snowy road surface.

[0237] Optionally, it is also possible to determine that the current driving road surface is not a snowy road surface when the classification result is a snowy road surface, and the confidence of the classification result is less than a preset confidence threshold, and / or the road surface limit adhesion coefficient is greater than or equal to a preset adhesion coefficient threshold, and / or the environmental parameters do not meet the target conditions.

[0238] In this case, if the classification result is a snowy road surface, but the confidence of the classification result is less than a preset confidence threshold, it means that the classification result is inaccurate, which means that the road surface where the target vehicle is currently located may not be a snowy road surface.

[0239] In this way, when at least one of the above three conditions is met, it can also be determined that the target vehicle is not currently on a snowy road surface, that is, the current driving road surface is not a snowy road surface.

[0240] For example, the preset confidence threshold is 0.7, the preset adhesion coefficient threshold is 0.4, and the preset temperature threshold is -10°C. The classification result of the image is snowy road surface, and the confidence of the classification result is 0.8. The road surface limit adhesion coefficient of the current driving road surface is 0.3, and the ambient temperature obtained by the temperature sensor of the target vehicle is -15°C. If the classification result of the image is snowy road surface, and the confidence of the classification result (0.8) is greater than the preset confidence threshold (0.7); and the road surface limit adhesion coefficient of the current driving road surface (0.3) is less than the preset adhesion coefficient threshold (0.4); and the ambient temperature (-15°C) is less than the preset temperature threshold (-10°C), then it can be determined that the current driving road surface is snowy road surface.

[0241] Optionally, the environmental parameter may include at least one of ambient temperature, humidity, and snow amount. The method for determining whether the environmental parameter meets the target condition may include at least one of the following possible methods.

[0242] A first possible implementation method is to collect the temperature of the environment in which the target vehicle is located through a temperature sensor of the target vehicle; when the temperature is less than a preset temperature threshold, it is determined that the environmental parameter meets the target condition; when the temperature is greater than or equal to the preset temperature threshold, it is determined that the environmental parameter does not meet the target condition.

[0243] Alternatively, the temperature sensor may be a temperature sensor located behind a front grille of the target vehicle, which is used to detect the temperature in the environment.

[0244] The preset temperature threshold can be set in advance, and the preset temperature threshold can be set according to the temperature of the environment under snowy conditions. The preset temperature threshold can be set lower, for example, the preset temperature threshold can be set to -10°C (degrees Celsius).

[0245] In this case, if the temperature is less than the preset temperature threshold, it means that the temperature is low and is within the temperature range of the environment under snowy conditions, which means that the target vehicle is likely to be in snowy conditions, and it can be determined that the environmental parameters meet the target conditions. If the temperature is greater than or equal to the preset temperature threshold, it means that the temperature is not low, that is, the temperature is higher than the temperature range of the environment under snowy conditions, which means that the target vehicle is not in snowy conditions, and it can be determined that the environmental parameters do not meet the target conditions.

[0246] A second possible implementation method is to collect the humidity of the environment in which the target vehicle is located through a humidity sensor of the target vehicle; when the humidity is greater than a preset humidity threshold, it is determined that the environmental parameter meets the target condition; when the humidity is less than or equal to the preset humidity threshold, it is determined that the environmental parameter does not meet the target condition.

[0247] Alternatively, the humidity sensor may be a humidity sensor located behind a front grille of the target vehicle, and configured to detect humidity in the environment.

[0248] The preset humidity threshold can be set in advance. The preset humidity threshold can be set according to the humidity of the environment under snowy conditions, and the preset humidity threshold can be set higher.

[0249] In this case, if the humidity is greater than the preset humidity threshold, it means that the humidity is high and is within the humidity range of the environment under snowy conditions, which means that the target vehicle is likely to be in snowy conditions, and it can be determined that the environmental parameters meet the target conditions. If the humidity is less than the preset humidity threshold, it means that the humidity is not high, that is, the humidity is lower than the humidity range of the environment under snowy conditions, which means that the target vehicle is not in snowy conditions, and it can be determined that the environmental parameters do not meet the target conditions.

[0250] A third possible implementation method is to collect the amount of snow in the environment where the target vehicle is located through the snow sensor of the target vehicle; when the amount of snow is greater than a preset snow amount threshold, it is determined that the environmental parameter meets the target condition; when the amount of snow is less than or equal to the preset snow amount threshold, it is determined that the environmental parameter does not meet the target condition.

[0251] Optionally, the snow sensor may be a snow sensor located at a windshield of the target vehicle, and is used to detect the amount of snow in the environment.

[0252] The preset snow amount threshold can be set in advance, and the preset snow amount threshold can be set higher.

[0253] In this case, if the amount of snow is greater than the preset snow amount threshold, it means that the snow amount is high, which means that the target vehicle is likely to be in snowy conditions, and it can be determined that the environmental parameters meet the target conditions. If the amount of snow is less than the preset snow amount threshold, it means that the snow amount is not high, which means that the target vehicle is not in snowy conditions, and it can be determined that the environmental parameters do not meet the target conditions.

[0254] Of course, in addition to determining whether the environmental parameters meet the target conditions through at least one of the above three implementation methods, it is also possible to determine whether the environmental parameters meet the target conditions through other methods, which is not limited in the embodiments of the present application.

[0255] (3) When the classification result is a slippery road surface, the confidence level of the classification result is greater than a preset confidence threshold, the road surface limit adhesion coefficient is less than a preset adhesion coefficient threshold, and the environmental parameters meet preset conditions, the current driving road surface is determined to be a slippery road surface.

[0256] Optionally, the environmental parameter may further include rainfall. The rainfall may be collected by a rain sensor of the target vehicle. The preset condition may be set as: the humidity is greater than or equal to a target humidity threshold and / or the rainfall is greater than or equal to a target rainfall threshold.

[0257] The target humidity threshold and the target rainfall threshold can be set in advance. Alternatively, the target humidity threshold and the target rainfall threshold can be set by a technician according to actual needs. For example, the target humidity threshold and the target rainfall threshold can both be set to be larger.

[0258] In this case, when the humidity is greater than or equal to the target humidity threshold and / or the rainfall is greater than or equal to the target rainfall threshold, it means that the humidity is high and / or the rainfall is high, which can indicate that the target vehicle is in an environment with high humidity and / or high rainfall, which can indicate that the road surface on which the target vehicle is traveling is likely to be a wet road surface or a rainy road surface, that is, the target vehicle is likely to be on a slippery road surface.

[0259] In addition, when the classification result is a slippery road surface and the confidence level of the classification result is greater than a preset confidence threshold, it indicates that the classification result obtained by classifying the image is relatively credible, that is, the classification result is relatively accurate, indicating that the image may contain features related to a slippery road surface. Therefore, it can be seen from the image that the current driving road surface may be a slippery road surface.

[0260] Furthermore, when the road surface limit adhesion coefficient is less than the preset adhesion coefficient threshold, it means that the road surface limit adhesion coefficient is small and is already within the range of the road surface limit adhesion coefficient of a wet road surface, which means that the current driving road surface is likely to be a wet road surface.

[0261] In this way, when all three of the above conditions are met, that is, when all the above three conditions can indicate that the current driving road surface is very likely to be a slippery road surface, it can be determined that the current driving road surface is a slippery road surface.

[0262] In addition, when at least one of the above three conditions is not satisfied, that is, when one of the above three conditions indicates that the current driving road surface is not a slippery road surface, it can be determined that the current driving road surface is not a slippery road surface.

[0263] It is worth noting that the road type of the current driving road can be determined through the above steps 201 to 203. Optionally, when it is determined that the current driving road is a snowy road, the target vehicle can be controlled to drive in a snowy mode.

[0264] Snow mode allows the vehicle to smoothly traverse snowy roads. In Snow mode, the throttle's response to the accelerator pedal is reduced, and the engine's power output is lower than normal, resulting in less transmission torque and, consequently, less friction on the wheels.

[0265] In this case, when the current driving road surface is a snowy road surface, the target vehicle drives on the snowy road surface, which can prevent the wheels from slipping, thereby maintaining stable driving of the vehicle and helping the vehicle to pass through the snowy road surface stably.

[0266] Optionally, when it is determined that the current driving road surface is a snowy road surface, the target vehicle may also be controlled to turn on the headlights and position lights.

[0267] In this case, turning on the headlights can help the driver observe the driving environment more clearly. In addition, turning on the clearance lights can serve as a reminder to the following vehicle to remind the following vehicle that there is a vehicle in front of it, thereby facilitating the safe driving of the target vehicle.

[0268] In an embodiment of the present application, the controller first obtains an image of the current driving surface of the target vehicle, current driving parameters, and environmental parameters of the target vehicle's environment, that is, obtains the driving parameters of the target vehicle and environmental information of the target vehicle's environment. Thereafter, based on the current driving parameters of the target vehicle, the ultimate road adhesion coefficient of the current driving surface is determined. Subsequently, based on the image, the ultimate road adhesion coefficient, and the environmental parameters, the road surface type of the current driving surface is determined. Since the ultimate road adhesion coefficient can indicate the maximum friction between the target vehicle's tires and the ground, the image and the environmental parameters of the target vehicle's environment can indicate environmental information of the target vehicle's environment. In this way, the current driving surface is subsequently determined by combining various features, so that the road surface type of the current driving surface can be determined more accurately, ensuring that the vehicle's driving parameters can be adjusted in a timely manner, thereby ensuring driving safety.

[0269] Figure 8 This is a schematic diagram of the structure of a road surface type determination device provided by an embodiment of the present application. The road surface type determination device can be implemented as part or all of a vehicle by software, hardware, or a combination of both. The vehicle can be as follows Figure 9 Vehicle shown. Figure 8 The device includes: a first acquisition module 801, a first determination module 802, and a second determination module 803.

[0270] The first acquisition module 801 is used to acquire an image of the current road surface on which the target vehicle is traveling, current driving parameters of the target vehicle, and environmental parameters of the environment in which the target vehicle is located;

[0271] A first determining module 802 is configured to determine a road surface limit adhesion coefficient of a current driving road surface based on current driving parameters of the target vehicle;

[0272] The second determining module 803 is configured to determine the road type of the current driving road based on the image, the road surface limit adhesion coefficient, and the environmental parameter.

[0273] Optionally, the device further comprises:

[0274] The second acquisition module is used to obtain the current speed, current wheel angle and wheelbase of the target vehicle;

[0275] a third determining module, configured to determine a current image acquisition range based on a current speed of the target vehicle, a current wheel angle, and a wheelbase of the target vehicle;

[0276] The first acquisition module 801 is used to:

[0277] According to the current image acquisition range, an image of the current road surface on which the target vehicle is traveling is acquired.

[0278] Optionally, the current driving parameters include the acceleration of the target vehicle, the speed of the target vehicle, and the wheel speeds of multiple wheels of the target vehicle. The first determining module 802 is configured to:

[0279] Divide the acceleration of the target vehicle by the acceleration due to gravity to obtain the actual adhesion coefficient of the current road surface at the current moment;

[0280] For any one of the plurality of wheels of the target vehicle, determining a slip ratio of the wheel based on a speed of the target vehicle and a wheel speed of the wheel;

[0281] The road surface limit adhesion coefficient is determined based on the slip ratios of the plurality of wheels of the target vehicle and the actual adhesion coefficient.

[0282] Optionally, the first determining module 802 is configured to:

[0283] When the number of wheels among the multiple wheels whose slip rates are greater than or equal to the preset slip rate threshold is greater than or equal to a preset number, determining the actual adhesion coefficient as the road surface limit adhesion coefficient;

[0284] When the number of wheels among the multiple wheels whose slip rate is greater than or equal to the preset slip rate threshold is less than a preset number, the maximum value between the actual adhesion coefficient and the target adhesion coefficient is determined as the road surface limit adhesion coefficient, and the target adhesion coefficient is the road surface limit adhesion coefficient at the previous moment.

[0285] Optionally, the second determining module 803 is configured to:

[0286] Classify the image and obtain the classification result and confidence of the image;

[0287] When the classification result is a snowy road surface, the confidence of the classification result is greater than a preset confidence threshold, the road surface limit adhesion coefficient is less than a preset adhesion coefficient threshold, and the environmental parameters meet the target conditions, it is determined that the current driving road surface is a snowy road surface.

[0288] Optionally, the environmental parameter includes at least one of temperature, humidity, and snow amount, and the second determining module 803 is configured to:

[0289] The temperature of the environment in which the target vehicle is located is collected by a temperature sensor of the target vehicle; if the temperature is less than a preset temperature threshold, it is determined that the environmental parameter meets the target condition; if the temperature is greater than or equal to the preset temperature threshold, it is determined that the environmental parameter does not meet the target condition;

[0290] The humidity of the environment in which the target vehicle is located is collected by a humidity sensor of the target vehicle; if the humidity is greater than a preset humidity threshold, it is determined that the environmental parameter meets the target condition; if the humidity is less than or equal to the preset humidity threshold, it is determined that the environmental parameter does not meet the target condition;

[0291] The amount of snow in the environment where the target vehicle is located is collected through the snow sensor of the target vehicle; when the amount of snow is greater than a preset snow amount threshold, it is determined that the environmental parameter meets the target condition; when the amount of snow is less than or equal to the preset snow amount threshold, it is determined that the environmental parameter does not meet the target condition.

[0292] Optionally, the second determining module 803 is configured to:

[0293] Performing feature extraction on the image to obtain a first image feature of the image;

[0294] Activate and pool the first image feature to obtain the second image feature;

[0295] Performing full connection and normalization on the second image features to obtain confidence levels of multiple predicted classification results, where the confidence levels are used to indicate the credibility of the corresponding predicted classification results;

[0296] The predicted classification result with the highest confidence among the multiple predicted classification results is determined as the classification result of the image.

[0297] Optionally, the device further comprises:

[0298] The control module is used to control the target vehicle to drive in a snow mode when it is determined that the current driving road surface is a snow road surface.

[0299] In an embodiment of the present application, an image of the current road surface of the target vehicle, current driving parameters, and environmental parameters of the target vehicle's environment are first obtained, that is, the driving parameters of the target vehicle and environmental information of the target vehicle's environment are obtained. Thereafter, based on the current driving parameters of the target vehicle, the road surface limit adhesion coefficient of the current road surface is determined. Subsequently, based on the image, the road surface limit adhesion coefficient, and the environmental parameters, the road surface type of the current road surface is determined. Since the road surface limit adhesion coefficient can indicate the maximum friction between the target vehicle's tires and the ground, the image and the environmental parameters of the target vehicle's environment can indicate environmental information of the target vehicle's environment. In this way, the current road surface is subsequently determined by combining various features, so that the road surface type of the current road surface can be determined more accurately, ensuring that the vehicle's driving parameters can be adjusted in a timely manner, thereby ensuring driving safety.

[0300] It should be noted that: when the road type determination device provided in the above embodiment determines the road type of the target vehicle currently traveling on, it only uses the division of the above-mentioned functional modules as an example. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0301] The functional units and modules in the above embodiments may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The above integrated units may be implemented in the form of hardware or software functional units. In addition, the specific names of the functional units and modules are only for the purpose of distinguishing them from each other and are not intended to limit the scope of protection of the embodiments of this application.

[0302] The road surface type determination device and the road surface type determination method provided in the above embodiments belong to the same concept. The specific working process and technical effects brought about by the units and modules in the above embodiments can be found in the method embodiment part and will not be repeated here.

[0303] Figure 9 It is a structural schematic diagram of a vehicle provided in an embodiment of the present application.

[0304] For example, Figure 9 As shown, the vehicle includes: a memory 91 and a processor 90, wherein the memory 91 stores an executable program code 92, and the processor 90 is used to call and execute the executable program code 92 to perform the above-mentioned road type determination method.

[0305] This embodiment can divide the vehicle into functional modules based on the above-described method example. For example, each functional module can be mapped to a specific function, or two or more functions can be integrated into a single processing module. The integrated module can be implemented in hardware. It should be noted that the module division in this embodiment is illustrative and represents only one logical functional division. In actual implementation, other division methods may be used.

[0306] In the case of dividing each functional module into corresponding functional modules, the vehicle may include: a first acquisition module, a first determination module, and a second determination module. It should be noted that all relevant contents of each step involved in the above method embodiment can be referred to the functional description of the corresponding functional module and will not be repeated here.

[0307] The vehicle provided in this embodiment is used to execute the above-mentioned method for determining the road type, and thus can achieve the same effect as the above-mentioned implementation method.

[0308] In the case of an integrated unit, the vehicle may include a processing module and a storage module. The processing module may be used to control and manage the vehicle's movements, while the storage module may be used to support the vehicle's execution of program codes and data.

[0309] The processing module may be a processor or controller that implements or executes various exemplary logic blocks, modules, and circuits disclosed herein. The processor may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a digital signal processing (DSP) and a microprocessor, and the storage module may be a memory.

[0310] This embodiment also provides a computer-readable storage medium, which stores computer program code. When the computer program code runs on a computer, the computer executes the above-mentioned related method steps to implement the above-mentioned method for determining the road surface type in the above-mentioned embodiment.

[0311] This embodiment also provides a computer program product. When the computer program product is run on a computer, it enables the computer to execute the above-mentioned related steps to implement the above-mentioned method for determining the road surface type in the above-mentioned embodiment.

[0312] Among them, the vehicle, computer-readable storage medium, computer program product or chip provided in this embodiment are all used to execute the corresponding methods provided above. Therefore, the beneficial effects that can be achieved can refer to the beneficial effects in the corresponding methods provided above, and will not be repeated here.

[0313] Through the description of the above implementation methods, technical personnel in the relevant field can understand that for the convenience and simplicity of description, only the division of the above-mentioned functional modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0314] In the embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of modules or units is only a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0315] The above content is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A method for determining a road surface type, characterized in that: The method comprises: Acquire an image of a current driving road surface of a target vehicle, current driving parameters of the target vehicle, and environmental parameters of an environment in which the target vehicle is located; determining a road surface limit adhesion coefficient of the current driving road surface based on the current driving parameters of the target vehicle; determining a road type of the current driving road based on the image, the road surface limit adhesion coefficient, and the environmental parameters; The determining the road type of the current driving road based on the image, the road surface limit adhesion coefficient, and the environmental parameter includes: Classifying the image to obtain a classification result and a confidence level of the image; When the classification result is a snowy road surface, the confidence of the classification result is greater than a preset confidence threshold, the road surface limit adhesion coefficient is less than a preset adhesion coefficient threshold, and the environmental parameters meet the target conditions, it is determined that the current driving road surface is a snowy road surface.

2. The method according to claim 1, wherein Before acquiring the image of the current driving road surface of the target vehicle, the current driving parameters of the target vehicle, and the environmental parameters of the target vehicle, the method further includes: Obtaining the current speed, current wheel angle, and wheelbase of the target vehicle; Determining a current image acquisition range based on a current speed, a current wheel angle, and a wheelbase of the target vehicle; The method for obtaining an image of a current road surface on which a target vehicle is traveling comprises: According to the current image acquisition range, an image of the current driving road of the target vehicle is acquired.

3. The method according to claim 1, wherein The current driving parameters include the acceleration of the target vehicle, the speed of the target vehicle, and the wheel speeds of multiple wheels of the target vehicle. Determining the road surface limit adhesion coefficient of the current driving road surface based on the current driving parameters of the target vehicle includes: Dividing the acceleration of the target vehicle by the acceleration of gravity to obtain the actual adhesion coefficient of the current driving road surface at the current moment; For any one of the plurality of wheels of the target vehicle, determining a slip ratio of the wheel based on a speed of the target vehicle and a wheel speed of the wheel; The road surface limit adhesion coefficient is determined based on the slip ratios of the plurality of wheels of the target vehicle and the actual adhesion coefficient.

4. The method according to claim 3, wherein The determining the road surface limit adhesion coefficient based on the slip rates of the plurality of wheels of the target vehicle and the actual adhesion coefficient includes: When the number of wheels among the plurality of wheels whose slip rates are greater than or equal to a preset slip rate threshold is greater than or equal to a preset number, determining the actual adhesion coefficient as the road surface limit adhesion coefficient; When the number of wheels among the multiple wheels whose slip rate is greater than or equal to the preset slip rate threshold is less than the preset number, the maximum value between the actual adhesion coefficient and the target adhesion coefficient is determined as the road surface limit adhesion coefficient, and the target adhesion coefficient is the road surface limit adhesion coefficient at the previous moment.

5. The method according to claim 1, wherein The environmental parameter includes at least one of temperature, humidity, and snow amount. The method for determining whether the environmental parameter meets the target condition includes at least one of the following: collecting the temperature of the environment in which the target vehicle is located through a temperature sensor of the target vehicle; if the temperature is less than a preset temperature threshold, determining that the environmental parameter meets the target condition; if the temperature is greater than or equal to the preset temperature threshold, determining that the environmental parameter does not meet the target condition; collecting the humidity of the environment in which the target vehicle is located by a humidity sensor of the target vehicle; if the humidity is greater than a preset humidity threshold, determining that the environmental parameter meets the target condition; if the humidity is less than or equal to the preset humidity threshold, determining that the environmental parameter does not meet the target condition; The amount of snow in the environment where the target vehicle is located is collected through the snow sensor of the target vehicle; when the amount of snow is greater than a preset snow amount threshold, it is determined that the environmental parameters meet the target conditions; when the amount of snow is less than or equal to the preset snow amount threshold, it is determined that the environmental parameters do not meet the target conditions.

6. The method according to claim 1, wherein The classifying the image to obtain the classification result and confidence level of the image includes: Performing feature extraction on the image to obtain a first image feature of the image; activating and pooling the first image features to obtain second image features; Performing full connection and normalization on the second image features to obtain confidence levels of multiple predicted classification results, where the confidence levels are used to indicate the credibility of the corresponding predicted classification results; The predicted classification result with the greatest confidence among the multiple predicted classification results is determined as the classification result of the image.

7. The method according to claim 1, wherein After determining the road type of the current driving road based on the image, the road surface limit adhesion coefficient, and the environmental parameters, the method further includes: When it is determined that the current driving road surface is a snowy road surface, the target vehicle is controlled to drive in a snow mode.

8. A road surface type determination device, characterized in that: The device comprises: A first acquisition module is used to acquire an image of the current road surface on which the target vehicle is traveling, current driving parameters of the target vehicle, and environmental parameters of the environment in which the target vehicle is located; A first determining module is configured to determine a road surface limit adhesion coefficient of the current driving road surface based on the current driving parameters of the target vehicle; a second determining module, configured to determine a road type of the current driving road based on the image, the road surface limit adhesion coefficient, and the environmental parameters; The second determination module is specifically used to: classify the image to obtain the classification result and confidence of the image; when the classification result is a snowy road surface, the confidence of the classification result is greater than a preset confidence threshold, the road surface limit adhesion coefficient is less than a preset adhesion coefficient threshold, and the environmental parameters meet the target conditions, determine that the current driving road surface is a snowy road surface.

9. A vehicle, characterized in that: The vehicle comprises: a memory for storing executable program code; A processor is configured to call and run the executable program code from the memory, so that the vehicle executes the method according to any one of claims 1 to 7.

Images