Visibility processing method and apparatus, electronic device, and storage medium

Abstract

The present application provides a kind of visibility processing method, device, electronic equipment and storage medium.The method comprises: obtaining meteorological information, the driving information of vehicle and the first threshold value corresponding to visibility;Determine the first weight corresponding to meteorological information and the second weight corresponding to driving information, wherein the first weight is used to indicate the influence degree of meteorological information on visibility, and the second weight is used to indicate the influence degree of driving information on visibility;Determine the first parameter based on meteorological information, driving information, first weight and second weight;Based on the first threshold value and the first parameter, determine the second threshold value corresponding to visibility, and classify visibility based on the second threshold value.The method of the present application can dynamically respond to the changes of meteorological information and driving information to optimize the first threshold value of visibility, effectively improve the accuracy of visibility classification, and provide more reliable basis for driving personnel's traffic decision.

Description

Technical Field

[0001] This application relates to the field of traffic safety, and more particularly to a visibility processing method, apparatus, electronic device, and storage medium. Background Technology

[0002] In the field of traffic safety, changes in visibility can affect drivers' judgment of road conditions. Accurate visibility classification forecasts can effectively guide drivers to adjust their driving behavior, such as controlling speed and maintaining a safe distance, thereby significantly reducing the incidence of traffic accidents.

[0003] In related technologies, multiple visibility ranges are obtained by using a preset fixed threshold, the visibility range to which the visibility belongs is determined, and thus the visibility is classified. The preset fixed threshold is set by relevant technical personnel based on experience.

[0004] However, the preset fixed threshold cannot capture subtle changes in different visibility ranges, resulting in the visibility classification results not matching the actual risks or visual experience in the actual scene, thus reducing the accuracy of the visibility classification results. Summary of the Invention

[0005] This application provides a visibility processing method, apparatus, electronic device, and storage medium to improve the accuracy of visibility classification results.

[0006] Firstly, this application provides a visibility processing method, the method comprising:

[0007] Acquire meteorological information, vehicle driving information, and a first threshold corresponding to visibility;

[0008] A first weight corresponding to the meteorological information and a second weight corresponding to the driving information are determined. The first weight is used to indicate the degree of influence of the meteorological information on the visibility, and the second weight is used to indicate the degree of influence of the driving information on the visibility.

[0009] Based on the meteorological information, the driving information, the first weight, and the second weight, the first parameter is determined;

[0010] Based on the first threshold and the first parameter, a second threshold corresponding to the visibility is determined, and the visibility is classified based on the second threshold.

[0011] In this scheme, meteorological information, vehicle driving information, and a first threshold corresponding to visibility are acquired. A first weight corresponding to the meteorological information and a second weight corresponding to the driving information are determined. The first weight indicates the degree of influence of meteorological information on visibility, and the second weight indicates the degree of influence of driving information on visibility. Based on the meteorological information, driving information, first weight, and second weight, a first parameter is determined. Based on the first threshold and the first parameter, a second threshold corresponding to visibility is determined, and visibility is classified based on the second threshold. In this method, by acquiring meteorological information and vehicle driving information, determining their weights, and calculating the first parameter accordingly, the first threshold for visibility is adjusted to obtain the second threshold for classification. This dynamically responds to changes in meteorological and driving information to optimize the fixed threshold for visibility, effectively improving the accuracy of visibility classification and providing a more reliable basis for drivers' traffic decisions.

[0012] In one implementation, the meteorological information includes multiple sub-meteorological information, and the driving information includes multiple sub-driving information; based on the meteorological information, the driving information, the first weight, and the second weight, a first parameter is determined, including:

[0013] In the first weight, the first sub-weight corresponding to each sub-meteorological information is determined;

[0014] In the second weighting, the second sub-weight corresponding to each sub-driving information is determined;

[0015] The first parameter is determined based on the multiple sub-meteorological information, multiple first sub-weights, the multiple sub-driving information, and multiple second sub-weights.

[0016] In one implementation, determining the first parameter based on the plurality of sub-meteorological information, the plurality of first sub-weights, the plurality of sub-driving information, and the plurality of second sub-weights includes:

[0017] The product of the sub-meteorological information and the first sub-weight corresponding to the sub-meteorological information is determined to obtain multiple second parameters;

[0018] The product of the sub-driving information and the second sub-weight corresponding to the sub-driving information is determined to obtain multiple third parameters;

[0019] The first parameter is determined based on the plurality of second parameters and the plurality of third parameters.

[0020] In one implementation, determining a second threshold corresponding to the visibility based on the first threshold and the first parameter includes:

[0021] The current vehicle speed is obtained from the driving information, and the current humidity is obtained from the weather information;

[0022] The second threshold is determined based on the current vehicle speed, the current humidity, the first threshold, and the first parameter.

[0023] In this scheme, the current vehicle speed is obtained from driving information, and the current humidity is obtained from meteorological information. Based on the current vehicle speed, current humidity, a first threshold, and a first parameter, a second threshold is determined. This method further optimizes the fixed visibility threshold by using the two key variables—current vehicle speed and humidity—which significantly affect the visibility threshold. The resulting second threshold fully considers the driver's reaction characteristics, achieving a more refined dynamic adjustment of the visibility threshold and further improving the accuracy of visibility classification.

[0024] In one implementation, determining the second threshold based on the current vehicle speed, current humidity, the first threshold, and the first parameter includes:

[0025] A third threshold is determined based on the current vehicle speed and the current humidity.

[0026] Based on the first threshold and the first parameter, a fourth threshold is determined;

[0027] Based on the third threshold and the fourth threshold, a second threshold corresponding to the visibility is determined.

[0028] In this scheme, the current environment is divided into grids based on the current vehicle speed, resulting in multiple grids within the vehicle's driving range. The visibility corresponding to each grid is then acquired. Based on a second threshold and the visibility of each grid, the visibility of each grid is classified. This method dynamically adjusts the grid width according to vehicle speed, balancing refined perception of environmental details in low-speed scenarios with efficient analysis of long-distance vision at high speeds. This significantly improves the spatiotemporal adaptability of visibility data collection, and the driving prompts provided by the grid offer a more reliable basis for drivers' traffic decisions.

[0029] In one implementation, the visibility is classified based on the second threshold, including:

[0030] Based on the current vehicle speed, the current environment is divided into grids to obtain multiple grids within the vehicle's driving range;

[0031] Obtain the visibility corresponding to each grid;

[0032] Based on the second threshold and the visibility corresponding to each grid, the visibility of each grid is classified.

[0033] In this scheme, the current environment is divided into grids based on the current vehicle speed, resulting in multiple grids within the vehicle's driving range. The visibility corresponding to each grid is then acquired. Based on a second threshold and the visibility of each grid, the visibility of each grid is classified. This method dynamically adjusts the grid width according to vehicle speed, balancing refined perception of environmental details in low-speed scenarios with efficient analysis of long-distance vision at high speeds. This significantly improves the spatiotemporal adaptability of visibility data collection, and the driving prompts provided by the grid offer a more reliable basis for drivers' traffic decisions.

[0034] In one implementation, for any two adjacent first and second grids, after classifying the visibility of each grid, the method further includes:

[0035] When the vehicle travels from the range of the first grid to the range of the second grid, it is determined whether the visibility type of the first grid and the visibility type of the second grid are the same;

[0036] If they are different, a notification message indicating the change in visibility will be generated.

[0037] In this scheme, for any two adjacent first grids and second grids, when a vehicle travels from the range of the first grid to the range of the second grid, it is determined whether the visibility type of the first grid and the visibility type of the second grid are the same. If they are different, a visibility change prompt is generated, so that the driver can perceive the sudden change in visibility in time and gain valuable reaction time to adjust the driving strategy.

[0038] Secondly, embodiments of this application provide a visibility processing device, comprising:

[0039] The acquisition module is used to acquire meteorological information, vehicle driving information, and a first threshold corresponding to visibility.

[0040] The determining module is used to determine a first weight corresponding to the meteorological information and a second weight corresponding to the driving information, wherein the first weight is used to indicate the degree of influence of the meteorological information on the visibility, and the second weight is used to indicate the degree of influence of the driving information on the visibility;

[0041] The determining module is further configured to determine a first parameter based on the meteorological information, the driving information, the first weight, and the second weight;

[0042] The processing module is configured to determine a second threshold corresponding to the visibility based on the first threshold and the first parameter, and classify the visibility based on the second threshold.

[0043] The visibility processing device provided in this application embodiment can execute the technical solution shown in the above method embodiment. Its implementation principle and beneficial effects are similar, and will not be described again here.

[0044] In one implementation, a module is defined, specifically for:

[0045] In the first weight, the first sub-weight corresponding to each sub-meteorological information is determined;

[0046] In the second weighting, the second sub-weight corresponding to each sub-driving information is determined;

[0047] The first parameter is determined based on the multiple sub-meteorological information, multiple first sub-weights, the multiple sub-driving information, and multiple second sub-weights.

[0048] In one implementation, a module is defined, specifically for:

[0049] The product of the sub-meteorological information and the first sub-weight corresponding to the sub-meteorological information is determined to obtain multiple second parameters;

[0050] The product of the sub-driving information and the second sub-weight corresponding to the sub-driving information is determined to obtain multiple third parameters;

[0051] The first parameter is determined based on the plurality of second parameters and the plurality of third parameters.

[0052] In one implementation, the processing module is specifically used for:

[0053] The current vehicle speed is obtained from the driving information, and the current humidity is obtained from the weather information;

[0054] The second threshold is determined based on the current vehicle speed, the current humidity, the first threshold, and the first parameter.

[0055] In one implementation, the processing module is specifically used for:

[0056] A third threshold is determined based on the current vehicle speed and the current humidity.

[0057] Based on the first threshold and the first parameter, a fourth threshold is determined;

[0058] Based on the third threshold and the fourth threshold, a second threshold corresponding to the visibility is determined.

[0059] In one implementation, the processing module is specifically used for:

[0060] Based on the current vehicle speed, the current environment is divided into grids to obtain multiple grids within the vehicle's driving range;

[0061] Obtain the visibility corresponding to each grid;

[0062] Based on the second threshold and the visibility corresponding to each grid, the visibility of each grid is classified.

[0063] In one implementation, for any two adjacent first and second grids, the processing module is further configured to:

[0064] When the vehicle travels from the range of the first grid to the range of the second grid, it is determined whether the visibility type of the first grid and the visibility type of the second grid are the same;

[0065] If they are different, a notification message indicating the change in visibility will be generated.

[0066] The visibility processing device provided in this application embodiment can execute the technical solution shown in the above method embodiment. Its implementation principle and beneficial effects are similar, and will not be described again here.

[0067] Thirdly, this application provides an electronic device, including: a processor, and a memory communicatively connected to the processor;

[0068] The memory stores instructions that the computer executes;

[0069] The processor executes computer-executable instructions stored in memory to implement the method as described in the first aspect.

[0070] The electronic device provided in this application embodiment can execute the technical solution shown in the above method embodiment. Its implementation principle and beneficial effects are similar, and will not be repeated here.

[0071] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the method as described in the first aspect.

[0072] When the computer-executable instructions in the computer-readable storage medium provided in this application are executed by a processor, the technical solutions shown in the above method embodiments can be implemented. The implementation principle and beneficial effects are similar, and will not be repeated here.

[0073] Fifthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the method of the first aspect.

[0074] When the computer program in the computer program product provided in this application embodiment is executed by the processor, it can implement the technical solution shown in the above method embodiment. Its implementation principle and beneficial effects are similar, and will not be repeated here. Attached Figure Description

[0075] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0076] Figure 1 A schematic flowchart illustrating the visibility processing method provided in an embodiment of this application;

[0077] Figure 2 A flowchart illustrating the method for determining the second threshold provided in this application;

[0078] Figure 3 A schematic diagram illustrating the method for classifying visibility provided in this application;

[0079] Figure 4 A flowchart illustrating the method for determining a preset adaptive neural network provided in an embodiment of this application;

[0080] Figure 5 A schematic diagram of a visibility processing device provided in an embodiment of this application;

[0081] Figure 6 This is a structural diagram of an electronic device provided in an embodiment of this application.

[0082] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0083] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0084] In the field of traffic safety, changes in visibility can affect drivers' judgment of road conditions. Accurate visibility classification forecasts can effectively guide drivers to adjust their driving behavior, such as controlling speed and maintaining a safe distance, thereby significantly reducing the incidence of traffic accidents.

[0085] In related technologies, multiple visibility ranges are obtained by using a preset fixed threshold, the visibility range to which the visibility belongs is determined, and thus the visibility is classified. The preset fixed threshold is set by relevant technical personnel based on experience.

[0086] However, the preset fixed threshold cannot capture subtle changes in different visibility ranges, resulting in the visibility classification results not matching the actual risks or visual experience in the actual scene, thereby reducing the accuracy of visibility classification results and the effectiveness of early warning.

[0087] Based on the above-mentioned technical problems, the technical concept of the embodiments of this application is as follows:

[0088] The method involves acquiring meteorological information, vehicle driving information, and a first threshold corresponding to visibility. It then determines a first weight for the meteorological information and a second weight for the driving information, where the first weight indicates the degree of influence of meteorological information on visibility, and the second weight indicates the degree of influence of driving information on visibility. Based on the meteorological information, driving information, the first weight, and the second weight, a first parameter is determined. Based on the first threshold and the first parameter, a second threshold corresponding to visibility is determined, and visibility is classified based on the second threshold. This method, by acquiring meteorological information and vehicle driving information, determining their weights, and calculating the first parameter accordingly, adjusts the first threshold for visibility to obtain the second threshold for classification. This dynamic response to changes in meteorological and driving information optimizes the fixed threshold for visibility, effectively improving the accuracy of visibility classification and providing a more reliable basis for drivers' traffic decisions.

[0089] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0090] Figure 1 This is a flowchart illustrating the visibility processing method provided in the embodiments of this application, as shown below. Figure 1 As shown, the method includes:

[0091] S101. Obtain meteorological information, vehicle driving information, and the first threshold corresponding to visibility.

[0092] The execution subject of this application embodiment can be an electronic device or a visibility processing device installed in an electronic device. The visibility processing device can be implemented by software or by a combination of software and hardware.

[0093] Meteorological information includes multiple sub-meteorological information. For example, these sub-meteorological information include temperature, humidity, wind speed, air pressure, precipitation, time, terrain features, road signs, and the historical frequency of fog occurrences on the road sections in which vehicles travel.

[0094] Terrain features include slope, curvature, and altitude. Road markings include solid and dashed lines, guardrails, etc.

[0095] The vehicle's driving information includes multiple sub-driving information. For example, these sub-driving information may include the current vehicle speed, the density of surrounding vehicles, and the following driving distance.

[0096] Optionally, weather-related information can be obtained through a vehicle-mounted weather station.

[0097] Optionally, the terrain features of the current location can be obtained through a high-precision map.

[0098] Optionally, road signs can be identified using image recognition technology. For example, a road sign identified as having a guardrail can be represented as 1, and a road sign identified as having no guardrail can be represented as 0.

[0099] The frequency of historical fog occurrences along the vehicle's current route can be used as an example, representing the frequency of fog occurrences along the route in the past 30 days. Alternatively, the frequency of fog occurrences along the route in the past 30 days can be obtained by querying a historical meteorological database.

[0100] Optionally, the current vehicle speed can be obtained via the Controller Area Network (CAN) bus.

[0101] Optionally, the density of surrounding vehicles can be obtained through vehicle-to-everything (V2X) technology, where the density of surrounding vehicles is the number of vehicles per kilometer of the current road segment.

[0102] Optionally, the distance between the vehicle and the vehicle in front can be obtained through millimeter-wave radar or visual sensors.

[0103] The first threshold for visibility is obtained based on human experience. For example, the first threshold is 2 meters.

[0104] In one implementation, obtaining meteorological information, vehicle driving information, and a first threshold corresponding to visibility includes: obtaining raw meteorological information, raw vehicle driving information, and a first threshold corresponding to visibility; and normalizing the raw meteorological information and raw vehicle driving information to obtain meteorological information and vehicle driving information.

[0105] The raw meteorological information includes multiple sub-raw meteorological information.

[0106] The vehicle's original driving information includes multiple sub-original driving information.

[0107] For example, normalizing the original meteorological information can yield the corresponding sub-meteorological information.

[0108] Optionally, the original meteorological information and the sub-meteorological information satisfy the following formula 1:

[0109]

[0110] in, Let x represent the i-th sub-meteorological information, x represent the i-th sub-original meteorological information, and max represent the maximum value. i min represents the maximum value corresponding to the i-th sub-original meteorological information. i This represents the minimum value corresponding to the i-th sub-original meteorological information.

[0111] The maximum and minimum values ​​corresponding to the i-th sub-original meteorological information are determined by using a pre-set adaptive neural network. The pre-set adaptive neural network is obtained through training.

[0112] It should be noted that the sub-original driving information and its correspondence with the sub-driving information are similar to the correspondence with the sub-original meteorological information and the sub-meteorological information, and will not be elaborated here.

[0113] S102. Determine the first weight corresponding to meteorological information and the second weight corresponding to driving information.

[0114] The first weight is used to indicate the degree of influence of meteorological information on visibility, and the second weight is used to indicate the degree of influence of driving information on visibility.

[0115] In one implementation, a first weight corresponding to meteorological information and a second weight corresponding to driving information are determined by pre-setting an adaptive neural network model.

[0116] The first weight is used to indicate the degree of influence of meteorological information on visibility. It can be understood as the first weight being used to indicate the degree of influence of meteorological information on the first threshold corresponding to visibility.

[0117] In one implementation, the first weight includes N first sub-weights. Specifically, the i-th first sub-weight indicates the degree of influence of the i-th sub-meteorological information on the first threshold corresponding to visibility, where i = 1, 2, ..., N.

[0118] The second weight is used to indicate the degree of influence of driving information on visibility. It can be understood that the second weight is used to indicate the degree of influence of driving information on the first threshold corresponding to visibility.

[0119] In one implementation, the second weight includes M second sub-weights. Specifically, the j-th second sub-weight indicates the degree of influence of the j-th sub-driving information on the first threshold corresponding to visibility, where j = 1, 2, ..., M.

[0120] S103. Based on meteorological information, driving information, first weight and second weight, determine the first parameter.

[0121] In one implementation, a first parameter is determined by pre-setting an adaptive neural network model based on meteorological information, driving information, a first weight, and a second weight.

[0122] In one implementation, a first parameter is determined based on meteorological information, driving information, a first weight, and a second weight, including: determining a first sub-weight corresponding to each sub-meteorological information in the first weight; determining a second sub-weight corresponding to each sub-driving information in the second weight; and determining the first parameter based on multiple sub-meteorological information, multiple first sub-weights, multiple sub-driving information, and multiple second sub-weights.

[0123] In one implementation, a first parameter is determined based on multiple sub-meteorological information, multiple first sub-weights, multiple sub-driving information, and multiple second sub-weights. This includes: determining the product of the sub-meteorological information and the first sub-weights corresponding to the sub-meteorological information to obtain multiple second parameters; determining the product of the sub-driving information and the second sub-weights corresponding to the sub-driving information to obtain multiple third parameters; and determining the first parameter based on the multiple second parameters and the multiple third parameters.

[0124] Optionally, the sub-meteorological information, the first sub-weight corresponding to the sub-meteorological information, and the second parameter satisfy the following formula 2:

[0125]

[0126] in, W1 represents the i-th second parameter determined by the product of the sub-meteorological information and the corresponding first sub-weight. i X1 represents the first sub-weight of the i-th sub-meteorological information. i This represents the i-th sub-meteorological information.

[0127] Optionally, the sub-driving information, the corresponding second sub-weight, and the third parameter satisfy the following formula 3:

[0128]

[0129] in, W2 represents the j-th third parameter determined based on the sub-driving information and the corresponding second sub-weight. j This represents the j-th second sub-weight, which is the second sub-weight corresponding to the j-th sub-driving information. X2 j This represents the j-th sub-driving information.

[0130] In one implementation, the first threshold may include multiple first sub-thresholds.

[0131] When the first threshold includes K first sub-thresholds, each first sub-weight includes K meteorological-threshold weights. The k-th meteorological-threshold weight in the i-th first sub-weight indicates the degree of influence of the i-th meteorological information on the k-th first sub-threshold corresponding to visibility. Similarly, each second sub-weight includes K driving-threshold weights. The k-th driving-threshold weight in the j-th second sub-weight indicates the degree of influence of the j-th driving information on the k-th first sub-threshold corresponding to visibility. Where k = 1, 2, ..., K.

[0132] Optionally, the sub-meteorological information, the K meteorological-threshold weights corresponding to the sub-meteorological information, and the second parameter satisfy the following formula:

[0133] Formula 4:

[0134]

[0135] in, w1 represents the i-th second parameter determined based on the sub-meteorological information and the K meteorological-threshold weights corresponding to the sub-meteorological information. ik This represents the k-th meteorological-threshold weight in the i-th first sub-weight.

[0136] Optionally, the sub-driving information, the K driving-threshold weights corresponding to the sub-driving information, and the third parameter are as follows: Formula 5:

[0137]

[0138] in, w2 represents the j-th third parameter determined based on the sub-driving information and the K corresponding driving-threshold weights. jk This represents the k-th driving-threshold weight in the j-th second sub-weight.

[0139] When the first threshold includes K first sub-thresholds, the first parameter includes K first sub-parameters.

[0140] Optionally, multiple second parameters, multiple third parameters, and the first sub-parameter satisfy the following formula 6:

[0141]

[0142] Where, ΔT k Let λ represent the k-th first sub-parameter. k This represents the scaling factor corresponding to the k-th first sub-threshold, where r takes the value 1 or 2, and b k This represents the preset bias term corresponding to the k-th first sub-threshold, and tanh represents the hyperbolic tangent function.

[0143] By using a preset adaptive neural network model, the scaling factor corresponding to the first threshold and the preset bias term corresponding to the first threshold are determined. Specifically, the scaling factor corresponding to each first sub-threshold and the preset bias term corresponding to each first sub-threshold are determined.

[0144] S104. Based on the first threshold and the first parameter, determine the second threshold corresponding to the visibility, and classify the visibility based on the second threshold.

[0145] In one implementation, when the first threshold includes K first sub-thresholds, the first parameter includes K first sub-parameters, and the second threshold includes K second sub-thresholds.

[0146] The following describes one method for determining the second threshold corresponding to visibility based on the first threshold and the first parameter:

[0147] When the first threshold includes K first sub-thresholds, the first parameter includes K first sub-parameters, and the second threshold includes K second sub-thresholds;

[0148] The first sub-threshold, the first sub-parameter, and the second sub-threshold can satisfy the following formula 7:

[0149]

[0150] in, T represents the k-th second sub-threshold. k This represents the k-th first sub-threshold.

[0151] The following example illustrates one way to classify visibility based on a second threshold.

[0152] In one implementation, classifying visibility based on a second threshold can be as follows: obtaining the visibility within the vehicle's driving range; determining multiple visibility ranges based on multiple second sub-thresholds; determining the visibility range to which the visibility belongs based on the multiple visibility ranges; classifying the visibility based on the visibility range to which the visibility belongs, and determining the visibility type.

[0153] Optionally, the visibility type can be retrieved from a pre-stored first correspondence based on the visibility range to which the visibility belongs. The first correspondence includes multiple visibility ranges, multiple visibility types, and correspondences between multiple visibility ranges and multiple visibility types.

[0154] For example, the second threshold includes four second sub-thresholds: the first second sub-threshold is, for example, 50 meters; the second second sub-threshold is, for example, 100 meters; the third second sub-threshold is, for example, 200 meters; and the fourth second sub-threshold is, for example, 500 meters. This determines five visibility ranges: less than or equal to 50 meters, greater than 50 meters but less than or equal to 100 meters, greater than 100 meters but less than or equal to 200 meters, and greater than 200 meters but less than or equal to 500 meters. The first correspondence can be shown in Table 1.

[0155] Table 1 First Correspondence Relationship

[0156] Visibility range Visibility type Less than or equal to 50 meters Level 4 Greater than 50 meters and less than or equal to 100 meters Level 3 Greater than 100 meters and less than or equal to 200 meters Level 2 Greater than 200 meters and less than or equal to 500 meters Level 1 Greater than 500 meters Level 0

[0157] For example, when the visibility range is less than or equal to 50 meters, the visibility type can be determined as level 4 based on Table 1.

[0158] Furthermore, after determining the visibility type, an alarm can be issued based on the visibility type.

[0159] The following is an example of how to generate alarms based on visibility type:

[0160] When visibility is level 4, drivers are advised to turn on fog lights, low beam headlights, side marker lights, front and rear position lights, and hazard warning lights, and to stop in the nearest safe area, such as a highway service area or emergency stopping lane. If stopping is not possible, drivers must reduce their speed to a minimum, maintain a very low speed, keep a distance of at least 200 meters from the vehicle in front, and avoid sudden braking or lane changes.

[0161] When visibility is level 3, drivers are advised to turn on fog lights, low beam headlights, side marker lights, front and rear position lights, and hazard warning lights. The speed must not exceed 20 kilometers per hour. If on a highway, drivers should exit the highway as soon as possible from the nearest exit.

[0162] When visibility is level 2, drivers are advised to turn on fog lights, low beam headlights, side marker lights, front and rear position lights, and hazard warning lights; the vehicle speed must not exceed 40 kilometers per hour; and a distance of more than 100 meters must be maintained from the vehicle in front in the same lane.

[0163] When visibility is level 1, drivers are advised to turn on fog lights, low beam headlights, side marker lights, and front and rear position lights; the speed must not exceed 60 kilometers per hour; and a distance of more than 50 meters must be maintained from the vehicle in front in the same lane.

[0164] No alarm will be issued when the visibility type is 0.

[0165] The beneficial effects of this embodiment are as follows: In this embodiment, meteorological information, vehicle driving information, and a first threshold corresponding to visibility are obtained; a first weight corresponding to meteorological information and a second weight corresponding to driving information are determined, wherein the first weight is used to indicate the degree of influence of meteorological information on visibility, and the second weight is used to indicate the degree of influence of driving information on visibility; a first parameter is determined based on meteorological information, driving information, the first weight, and the second weight; a second threshold corresponding to visibility is determined based on the first threshold and the first parameter, and visibility is classified based on the second threshold. In the above method, by obtaining meteorological information and vehicle driving information, determining their weights and calculating the first parameter accordingly, and then adjusting the first threshold of visibility to obtain the second threshold and classifying accordingly, the fixed threshold of visibility (i.e., the first threshold) can be dynamically optimized in response to changes in meteorological information and driving information, effectively improving the accuracy of visibility classification and providing a more reliable basis for drivers' traffic decisions.

[0166] Below, in conjunction with Figure 2 ,exist Figure 1 Based on the embodiments, another method for determining the second threshold corresponding to visibility based on the first threshold and the first parameter will be described in detail.

[0167] Figure 2 A flowchart illustrating the method for determining the second threshold provided in this application is shown below. Figure 2 As shown, the method includes:

[0168] S201. Obtain the current vehicle speed from the driving information and the current humidity from the weather information.

[0169] S202. Based on the current vehicle speed, current humidity, first threshold and first parameter, determine the second threshold.

[0170] In one implementation, a second threshold is determined based on the current vehicle speed, the current humidity, a first threshold, and a first parameter, including:

[0171] Determine the third threshold based on the current vehicle speed and current humidity;

[0172] Based on the first threshold and the first parameter, determine the fourth threshold;

[0173] Based on the third and fourth thresholds, the second threshold corresponding to visibility is determined.

[0174] In one implementation, the current vehicle speed, the current humidity, and the third threshold satisfy the following formula 8:

[0175]

[0176] Among them, T fogThis represents the third threshold, v represents the current vehicle speed, and t represents the third threshold. react This represents the driver's braking reaction time, α represents the attenuation factor, and RH represents the current humidity.

[0177] The driver's braking reaction time is determined based on experience, for example, 2.5 seconds.

[0178] The attenuation factor can be determined empirically or based on a preset adaptive neural network.

[0179] In one implementation, the fourth threshold includes K fourth sub-thresholds. The fourth threshold is determined based on the first threshold and the first parameter. This can be understood as determining the k-th fourth sub-threshold based on the k-th first sub-threshold and the k-th first sub-parameter for the k-th fourth sub-threshold. Where k = 1, ..., K.

[0180] In one implementation, the first sub-threshold, the first parameter, and the fourth sub-threshold satisfy the following formula 9:

[0181]

[0182] in, This represents the k-th fourth sub-threshold.

[0183] In one implementation, when the fourth threshold includes K fourth sub-thresholds, the second threshold corresponding to visibility is determined based on the third and fourth thresholds. This can be understood as follows:

[0184] For the k-th fourth sub-threshold;

[0185] Based on the third threshold and the kth fourth threshold, the kth second threshold corresponding to visibility is determined.

[0186] In one implementation, the third threshold, the fourth threshold, and the second threshold satisfy the following formula 10:

[0187]

[0188] Where γ represents the dynamic weighting coefficient.

[0189] The dynamic weight coefficient is greater than or equal to 0 or less than or equal to 1, and is determined according to the preset adaptive neural network.

[0190] The beneficial effects of this embodiment are as follows: In this embodiment, the current vehicle speed is obtained from driving information, and the current humidity is obtained from meteorological information; based on the current vehicle speed, current humidity, a first threshold, and a first parameter, a second threshold is determined. In the above method, by using the two key variables, current vehicle speed and humidity, which significantly affect the visibility threshold, the fixed visibility threshold is further optimized. The resulting second threshold fully considers the driver's reaction characteristics, achieving a more refined dynamic adjustment of the visibility threshold and further improving the accuracy of visibility classification.

[0191] Below, in conjunction with Figure 3 ,exist Figure 1 Based on the embodiments, another method for classifying visibility based on a second threshold will be described in detail.

[0192] Figure 3 A schematic diagram illustrating the method for classifying visibility provided in this application, such as... Figure 3 As shown, the method includes:

[0193] S301. Based on the current vehicle speed, the current environment is divided into grids to obtain multiple grids within the vehicle's driving range.

[0194] For example, when the current vehicle speed is less than 60 kilometers per hour, the current environment is divided into 500-meter grids to obtain multiple 500-meter grids within the vehicle's driving range; when the current vehicle speed is greater than or equal to 60 kilometers per hour, the current environment is divided into 1000-meter grids to obtain multiple 1000-meter grids within the vehicle's driving range.

[0195] S302. Obtain the visibility corresponding to each grid.

[0196] S303. Based on the second threshold and the visibility corresponding to each grid, classify the visibility of each grid.

[0197] In one implementation, the visibility of each grid is classified based on a second threshold and the visibility corresponding to each grid, including:

[0198] Multiple visibility ranges are determined based on multiple second sub-thresholds;

[0199] Based on multiple visibility ranges, determine the visibility range to which the visibility of each grid belongs;

[0200] Based on the visibility range to which each grid belongs, the visibility of each grid is classified to determine the visibility type of each grid.

[0201] It should be noted that the specific execution process of classifying the visibility of each grid based on the second threshold and the visibility corresponding to each grid is similar to the specific execution process of classifying the visibility based on the second threshold in S104, and will not be repeated here.

[0202] In one implementation, for any two adjacent first and second grids, after classifying the visibility of each grid, the following can be included:

[0203] When a vehicle travels from the range of the first grid to the range of the second grid, determine whether the visibility type of the first grid and the visibility type of the second grid are the same;

[0204] If they are different, a notification message indicating the change in visibility will be generated.

[0205] For example, a visibility change warning message could be: "You are about to enter an area with visibility level 2. It is recommended to turn on your fog lights, low beam headlights, side marker lights, and front and rear position lights. Your speed should not exceed 60 kilometers per hour. Keep a distance of more than 100 meters from the vehicle in front in the same lane."

[0206] In one implementation, after classifying the visibility of each grid, it may also include:

[0207] Determine the color of each grid based on its visibility type;

[0208] A visual map of fog distribution is generated based on the color of each grid cell.

[0209] Optionally, the color of each grid can be retrieved from a pre-stored second correspondence based on the visibility type of each grid. The second correspondence includes multiple visibility types, multiple colors, and correspondences between multiple visibility types and multiple colors.

[0210] For example, the second correspondence can be shown in Table 2:

[0211] Table 2 Second Correspondence

[0212] Visibility type color Level 4 Dark red Level 3 light red Level 2 orange color Level 1 yellow Level 0 green

[0213] For example, when the visibility type of the grid is level 2, the color of the grid can be determined to be orange based on Table 2.

[0214] The beneficial effects of this embodiment are as follows: In this embodiment, based on the current vehicle speed, the current environment is divided into grids to obtain multiple grids within the vehicle's driving range; the visibility corresponding to each grid is obtained; and the visibility of each grid is classified based on a second threshold and the visibility corresponding to each grid. In the above method, the grid division width is dynamically adjusted according to the vehicle speed, which takes into account both the refined perception of environmental details in low-speed scenarios and the efficient analysis needs of long-distance vision when driving at high speeds, significantly improving the spatiotemporal adaptability of visibility data collection. The driving prompts given based on the grid provide a more reliable basis for the driver's traffic decisions. In addition, for any two adjacent first and second grids, when the vehicle travels from the range of the first grid to the range of the second grid, it is determined whether the visibility type of the first grid is the same as that of the second grid. If they are different, a visibility change prompt is generated, enabling the driver to perceive sudden changes in visibility in a timely manner and gain valuable reaction time to adjust the driving strategy. For example, when a vehicle moves from a visibility level 0 grid into a visibility level 2 grid, the driver can slow down and turn on the fog lights in advance based on the prompts, avoiding misjudgment and emergency braking caused by a sudden drop in visibility, thus effectively improving driving safety.

[0215] Below, in conjunction with Figure 4 The process of training an adaptive neural network model to determine the preset adaptive neural network model is explained in detail.

[0216] Figure 4 This is a flowchart illustrating the method for determining a preset adaptive neural network provided in an embodiment of this application, as shown below. Figure 4 As shown, the method includes:

[0217] S401. Determine multiple sets of samples, each set of samples including sample meteorological information, sample driving information, sample first threshold, and sample results.

[0218] The sample results are of the true visibility type.

[0219] Optionally, multiple groups of samples are determined, including:

[0220] Identify the original multiple groups of samples;

[0221] For missing data in the original multi-group samples, interpolation and mean imputation methods are used to obtain complete multi-group samples;

[0222] By using the inverse distance weighted interpolation method, multiple sets of samples are processed to obtain multiple sets of samples.

[0223] Optionally, multiple sub-information items with strong correlation to the first threshold of the sample can be determined from the sample meteorological information and sample driving information by using correlation analysis and / or the feature importance ranking method of XGBoost, thereby determining the final multiple sets of samples.

[0224] S402. Using an adaptive neural network model, multiple sets of training results for this iteration are obtained based on multiple sets of sample meteorological information, sample driving information, and the first threshold of the samples.

[0225] The training result is used to predict the visibility type.

[0226] S403. Based on the multiple training results of this iteration, the loss function is used to iterate repeatedly until the preset number of iterations is reached, and the final multiple training results are determined.

[0227] In one implementation, taking the iteration of the meteorological-threshold weight as an example, the update of the meteorological-threshold weight in the (t+1)th iteration is explained:

[0228] The update of the meteorological threshold weights satisfies the following formula 11:

[0229]

[0230] in, This represents the k-th meteorological-threshold weight in the i-th first sub-weight obtained in the (t+1)-th iteration. η represents the k-th meteorological-threshold weight in the i-th first sub-weight obtained in the t-th iteration. (t+1) This represents the learning rate after the (t+1)th adjustment. This represents the partial derivative of the loss function with respect to the k-th meteorological-threshold weight in the i-th first sub-weight.

[0231] The learning rate is adjusted according to the adaptive learning rate strategy, and can be adjusted based on the convergence of the adaptive learning model. The initial learning rate is determined empirically.

[0232] For example, the loss function is the cross-entropy loss function.

[0233] S404. Based on the final training results and sample results, determine whether the adaptive neural network model meets the preset conditions.

[0234] If yes, execute S406; otherwise, execute S405.

[0235] The following examples 1A, 1B, and 1C illustrate the process of determining whether an adaptive neural network model meets preset conditions based on the final multiple sets of training results and multiple sets of sample results.

[0236] Example 1A: Based on the final training results and sample results, determine whether the adaptive neural network model meets the preset conditions, including:

[0237] Based on the final training results and sample results, the threat score (TS) corresponding to each visibility type is obtained.

[0238] Based on the threat score corresponding to each visibility type, determine whether the adaptive neural network model meets the preset conditions.

[0239] Optionally, based on the final multiple training results and multiple sample results, a threat score corresponding to each visibility type is obtained, including:

[0240] Threat score for a given visibility type;

[0241] Based on the final training results and sample results, determine the number of hits, misses, and false alarms for this visibility type.

[0242] The threat score corresponding to a visibility type is determined based on the number of hits, misses, and false alarms for that visibility type.

[0243] A hit for this visibility type can be exemplified as follows: if the sample result for this group is level 4, and the final training result for this group is also level 4, then it is determined to be a level 4 hit. A miss for this visibility type can be exemplified as follows: if the sample result for this group is level 4, and the final training result for this group is not level 4, then it is determined to be a level 4 miss. A false negative for this visibility type can be exemplified as follows: if the sample result for this group is not level 4, and the final training result for this group is level 4, then it is determined to be a level 4 false positive.

[0244] Optionally, the number of hits, the number of missed detections, the number of false alarms, and the threat score corresponding to the visibility type satisfy the following formula 12:

[0245]

[0246] Where TS represents the threat score corresponding to the visibility type, H represents the number of hits for the visibility type, M represents the number of missed detections for the visibility type, and F represents the number of false alarms for the visibility type.

[0247] The following examples, 2A and 2B, illustrate the process of determining whether an adaptive neural network model meets preset conditions based on the threat score corresponding to each visibility type.

[0248] Example 2A: Based on the threat score corresponding to each visibility type, determine whether the adaptive neural network model meets the preset conditions, including:

[0249] Determine the arithmetic mean of the threat scores for all visibility types to obtain the mean threat score;

[0250] When the mean threat score is less than or equal to the first preset value, it is determined that the adaptive neural network model does not meet the preset conditions;

[0251] When the mean threat score is greater than the first preset value, the adaptive neural network model is determined to meet the preset conditions.

[0252] When determining whether the adaptive neural network model meets the preset conditions based on the threat score corresponding to each visibility type, the first preset value is, for example, 60%.

[0253] Example 2B: Based on the threat score corresponding to each visibility type, determine whether the adaptive neural network model meets the preset conditions, including:

[0254] Based on the threat score corresponding to each visibility type and the second preset value for each visibility type, determine whether the threat score corresponding to each visibility type meets the preset conditions;

[0255] When the threat scores for each visibility type meet the preset conditions, the adaptive neural network model is determined to meet the preset conditions.

[0256] If any one or more of the threat scores corresponding to each visibility type do not meet the preset conditions, the adaptive neural network model is determined to not meet the preset conditions.

[0257] Optionally, the smaller the visibility range value corresponding to a visibility type, the larger the second preset value for that visibility type.

[0258] Based on the threat score corresponding to each visibility type and the second preset value for each visibility type, determine whether the threat score corresponding to each visibility type meets the preset conditions, including:

[0259] For a visibility type;

[0260] When the threat score corresponding to the visibility type is less than or equal to the second preset value of the visibility type, it is determined that the threat score corresponding to the visibility type does not meet the preset condition;

[0261] When the threat score corresponding to a visibility type is greater than the second preset value for that visibility type, the threat score corresponding to the visibility type meets the preset condition.

[0262] Example 1B: Based on the final training results and sample results, determine whether the adaptive neural network model meets the preset conditions, including:

[0263] Based on the final training results and sample results, the F1 score corresponding to each visibility type is obtained.

[0264] Based on the F1 score corresponding to each visibility type, determine whether the adaptive neural network model meets the preset conditions.

[0265] Optionally, based on the final training results and sample results, the F1 score corresponding to each visibility type is obtained, including:

[0266] The F1 score for a given visibility type;

[0267] Based on the final training results and sample results, determine the number of hits, misses, and false alarms for this visibility type.

[0268] The accuracy rate corresponding to a visibility type is determined based on the number of hits and false alarms for that visibility type.

[0269] The recall rate corresponding to a visibility type is determined based on the number of hits and the number of misses for that visibility type.

[0270] The F1 score for each visibility type is determined based on its precision and recall.

[0271] Optionally, the number of hits, the number of false alarms, and the accuracy corresponding to the visibility type satisfy the following formula 13:

[0272]

[0273] Precision refers to the accuracy rate corresponding to this visibility type.

[0274] Optionally, the number of hits, the number of misses, and the recall rate corresponding to the visibility type satisfy the following formula 14:

[0275]

[0276] Here, Recall represents the recall rate corresponding to this visibility type.

[0277] Optionally, the precision, recall, and F1 score corresponding to the visibility type satisfy the following formula 15:

[0278]

[0279] Here, F1 represents the F1 score corresponding to this visibility type.

[0280] It should be noted that the process of determining whether the adaptive neural network model meets the preset conditions based on the F1 score corresponding to each visibility type is similar to the process of determining whether the adaptive neural network model meets the preset conditions based on the threat score corresponding to each visibility type, and will not be described in detail here.

[0281] When determining whether the adaptive neural network model meets the preset conditions based on the threat score corresponding to each visibility type, the first preset value is, for example, 50%.

[0282] Example 1C: Based on the final training results and sample results, determine whether the adaptive neural network model meets the preset conditions, including:

[0283] Based on the final training results and sample results, the Receiver Operating Characteristic Curve (ROC) is obtained.

[0284] Based on the ROC curve, determine the area under the curve (AUC).

[0285] When the area under the curve is greater than the third preset value, the adaptive neural network model is determined to meet the preset conditions.

[0286] When the area under the curve is less than or equal to the third preset value, the adaptive neural network model is determined not to meet the preset conditions.

[0287] Optionally, based on the final training results and sample results from multiple sets, an ROC curve is obtained, including:

[0288] For a given visibility type, based on the final training results and sample results, determine the number of hits, misses, false positives, and hits for non-visibility types for that visibility type. Based on the number of hits and misses for that visibility type, determine the True Positive Rate (TPR) for that visibility type. Based on the number of false positives and hits for non-visibility types for that visibility type, determine the False Positive Rate (FPR) for that visibility type. Based on the True Positive Rate and False Positive Rate for that visibility type, determine the curve points for that visibility type.

[0289] Plot the ROC curve based on the curve points corresponding to each visibility type.

[0290] For example, the true positive rate corresponding to the visibility type is 0.83, the false positive rate corresponding to the visibility type is 0.21, and the curve point corresponding to this visibility type is (0.21, 0.83).

[0291] The third preset value is, for example, 0.8.

[0292] In one implementation, Example 1B and Example 1C can be combined to determine whether the adaptive neural network model meets the preset conditions, specifically:

[0293] If it is determined from both Example 1B and Example 1C that the adaptive neural network model meets the preset conditions, then it is finally determined that the adaptive neural network model meets the preset conditions.

[0294] Otherwise, it will be determined that the adaptive neural network model does not meet the preset conditions.

[0295] S405. After modifying the parameter values ​​of the adaptive neural network model, execute S402.

[0296] Optionally, the parameter values ​​of the adaptive neural network model include, but are not limited to, the number of attention heads, the hidden layer dimension, the initial learning rate, the decay factor, and the dynamic weight coefficients.

[0297] S406. Set the current adaptive neural network model as the preset adaptive neural network model.

[0298] Furthermore, the adaptive neural network model is trained and the preset adaptive neural network model is verified using a cross-validation method.

[0299] Cross-validation methods include, for example, K-fold cross-validation or time series cross-validation.

[0300] The beneficial effects of this embodiment are as follows: In related technologies, adaptive neural network models are trained based on root mean square error (RMSE) and bias. RMSE, as a measure based on the square of the error, has a strong amplification effect on errors with large values. When processing visibility data with a large range, it will generate high-weight error feedback due to small fluctuations in high visibility scenarios, causing the model training to be overly biased towards high visibility data, which weakens the learning of key features of low visibility data. Bias only focuses on the average difference direction between the forecast value and the actual value. It cannot quantify the severity of the error in a single prediction, nor can it distinguish the degree of error impact under different visibility conditions. Especially in scenarios such as low visibility, which have a decisive impact on traffic safety, it is easy to misjudge the model performance by ignoring the magnitude of the error, making it difficult for the model to accurately cope with complex weather conditions in practical applications. In this embodiment, an adaptive neural network model is trained based on threat score, FI score, and / or ROC curve, effectively avoiding the limitations of evaluation indicators (root mean square error and bias) in related technologies. Threat score, FI score, and ROC curve can accurately measure the forecast effectiveness of the preset adaptive neural network model under different visibility types. Especially in scenarios where safety decisions are critical, such as low visibility, the model can intuitively reflect the ability of the preset adaptive neural network model to capture key weather changes, thereby improving the accuracy of visibility classification.

[0301] Figure 5 This is a schematic diagram of a visibility processing device provided in an embodiment of this application. Figure 5 As shown, the visibility processing device 50 includes an acquisition module 501, a determination module 502, and a processing module 503.

[0302] The acquisition module 501 is used to acquire meteorological information, vehicle driving information, and a first threshold corresponding to visibility.

[0303] The determining module 502 is used to determine a first weight corresponding to the meteorological information and a second weight corresponding to the driving information. The first weight is used to indicate the degree of influence of the meteorological information on the visibility, and the second weight is used to indicate the degree of influence of the driving information on the visibility.

[0304] The determining module 502 is further configured to determine a first parameter based on the meteorological information, the driving information, the first weight, and the second weight;

[0305] The processing module 503 is used to determine a second threshold corresponding to the visibility based on the first threshold and the first parameter, and to classify the visibility based on the second threshold.

[0306] The visibility processing device 50 provided in this application embodiment can execute the technical solution shown in the above method embodiment. Its implementation principle and beneficial effects are similar, and will not be described again here.

[0307] In one implementation, module 502 is specifically used for:

[0308] In the first weight, the first sub-weight corresponding to each sub-meteorological information is determined;

[0309] In the second weighting, the second sub-weight corresponding to each sub-driving information is determined;

[0310] The first parameter is determined based on the multiple sub-meteorological information, multiple first sub-weights, the multiple sub-driving information, and multiple second sub-weights.

[0311] In one implementation, module 502 is specifically used for:

[0312] The product of the sub-meteorological information and the first sub-weight corresponding to the sub-meteorological information is determined to obtain multiple second parameters;

[0313] The product of the sub-driving information and the second sub-weight corresponding to the sub-driving information is determined to obtain multiple third parameters;

[0314] The first parameter is determined based on the plurality of second parameters and the plurality of third parameters.

[0315] In one implementation, the processing module 503 is specifically used for:

[0316] The current vehicle speed is obtained from the driving information, and the current humidity is obtained from the weather information;

[0317] The second threshold is determined based on the current vehicle speed, the current humidity, the first threshold, and the first parameter.

[0318] In one implementation, the processing module 503 is specifically used for:

[0319] A third threshold is determined based on the current vehicle speed and the current humidity.

[0320] Based on the first threshold and the first parameter, a fourth threshold is determined;

[0321] Based on the third threshold and the fourth threshold, a second threshold corresponding to the visibility is determined.

[0322] In one implementation, the processing module 503 is specifically used for:

[0323] Based on the current vehicle speed, the current environment is divided into grids to obtain multiple grids within the vehicle's driving range;

[0324] Obtain the visibility corresponding to each grid;

[0325] Based on the second threshold and the visibility corresponding to each grid, the visibility of each grid is classified.

[0326] In one implementation, for any two adjacent first and second grids, the processing module 503 is further configured to:

[0327] When the vehicle travels from the range of the first grid to the range of the second grid, it is determined whether the visibility type of the first grid and the visibility type of the second grid are the same;

[0328] If they are different, a notification message indicating the change in visibility will be generated.

[0329] The visibility processing device 50 provided in this application embodiment can execute the technical solution shown in the above method embodiment. Its implementation principle and beneficial effects are similar, and will not be described again here.

[0330] Figure 6 This is a structural diagram of an electronic device provided in an embodiment of this application. As shown, the electronic device 60 includes a processor 601 and a memory 602. The processor 601 is communicatively connected to the memory 602, and the memory 602 is used to store computer execution instructions. The processor 601 is configured to execute the technical solutions in any of the aforementioned method embodiments by executing the computer execution instructions stored in the memory 602.

[0331] Optionally, the memory 602 can be either standalone or integrated with the processor 601. Optionally, when the memory 602 is a device independent of the processor 601, the electronic device 60 may further include a bus 603 for connecting the aforementioned devices.

[0332] The electronic device is used to execute the technical solutions in any of the foregoing method embodiments. Its implementation principle and technical effect are similar, and will not be described again here.

[0333] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the technical solutions provided in any of the foregoing method embodiments.

[0334] This application also provides a computer program product, including a computer program, which, when executed by a processor, is used to implement the technical solutions provided in the foregoing method embodiments.

[0335] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily essential to this application.

[0336] It should be further noted that although the steps in the flowchart are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowchart may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0337] It should be understood that the above-described device embodiments are merely illustrative, and the device of this application can also be implemented in other ways. For example, the division of units / modules in the above embodiments is only a logical functional division, and there may be other division methods in actual implementation. For example, multiple units, modules, or components may be combined, or integrated into another system, or some features may be ignored or not executed.

[0338] Furthermore, unless otherwise specified, the functional units / modules in the various embodiments of this application can be integrated into one unit / module, or each unit / module can exist physically separately, or two or more units / modules can be integrated together. The integrated units / modules described above can be implemented in hardware or as software program modules.

[0339] When integrated units / modules are implemented in hardware, the hardware can be digital circuits, analog circuits, etc. The physical implementation of the hardware structure includes, but is not limited to, transistors, memristors, etc. Unless otherwise specified, the processor can be any suitable hardware processor, such as a CPU, GPU, FPGA, DSP, and ASIC, etc. Unless otherwise specified, the storage unit can be any suitable magnetic or magneto-optical storage medium, such as Resistive Random Access Memory (RRAM), Dynamic Random Access Memory (DRAM), Static Random Access Memory (SRAM), Enhanced Dynamic Random Access Memory (EDRAM), High-Bandwidth Memory (HBM), Hybrid Memory Cube (HMC), etc.

[0340] If the integrated unit / module is implemented as a software program module and sold or used as an independent product, it can be stored in a computer-readable storage device (CMD). Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned memory includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.

[0341] In the above embodiments, the descriptions of each embodiment have their own emphasis. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments. The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as these combinations of technical features do not contradict each other, they should be considered within the scope of this specification.

[0342] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this application are indicated by the following claims.

[0343] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

Claims

1. A visibility processing method, characterized in that, include: Acquire meteorological information, vehicle driving information, and a first threshold corresponding to visibility. The meteorological information includes multiple sub-meteorological information, the driving information includes multiple sub-driving information, and the first threshold includes multiple first sub-thresholds. By using a preset adaptive neural network model, a first weight corresponding to the meteorological information and a second weight corresponding to the driving information are determined. The first weight is used to indicate the degree of influence of the meteorological information on the visibility, and the second weight is used to indicate the degree of influence of the driving information on the visibility. The first weight includes the first sub-weight corresponding to each sub-meteorological information, and the second weight includes the second sub-weight corresponding to each sub-driving information. The product of the sub-meteorological information and the first sub-weight corresponding to the sub-meteorological information is determined to obtain multiple second parameters, and the product of the sub-driving information and the second sub-weight corresponding to the sub-driving information is determined to obtain multiple third parameters; The first parameter is determined by the preset adaptive neural network model based on the plurality of second parameters, the plurality of third parameters, the preset bias term corresponding to each first sub-threshold, and the scaling factor corresponding to each first sub-threshold. The first parameter includes the first sub-parameter corresponding to each first sub-threshold. Based on the first threshold and the first parameter, a second threshold corresponding to the visibility is determined, and the visibility is classified based on the second threshold. The second threshold includes multiple second sub-thresholds.

2. The method according to claim 1, characterized in that, Determining a second threshold corresponding to the visibility based on the first threshold and the first parameter includes: The current vehicle speed is obtained from the driving information, and the current humidity is obtained from the weather information; The second threshold is determined based on the current vehicle speed, the current humidity, the first threshold, and the first parameter.

3. The method according to claim 2, characterized in that, Determining the second threshold based on the current vehicle speed, current humidity, the first threshold, and the first parameter includes: A third threshold is determined based on the current vehicle speed and the current humidity. Based on the first threshold and the first parameter, a fourth threshold is determined; Based on the third threshold and the fourth threshold, a second threshold corresponding to the visibility is determined.

4. The method according to claim 1, characterized in that, The visibility is classified based on the second threshold, including: Based on the current vehicle speed, the current environment is divided into grids to obtain multiple grids within the vehicle's driving range; Obtain the visibility for each grid cell; Based on the second threshold and the visibility corresponding to each grid, the visibility of each grid is classified.

5. The method according to claim 4, characterized in that, For any two adjacent first and second grids; after classifying the visibility of each grid, the method further includes: When the vehicle travels from the range of the first grid to the range of the second grid, it is determined whether the visibility type of the first grid and the visibility type of the second grid are the same; If they are different, a notification message indicating the change in visibility will be generated.

6. A visibility processing device, characterized in that, include: The acquisition module is used to acquire meteorological information, vehicle driving information, and a first threshold corresponding to visibility. The meteorological information includes multiple sub-meteorological information, the driving information includes multiple sub-driving information, and the first threshold includes multiple first sub-thresholds. The determination module is used to determine a first weight corresponding to the meteorological information and a second weight corresponding to the driving information through a preset adaptive neural network model. The first weight is used to indicate the degree of influence of the meteorological information on the visibility, and the second weight is used to indicate the degree of influence of the driving information on the visibility. The first weight includes the first sub-weight corresponding to each sub-meteorological information, and the second weight includes the second sub-weight corresponding to each sub-driving information. The determining module is further configured to determine the product of the sub-meteorological information and the first sub-weight corresponding to the sub-meteorological information to obtain multiple second parameters, and to determine the product of the sub-driving information and the second sub-weight corresponding to the sub-driving information to obtain multiple third parameters; The determining module is further configured to determine a first parameter based on the plurality of second parameters, the plurality of third parameters, the preset bias term corresponding to each first sub-threshold, and the scaling factor corresponding to each first sub-threshold through the preset adaptive neural network model, wherein the first parameter includes the first sub-parameter corresponding to each first sub-threshold; The processing module is configured to determine a second threshold corresponding to the visibility based on the first threshold and the first parameter, and classify the visibility based on the second threshold, wherein the second threshold includes a plurality of second sub-thresholds.

7. An electronic device, characterized in that, include: Memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1-5.

Images