Methods, devices, media and equipment for early warning of safe driving of vehicles under adverse conditions

Abstract

This invention discloses a method, device, medium, and equipment for providing early warning of safe driving in adverse environments. The method includes the following steps: acquiring real-time environmental images and real-time driving data through the onboard equipment of the target vehicle; calling a pre-trained weather classification model to generate the current weather type of the target road; generating the current road surface friction coefficient and current visibility; generating an optimal safe speed limit based on the current road surface friction coefficient, current visibility, and real-time driving data; and generating a safe driving warning scheme based on the optimal safe speed limit. This invention can generate the current road surface friction coefficient and current visibility based on the real-time environmental images of the target road, calculate and generate the optimal safe speed limit based on the above data and the real-time driving data of the target vehicle, and then generate a current safe driving warning scheme for the target vehicle based on the optimal safe speed limit. This eliminates the need to rely on the update frequency and range of regional meteorological information, thereby efficiently assisting drivers in driving safely in adverse environments.

Description

Technical Field

[0001] This invention relates to the field of vehicle driver assistance, and in particular to methods, devices, media and equipment for early warning of safe driving of vehicles under adverse conditions. Background Technology

[0002] Adverse weather conditions such as rain, fog, ice, and snow cause changes in road surface temperature and humidity, leading to a decrease in the road surface friction coefficient and adhesion coefficient. This, in turn, reduces the friction and adhesion between the wheels and the road surface, resulting in reduced vehicle braking performance. Furthermore, it severely affects the driver's perception of the driving environment; for example, in foggy or hazy weather, the driver's field of vision is limited by visibility. The deterioration of driving conditions and the reduction in the driver's perception accuracy can easily lead to an imbalance in driving stability, triggering incorrect driving decisions such as speeding and improper braking, resulting in widespread traffic congestion and serious traffic accidents. To address driving safety issues in adverse weather conditions, traffic management departments currently primarily guide driving through section speed limits. The calculation of section speed limits relies on the published regional meteorological information, and the update range and frequency of this information depend on the distribution of meteorological monitoring stations and the frequency of information collection.

[0003] While interval-based vehicle speed limiting methods offer real-time performance, accuracy, and granular information for macroscopic control of passing vehicles, they struggle to precisely control each individual vehicle. Therefore, achieving a balance between safety and efficiency in terms of control precision is difficult. Existing vehicle-based speed warning systems are developed solely from the perspective of vehicle dynamics, with limited integration of environmental factors such as road surface friction coefficient and weather visibility. Consequently, several technical challenges remain to be addressed. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to overcome the defects of the prior art and provide a method, device, medium and equipment for early warning of safe driving of vehicles under adverse conditions.

[0005] To address the aforementioned technical problems, the first aspect of the present invention provides the following technical solution:

[0006] A method for early warning of safe vehicle operation under adverse conditions includes the following steps:

[0007] Real-time environmental images and real-time driving data of the target road are obtained through the onboard equipment of the target vehicle.

[0008] Invoke a pre-trained weather classification model and generate the current weather type of the target road based on the real-time environmental image;

[0009] Generate the current road surface friction coefficient and current visibility of the target road based on the current weather type;

[0010] The optimal safe speed limit for the target vehicle is generated based on the current road surface friction coefficient, the current visibility, and the real-time driving data.

[0011] Based on the optimal safe speed limit, a safe driving warning scheme for the target vehicle on the target road is generated.

[0012] Beneficial effects: This invention can generate the current road surface friction coefficient and current visibility based on the real-time environmental image of the target road, and calculate and generate the optimal safe speed limit based on the above data and the real-time driving data of the target vehicle. Then, it generates the current safe driving warning scheme for the target vehicle based on the optimal safe speed limit. It does not rely on the update frequency and update range of regional meteorological information, and at the same time, it has higher real-time performance, accuracy and safety, thus efficiently assisting drivers to drive safely in adverse environments.

[0013] As a preferred embodiment of the present invention, generating the current road surface friction coefficient of the target road based on the current weather type includes the following sub-steps:

[0014] Query the preset first mapping relationship table to generate the target precipitation intensity corresponding to the current weather type. The first mapping relationship table includes preset precipitation intensities corresponding to different weather types.

[0015] Obtain road information for the target road, and calculate the current road surface water film thickness based on the target precipitation intensity and the road information;

[0016] The preset second mapping table is queried to generate the target pendulum meter reading corresponding to the current road surface water film thickness. The second mapping table includes preset pendulum meter readings corresponding to different road surface water film thicknesses.

[0017] The current road surface friction coefficient of the target road is calculated based on the readings of the target pendulum instrument.

[0018] Further benefits of the technical solution: This invention obtains the current road surface friction coefficient by measuring the road surface water film thickness under current weather conditions, which can more accurately calculate the safe speed limit of the target vehicle, so as to generate a more reasonable safe driving warning scheme.

[0019] As a preferred embodiment of the present invention, generating the current road surface friction coefficient of the target road based on the current weather type includes the following sub-steps:

[0020] The preset third mapping relationship table is queried to generate the current road surface friction coefficient corresponding to the current weather type. The third mapping relationship table includes preset road surface friction coefficients corresponding to different weather types.

[0021] Further beneficial effects of the technical solution: The present invention pre-establishes a third mapping relationship table based on historical experience data, which includes the correspondence between weather type and road surface friction coefficient. By querying the pre-established third mapping relationship table, the road surface friction coefficient can be quickly obtained, enabling devices with insufficient computing power to perform rapid calculations.

[0022] As a preferred embodiment of the present invention, generating the current visibility of the target road based on the current weather type includes the following sub-steps:

[0023] When the current weather type is rainy, the target precipitation intensity corresponding to the real-time environmental image is obtained, and the current visibility of the target road is generated based on the target precipitation intensity.

[0024] When the current weather type is non-rainy, a preset fog detection method based on image color space features is used to classify the current weather, generate a fog classification result, and generate the current visibility of the target road based on the fog classification result.

[0025] Further benefits of the technical solution: The present invention can determine the current visibility under different weather types, and adopts a targeted judgment method for rainy and non-rainy days, which is not only efficient but also highly accurate, and can more accurately calculate the safe speed limit of the target vehicle.

[0026] As a preferred embodiment of the present invention, the fog classification results include non-foggy days, light foggy days, and heavy foggy days. If the current weather is heavy foggy, generating the current visibility of the target road includes the following sub-steps:

[0027] At least one initial horizon in the real-time environment image is marked using a preset horizon detection algorithm;

[0028] The real-time environment image is binarized, and the binarization result is further processed based on a preset region growing method to mark at least one initial fog boundary line in the real-time environment image.

[0029] A preset lane line detection algorithm is used to mark the lane lines in the real-time environment image, and at least one determination line is set at the end of the lane line, and the at least one determination line forms a determination line area.

[0030] Obtain at least one initial horizon line and at least one initial fog boundary line, and acquire a target horizon line and a target fog boundary line that are located in the determination line region;

[0031] The current visibility of the target road is generated based on the first number of pixels of the target horizon and the second number of pixels of the target fog boundary.

[0032] Further advantages of the technical solution: The present invention adopts a visibility detection method based on three auxiliary lines: the horizon, the fog-ground boundary line, and the judgment line. By calculating the positional difference between the horizon and the fog-ground boundary line, the current visibility in foggy weather can be calculated in real time and accurately. It has high accuracy and wide applicability.

[0033] As a preferred embodiment of the present invention, the real-time driving data includes at least the first real-time driving speed, vehicle length, and first coordinate position of the nearest vehicle ahead, and the second real-time driving speed and second coordinate position of the target vehicle; generating the optimal safe speed limit for the target vehicle based on the current road surface friction coefficient, current visibility, and real-time driving data includes the following steps:

[0034] The first preset formula is invoked, and the initial safe speed limit of the target vehicle is calculated based on the current road surface friction coefficient and the current visibility.

[0035] The second preset formula is invoked, and the potential collision time between the target vehicle and the nearest vehicle ahead is calculated based on the real-time driving data and the initial safe speed limit.

[0036] Determine whether the potential conflict time meets the preset high-risk conditions. If not, the initial safe speed limit is taken as the optimal safe speed limit. If so, the third preset formula is called, and the alternative safe speed limit of the target vehicle is calculated based on the real-time driving data and the preset reaction time. The alternative safe speed limit is taken as the optimal safe speed limit.

[0037] Further benefits of the technical solution: The present invention pre-determines multiple mathematical formulas to calculate potential conflict times, and obtains the optimal safe speed limit by judging the numerical value of potential conflict times.

[0038] As a preferred embodiment of the present invention, the method for generating a safe driving warning scheme for the target vehicle on the target road based on the optimal safe speed limit specifically includes:

[0039] The second real-time driving speed of the target vehicle is obtained. When the second real-time driving speed is greater than the optimal safe speed limit, a first warning signal is generated and displayed with sound and light.

[0040] And / or acquire the real-time yaw rate of the target vehicle, and when the real-time yaw rate is greater than a preset yaw rate threshold, generate and display the corresponding second warning signal with sound and light.

[0041] Further benefits of the technical solution: The technical solution of the present invention can generate different warning signals according to different situations, such as driving speed and yaw rate, so as to more comprehensively and effectively protect the driver's driving safety in adverse environments.

[0042] A second aspect of the present invention also provides a vehicle safety driving warning device under adverse conditions, comprising a data acquisition module, a first generation module, a second generation module, an analysis module, and a warning module.

[0043] The acquisition module is used to acquire real-time environmental images and real-time driving data of the target road through the on-board equipment of the target vehicle;

[0044] The first generation module is used to call a pre-trained weather classification model and generate the current weather type of the target road based on the real-time environmental image;

[0045] The second generation module is used to generate the current road surface friction coefficient and current visibility of the target road according to the current weather type;

[0046] The analysis module is used to generate the optimal safe speed limit for the target vehicle based on the current road surface friction coefficient, the current visibility, and the real-time driving data.

[0047] The early warning module is used to generate a safe driving warning scheme for the target vehicle on the target road based on the optimal safe speed limit.

[0048] A third aspect of the present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the above-described method for warning safe driving of vehicles under adverse conditions.

[0049] A fourth aspect of the present invention also provides a vehicle safety driving warning device under adverse conditions, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the vehicle safety driving warning method under adverse conditions described above.

[0050] The objectives and other advantages of this application can be realized and obtained by means of the structures specifically pointed out in the written description, claims, and drawings. Attached Figure Description

[0051] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0052] Figure 1 This is a flowchart illustrating the steps of an embodiment of the vehicle safe driving warning method under adverse conditions according to the present invention;

[0053] Figure 2 This is a flowchart illustrating the steps of an embodiment of the present invention for generating the current road surface friction coefficient;

[0054] Figure 3 This is a flowchart illustrating another embodiment of the present invention for generating the current road surface friction coefficient;

[0055] Figure 4 This is a flowchart illustrating the steps of an embodiment of the method for generating current visibility according to the present invention;

[0056] Figure 5 This is a flowchart illustrating the steps of another embodiment of the method for generating current visibility according to the present invention;

[0057] Figure 6 This is a flowchart illustrating the steps of an embodiment of the method for generating the optimal safe speed limit according to the present invention;

[0058] Figure 7 This is a flowchart illustrating the steps of an embodiment of the method for generating a safe driving warning scheme according to the present invention;

[0059] Figure 8 This is a structural block diagram of a vehicle safety driving warning device under adverse conditions according to the present invention;

[0060] Figure 9 This is a structural block diagram of a vehicle safety driving early warning device under adverse conditions according to the present invention;

[0061] Figure 10 This is one of the schematic diagrams of the image effect after image binarization processing according to an embodiment of the present invention;

[0062] Figure 11 This is one of the schematic diagrams illustrating the image growth method processing effect of an embodiment of the present invention;

[0063] Figure 12 This is the second image effect after image binarization processing according to an embodiment of the present invention;

[0064] Figure 13 This is the second schematic diagram illustrating the image growth method processing effect according to an embodiment of the present invention;

[0065] Figure 14 This is a schematic diagram of the maximum safe speed model of the present invention. Detailed Implementation

[0066] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0067] In the attached diagram, all identical reference numerals refer to the same components.

[0068] like Figure 1 As shown, one embodiment of the present invention provides a method for warning of safe vehicle driving under adverse conditions, including the following steps:

[0069] Step 1: Acquire real-time environmental images and real-time driving data of the target road using the onboard equipment of the target vehicle. Specifically, real-time environmental images and real-time driving data around the target vehicle are collected using pre-installed onboard equipment for subsequent analysis. In embodiments of the present invention, depending on the pre-installed onboard equipment, the real-time environmental images can be real-time environmental images in front of and behind the target vehicle, or real-time environmental images in all directions around the target vehicle; the real-time driving data includes at least the target vehicle's real-time acceleration, real-time speed, and real-time yaw rate, etc.

[0070] Then, step 2 is executed: the pre-trained weather classification model is invoked, and the current weather type of the target road is generated based on the real-time environmental image. Specifically, the pre-trained weather classification model is used to process the acquired real-time environmental image to generate the current weather type of the target road, providing support for the subsequent generation of road surface friction coefficient and visibility. In this embodiment of the invention, the pre-trained weather classification model needs to be invoked to determine the real-time weather around the target vehicle. This determination is divided into two parts: determining whether the real-time weather is rainy or foggy, which is done by setting rainy weather determination models and foggy weather determination models respectively.

[0071] In this specific embodiment, the rainy day judgment model is used to determine whether the weather type is rainy. The establishment of the rainy day judgment model involves two steps: First, a meteorological classification dataset is established. Specifically, to obtain clear and distinguishable image data, images are taken using the vehicle's rearview mirror, which can be called rearview data acquisition. This acquisition method effectively avoids color difference changes and distance stretching effects caused by the windshield, reducing image noise. Second, the Darknet convolutional neural network algorithm is used for training. Data collected under different weather conditions and road conditions is then categorized into five types—sunny, light rain, moderate rain, heavy rain, and torrential rain—using LabelImage. 80% of the data is used as the training set, and 20% is used as the validation set. The Darknet network is used to train the data, and the trained weight file and network are then embedded into the YOLOv4 vision model to train the rainy day judgment model in this embodiment.

[0072] The fog weather identification model is used to determine whether the weather type is foggy. The fog weather identification model classifies fog based on HSV features. Specifically, comparative experiments show that the H (hue) component of clear weather images is concentrated in the 100-110 range and is generally higher than that of foggy weather images; the S (saturation) component of clear weather images is mainly distributed in the 100-150 range, while the S component of foggy weather images is concentrated in the 0-50 range due to the blurriness of the foggy images; both clear weather and foggy weather images show a bimodal phenomenon in the V (brightness) component, indicating that the roadbed has a significant impact on the brightness component. Based on these patterns, an HSV-based visibility classification and detection model can be created. The specific method for classifying fog weather types based on the HSV model is as follows: when the acquired environmental image is a video, the video is decomposed into images. Based on the different color information of each image, different H, S, and V components are extracted and visibility levels are determined. Each pixel is composed of three components: H, S, and V. By traversing each pixel of the image, the sum of the feature components is obtained. Based on existing research results, the corresponding calculation formulas (1)-(3) are shown below:

[0073]

[0074] In the above calculation formula: For the entire image The sum of component values, ; This represents the number of pixels whose corresponding component value is not 0. The length and width of the image; This is the mean of the components.

[0075] The classification criteria for foggy days are based on specific determinations of heavy fog, light fog, and fog-free weather. First, the calculation is performed using formula (1-3). , and Then, based on the judgment condition in the calculation formula (4), determine whether the heavy fog condition is met. If the conditions are met, it is considered heavy fog; otherwise, it is determined whether the conditions for fog-free conditions are met. If the conditions are met, it is considered a non-foggy day; otherwise, it is determined whether it meets the criteria for light fog. The criteria for determining fog are as follows: if the criteria are met, it is considered light fog; otherwise, it is considered heavy fog.

[0076]

[0077] Then proceed to step 3: Generate the current road surface friction coefficient and current visibility of the target road based on the current weather type.

[0078] In a preferred embodiment, such as Figure 2 As shown, generating the current road surface friction coefficient of the target road based on the current weather type includes the following sub-steps:

[0079] Step 311: Query the preset first mapping table to generate the target precipitation intensity corresponding to the current weather type. The first mapping table includes preset precipitation intensities corresponding to different weather types. In a specific embodiment of the present invention, the current weather type determined by the pre-trained weather classification model in step 2 is input into the preset first mapping table for querying, thereby obtaining the preset precipitation intensity corresponding to the target road.

[0080] First mapping table: Precipitation intensity corresponding to different weather types

[0081] Weather Classification Rainfall intensity (mm / min) sunny 0 Light rain 0.8~1.2 Moderate rain 1.3~2.0 heavy rain 2.1~3.0 rainstorm >3.0

[0082] Step 312: Obtain road information of the target road, and calculate the current road surface water film thickness based on the target precipitation intensity and road information. In a specific embodiment of the present invention, the on-board equipment on the target vehicle obtains the road information of the target road, thereby obtaining the road surface construction depth of the target road, and then calculates the water film thickness under different road environments and rainfall environments using calculation formula (5).

[0083]

[0084] In formula (5): h The thickness of the water film on the road surface (in millimeters); l The slope length (in meters) is obtained from the vehicle-mounted equipment. i The road surface slope (%) is obtained from the on-board equipment. q The current precipitation intensity (mm / min) is obtained by querying the preset first mapping table; TD The construction depth (in millimeters) of the target road is obtained by the onboard equipment.

[0085] Step 313: Query the preset second mapping table to generate the target pendulum meter reading corresponding to the current road surface water film thickness. The second mapping table includes preset pendulum meter readings corresponding to different road surface water film thicknesses. In a specific embodiment of the present invention, the current road surface water film thickness generated in step 312 is input into the preset second mapping table for querying, thereby obtaining the corresponding preset pendulum meter reading (BPM value).

[0086] Second mapping table: Correspondence between water film thickness and road surface friction coefficient

[0087]

[0088] Step 314: Calculate the current road surface friction coefficient of the target road based on the target pendulum meter reading. In a specific embodiment of the present invention, the corresponding preset pendulum meter reading (BPM value) obtained in step 313 is substituted into the calculation formula (6):

[0089]

[0090] in: This refers to the road friction coefficient, while BPN is the reading from the pendulum meter. The current road surface friction coefficient for the target road under the current weather conditions can be calculated.

[0091] In other embodiments, a third mapping table can be directly established to more conveniently obtain the current road surface friction coefficient. For example... Figure 3 As shown, generating the current road surface friction coefficient of the target road based on the current weather type includes the following sub-steps:

[0092] Step 315: Query the preset third mapping table to generate the current road surface friction coefficient corresponding to the current weather type. The third mapping table includes preset road surface friction coefficients corresponding to different weather types. In a specific embodiment of the present invention, the current weather type determined by the pre-trained weather classification model in step 2 is input into the preset third mapping table for querying, thereby obtaining the preset road surface friction coefficient corresponding to the target road.

[0093] Third mapping table: Road friction coefficients corresponding to different weather conditions

[0094] Weather type Road surface friction coefficient (μ) sunny 0.75 Light rain 0.59 Moderate rain 0.55 heavy rain 0.49 rainstorm 0.45

[0095] In another specific embodiment, such as Figure 4 As shown, generating the current visibility of the target road based on the current weather type includes the following sub-steps:

[0096] Step 321: When the current weather type is rainy, obtain the target precipitation intensity corresponding to the real-time environmental image, and generate the current visibility of the target road based on the target precipitation intensity. In a specific embodiment of the present invention, the corresponding target precipitation intensity obtained in the previous step is input into a preset fourth mapping relationship table for querying, thereby obtaining the corresponding current visibility of the target road. This embodiment of the present invention establishes a visibility relationship table under different rainfall intensities based on the observation data "Hourly Value Data of National Ground Stations in China" from the China Meteorological Data Network.

[0097] Fourth mapping table: Correspondence between rainfall intensity and visibility interval values

[0098] Rainfall intensity per minute (mm / min) <0.8 0.8-1.2 1.2-1.6 1.6-2.0 Visibility range (m) 1000-500 350-250 250-175 175-100

[0099] Step 322: When the current weather type is non-rainy, a preset fog detection method based on image color space features is used to classify the current weather, generating a fog classification result, and generating the current visibility of the target road based on the fog classification result. In a specific embodiment of the present invention, the obtained fog classification result is input into a preset fifth mapping relationship table for querying. The obtained fog classification result includes non-foggy days, light fog days, and heavy fog days. When the weather conditions are light fog and non-foggy, the visibility value can be directly obtained.

[0100] Fifth mapping table: Correspondence between fog weather classification and visibility range values

[0101] Weather conditions Dense fog Light mist Non-fog Visibility (m) 0-500 500-4000 >4000

[0102] If the current weather is foggy, generating the current visibility of the target road includes the following sub-steps:

[0103] Step 3221: At least one initial horizon in the real-time environment image is marked using a preset horizon detection algorithm. In a specific embodiment of the present invention, the preset horizon detection algorithm is specifically an image edge method, where the horizon is an abrupt edge line between the ground and sky regions. Regardless of whether the sky region on one side of the horizon has uniform grayscale or linear smooth changes, the abrupt edge between the sky region and the ground region always exists. Therefore, there is a clear grayscale jump between the two regions, which can be obtained through image edge detection. The present invention uses a horizon detection algorithm under low visibility conditions in the prior art to detect the horizon in the video frame image of the vehicle's forward-facing camera and marks at least one initial horizon in the image.

[0104] Step 3222 involves binarizing the real-time environment image and then performing a secondary processing on the binarization result based on a preset region growing method to mark at least one initial fog boundary line in the real-time environment image. In a specific embodiment of the present invention, binarizing the real-time environment image yields the following result: Figure 10 The black and white boundary lines shown are then used to process the binarized image using the region growing method to obtain... Figure 11 The black and white border shown is the initial fog boundary line.

[0105] Step 3223: Lane lines in the real-time environment image are marked using a preset lane line detection algorithm, and at least one decision line is set at the end of each lane line, forming a decision line region. In a specific embodiment of the invention, edges in the real-time environment image are first detected using Canny edge detection, for example... Figure 12As shown in the figure, the Hough transform can be used to separate geometric shapes with the same characteristics from all edge pixels in the above image. Then, straight lines are filtered, and neighborhood detection is performed on each pixel. When there are more than 500 consecutive red pixels, the straight line can be identified as a lane line, thus determining the specific functional form and location of the lane line. For example... Figure 13 As shown in the figure. In a specific embodiment of the present invention, due to the influence of surrounding mountain buildings, trees and other objects, there are serious errors in the initial fog boundary line (multiple initial fog boundary lines will be generated). In this case, a decision line is used to determine the actual fog boundary line. The decision line is a vertical line located between the two edge lines of the vehicle driving road, that is, between the two lane lines and perpendicular to the horizon. Once the lane lines are determined, the decision line area can be generated through the lane lines.

[0106] Step 3224: Obtain the target horizon and target fog boundary line that are located in the decision line area from at least one initial horizon and at least one initial fog boundary line.

[0107] Step 3225: Generate the current visibility of the target road based on the first pixel count of the target horizon and the second pixel count of the target fog boundary. In a specific embodiment of the present invention, when the visibility changes, the difference between the first pixel count of the target horizon and the second pixel count of the target fog boundary will change. The present invention embodiment, through analysis of a large amount of experimental observation data, found that the distance between the two lines, the target horizon, the target fog boundary, and the visibility have an inverse proportional relationship, and derived the calculation formula (7):

[0108]

[0109] Where R represents visibility, H represents the fog boundary, and W represents the number of pixels between the horizon and the ground. The coefficient 7496 (m) in this formula is used to calculate the current visibility of the target road under foggy conditions.

[0110] The current road surface friction coefficient and current visibility of the target road can be calculated using the methods described in the above embodiments.

[0111] Then proceed to step 4: Generate the optimal safe speed limit for the target vehicle based on the current road surface friction coefficient, current visibility, and real-time driving data, to support the subsequent development of a safe driving warning plan.

[0112] In one specific embodiment, such as Figure 6As shown, the real-time driving data includes at least the first real-time driving speed, vehicle length, and first coordinate position of the nearest vehicle ahead, and the second real-time driving speed and second coordinate position of the target vehicle; generating the optimal safe speed limit for the target vehicle based on the current road friction coefficient, current visibility, and real-time driving data includes the following steps:

[0113] Step 401: Invoke the first preset formula and calculate the initial safe speed limit for the target vehicle based on the current road surface friction coefficient and current visibility. In a specific embodiment of the present invention, such as... Figure 12 As shown, visibility is represented by R, which consists of three parts: reaction distance, etc. Braking distance and safe distance . It is the distance the target vehicle travels from the moment the driver observes the obstacle to the moment they begin to brake, at which point the target vehicle begins to brake. With driving speed It is related to the driver's reaction time, generally speaking, reaction time The typical timeframe is 0.1-0.4s, but this invention uses 0.25s. This invention derives a maximum safe driving speed model based on existing safe following distance calculation models. By substituting the current visibility and current road surface friction coefficient obtained in the previous steps into the first preset formula (maximum safe driving speed model), the initial safe speed limit can be calculated. :

[0114]

[0115] Step 402: Invoke the second preset formula and calculate the potential collision time between the target vehicle and the nearest vehicle ahead based on real-time driving data and the initial safe speed limit. In a specific embodiment of the present invention, the speed, acceleration, and distance of the vehicles ahead in the target lane are obtained through the on-board equipment, and the initial safe speed limit obtained in step 401 is used. replace Substituting the values ​​into the second preset formula (Conflict Risk Calculation Index TTC) for calculation, the initial safe speed limit is obtained. The time to potential conflict (TTC) between the target vehicle and the vehicle in front during travel.

[0116]

[0117] in: The target vehicle speed; The target vehicle's coordinates; The coordinates of the vehicle ahead; The length of the vehicle in front; The distance between the target vehicle and the vehicle in front; This represents the speed difference between the target vehicle and the vehicle in front.

[0118] Step 403: Determine whether the potential conflict time meets the preset high-risk condition. If not, the initial safe speed limit is taken as the optimal safe speed limit; if so, the third preset formula is called, that is, the alternative safe speed limits for the target vehicle are calculated based on real-time driving data and preset reaction time, and the alternative safe speed limits are taken as the optimal safe speed limit. In a specific embodiment of the present invention, when TTC is less than or equal to 2 seconds, it is considered a high-risk speed, and the optimal safe speed limit is calculated using the third preset formula. If TTC is greater than 2+ When the optimal safe speed limit is, then the value is... Meanwhile, when there are no other vehicles within the visibility range in the lane where the vehicle is located, the optimal safe speed limit is [value missing]. .

[0119] Third preset formula:

[0120] in: For optimal safety speed; The target vehicle's coordinates; The coordinates of the vehicle ahead; The length of the vehicle in front; The distance between the target vehicle and the vehicle in front; The speed difference between the target vehicle and the vehicle in front; This is the preset reaction time.

[0121] Finally, step 5 is executed: A safe driving warning scheme for the target vehicle on the target road is generated based on the optimal safe speed limit. In a preferred embodiment, such as... Figure 7 As shown, the specific steps for generating a safe driving warning scheme for the target vehicle on the target road based on the optimal safe speed limit are as follows:

[0122] Step 501: Obtain the second real-time driving speed corresponding to the target vehicle. When the second real-time driving speed is greater than the optimal safe speed limit, generate and display the corresponding first warning signal with sound and light. In a specific embodiment of the present invention, when the second real-time driving speed is greater than the optimal safe speed limit, the on-board device issues a danger warning to the driver of the target vehicle via sound alarm, such as "Current visibility is poor, road is slippery, please slow down and maintain a safe distance," and simultaneously flashes an indicator light to alert the driver of the target vehicle.

[0123] In other embodiments, the following warning scheme may be used simultaneously or individually: acquiring the real-time yaw rate of the target vehicle, and generating and displaying a corresponding second warning signal with sound and light when the real-time yaw rate is greater than a preset angular velocity threshold. In a specific embodiment of the present invention, when the real-time yaw rate is greater than the preset angular velocity threshold, the on-board device issues a hazard warning to the driver of the target vehicle via sound alarm, such as "Please pay attention to controlling the vehicle's driving direction," and simultaneously flashes an indicator light to alert the driver of the target vehicle.

[0124] As another aspect of this invention, this embodiment also provides a vehicle safety driving warning device under adverse conditions. The vehicle safety driving warning device under adverse conditions can be a software module, which includes several instructions stored in a memory. A processor can access the memory, call the instructions, and execute them to complete the warning methods described in the above embodiments.

[0125] In some embodiments, the vehicle safety driving warning device under adverse conditions can also be constructed from hardware devices. For example, it can be constructed from one or more chips, which can work in coordination to complete the vehicle safety driving warning method under adverse conditions described in the above embodiments. Furthermore, the vehicle safety driving warning device under adverse conditions can also be constructed from various logic devices, such as general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), microcontrollers, ARM (Acorn RISC Machine) or other programmable logic devices, discrete gate or transistor logic, discrete hardware components, or any combination of these components.

[0126] Figure 8 This is a schematic diagram of the structure of a vehicle safe driving warning device under adverse conditions provided in Embodiment 2 of the present invention, as shown below. Figure 8 As shown, it includes a data acquisition module 100, a first generation module 200, a second generation module 300, an analysis module 400, and an early warning module 500.

[0127] The acquisition module 100 is used to acquire real-time environmental images and real-time driving data of the target road through the on-board equipment of the target vehicle;

[0128] The first generation module 200 is used to call a pre-trained weather classification model and generate the current weather type of the target road based on real-time environmental images;

[0129] The second generation module 300 is used to generate the current road surface friction coefficient and current visibility of the target road according to the current weather type;

[0130] The analysis module 400 is used to generate the optimal safe speed limit for the target vehicle based on the current road surface friction coefficient, current visibility, and real-time driving data.

[0131] The early warning module 500 is used to generate a safe driving warning plan for the target vehicle on the target road based on the optimal safe speed limit.

[0132] Furthermore, the second generation module 300 specifically includes a friction coefficient generation module 301 and a visibility generation module 302. In one specific embodiment, the friction coefficient generation module 301 specifically includes:

[0133] The first query unit 3011 is used to query a preset first mapping relationship table and generate the target precipitation intensity corresponding to the current weather type. The first mapping relationship table includes preset precipitation intensities corresponding to different weather types.

[0134] The first calculation unit 3012 is used to acquire road information of the target road and calculate and generate the current road surface water film thickness based on the target precipitation intensity and the road information.

[0135] The second query unit 3013 is used to query a preset second mapping relationship table and generate the target pendulum meter reading corresponding to the current road surface water film thickness. The second mapping relationship table includes preset pendulum meter readings corresponding to different road surface water film thicknesses.

[0136] The second calculation unit 3014 is used to calculate and generate the current road surface friction coefficient of the target road based on the reading of the target pendulum instrument.

[0137] In a preferred embodiment, the friction coefficient generation module 301 includes a third query unit 3016.

[0138] The third query unit 3016 is used to query a preset third mapping relationship table to generate the current road surface friction coefficient corresponding to the current weather type. The third mapping relationship table includes preset road surface friction coefficients corresponding to different weather types.

[0139] In a preferred embodiment, the visibility generation module 302 is specifically used to: when the current weather type is rainy, acquire the target precipitation intensity corresponding to the real-time environmental image, and generate the current visibility of the target road based on the target precipitation intensity; and when the current weather type is non-rainy, use a preset fog detection method based on image color space features to classify the current weather, generate a fog classification result, and generate the current visibility of the target road based on the fog classification result.

[0140] In a preferred embodiment, the visibility generation module 302 specifically includes:

[0141] The first marking unit 3021 is used to mark at least one initial horizon in the real-time environment image using a preset horizon detection algorithm;

[0142] The processing unit 3022 is used to perform binarization processing on the real-time environment image, and to perform secondary processing on the binarization processing result based on a preset region growing method to mark at least one initial fog boundary line in the real-time environment image.

[0143] The second marking unit 3023 is used to mark the lane lines in the real-time environment image using a preset lane line detection algorithm, and to set at least one determination line at the end of the lane line, wherein the at least one determination line forms a determination line area.

[0144] The third marking unit 3024 is used to acquire a target horizon and a target fog boundary line that are located in the determination line area among at least one initial horizon and at least one initial fog boundary line;

[0145] The third calculation unit 3025 is used to generate the current visibility of the target road based on the first number of pixels of the target horizon and the second number of pixels of the target fog boundary.

[0146] In a preferred embodiment, the real-time driving data includes at least the first real-time driving speed, vehicle length, and first coordinate position of the nearest vehicle ahead, and the second real-time driving speed and second coordinate position of the target vehicle; the analysis module 400 specifically includes:

[0147] The fourth calculation unit 401 is used to call the first preset formula and calculate the initial safe speed limit of the target vehicle based on the current road surface friction coefficient and the current visibility.

[0148] The fifth calculation unit 402 is used to call the second preset formula and calculate the potential collision time between the target vehicle and the nearest vehicle in front based on the real-time driving data and the initial safe speed limit;

[0149] The execution unit 403 is used to determine whether the potential conflict time meets the preset high-risk conditions. If not, the initial safe speed limit is taken as the optimal safe speed limit. If so, the third preset formula is called, and the alternative safe speed limit of the target vehicle is calculated based on the real-time driving data and the preset reaction time. The alternative safe speed limit is taken as the optimal safe speed limit.

[0150] In a preferred embodiment, the warning module 500 specifically includes a first warning unit 501, which is used to obtain the second real-time driving speed corresponding to the target vehicle, and generate and display the corresponding first warning signal with sound and light when the second real-time driving speed is greater than the optimal safe speed limit.

[0151] In other embodiments, the warning module 500 may further include a second warning unit 502, which is used to acquire the real-time yaw rate of the target vehicle, and generate and display the corresponding second warning signal with sound and light when the real-time yaw rate is greater than a preset yaw rate threshold.

[0152] It should be noted that the above-mentioned vehicle safety driving warning device under adverse conditions can execute the vehicle safety driving warning method under adverse conditions provided in the embodiments of the present invention, and has the corresponding functional modules and beneficial effects of the method. Technical details not described in detail in the embodiments of the vehicle safety driving warning device under adverse conditions can be found in the vehicle safety driving warning method under adverse conditions provided in the embodiments of the present invention.

[0153] This invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the above-described vehicle safe driving warning method under adverse conditions.

[0154] like Figure 9 As shown, this embodiment of the invention also provides a vehicle safety driving warning device under adverse conditions, including a memory 610, a processor 620, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the above-described vehicle safety driving warning method under adverse conditions. The processor 620 and the memory 610 can be connected via a bus or other means. Figure 9 Taking a bus connection as an example, the memory 610, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules, such as the program instructions / modules corresponding to the early warning method in the embodiments of the present invention. The processor 620 executes various functional applications and data processing of the early warning device by running the non-volatile software programs, instructions, and modules stored in the memory 610, thereby realizing the functions of the early warning method provided in the above method embodiments and the various modules or units in the above device embodiments.

[0155] Memory 610 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some embodiments, memory 610 may optionally include memory remotely located relative to processor 620, which can be connected to processor 620 via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0156] The program instructions / modules are stored in the memory 610 and, when executed by one or more processors 61, execute the warning method in any of the above method embodiments.

[0157] This invention also provides a computer program product, which includes a computer program stored on a non-volatile computer-readable storage medium. The computer program includes program instructions that, when executed by an electronic device, cause the electronic device to perform any of the aforementioned warning methods.

[0158] The device or equipment embodiments described above are merely illustrative. The unit modules described as separate components may or may not be physically separate. The components shown as module units may or may not be physical units; that is, they may be located in one place or distributed across multiple network module units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0159] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented using software plus a general-purpose hardware platform, or of course, using hardware. Based on this understanding, the above technical solutions, in essence or the parts that contribute to the related technology, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0160] Finally, it should be noted that the above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for early warning of safe driving of a vehicle in a bad environment, characterized in that, The method comprises the following steps: obtaining real-time environment images and real-time driving data of a target road through a vehicle-mounted device of a target vehicle; calling a pre-trained weather classification model and generating a current weather type of the target road based on the real-time environment images; the weather classification model comprises a rain day judgment model and a fog day judgment model, the rain day judgment model is used to judge whether the weather type is a rain day and comprises the following sub-steps: using a computer vision algorithm Darknet convolutional neural network training, collecting data for different weather conditions and different road conditions, then using LabelImage to divide the collected pictures into five categories of sunny, light rain, moderate rain, heavy rain and heavy rain, wherein 80% of the data is established as a training set and 20% of the data is established as a validation set, the data is trained by using the Darknet network training, and the trained weight file and network are implanted into a Yolov4 visual model, so as to train the rain day judgment model; the fog day judgment model is based on the HSV feature for fog day classification, when the collected environment images are videos, the videos are decomposed into pictures, then different H, S and V components are extracted according to the color information of each picture, each pixel point is composed of H, S and V components, the total sum of the feature components is obtained by traversing each pixel of the image, and then the total sum is substituted into the calculation formula The following calculations are made , and , according to the formula the judgment condition in the table 1, judges whether to meet the heavy fog ( ) determination condition, if yes, it is heavy fog, otherwise judges whether to meet the no fog ( ) determination condition, if yes, it is non-fog day, otherwise judges whether to meet the light fog ( ) determination condition, if yes, it is light fog, otherwise it is heavy fog; generating a current road surface friction coefficient and a current visibility of the target road according to the current weather type; generating an optimal safety speed limit of the target vehicle according to the current road surface friction coefficient, the current visibility and the real-time driving data; generating a safe driving warning scheme of the target vehicle on the target road according to the optimal safety speed limit.

2. The method of claim 1, wherein, The generation of the current road surface friction coefficient of the target road according to the current weather type comprises the following sub-steps: querying a preset first mapping relationship table to generate a target precipitation intensity corresponding to the current weather type, the first mapping relationship table comprises preset precipitation intensities corresponding to different weather types; obtaining road information of the target road, and calculating a current road surface water film thickness according to the target precipitation intensity and the road information; querying a preset second mapping relationship table to generate a target pendulum instrument reading corresponding to the current road surface water film thickness, the second mapping relationship table comprises preset pendulum instrument readings corresponding to different road surface water film thicknesses; calculating the current road surface friction coefficient of the target road according to the target pendulum instrument reading.

3. The method of claim 1, wherein, The generation of the current road surface friction coefficient of the target road according to the current weather type comprises the following sub-steps: querying a preset third mapping relationship table to generate a current road surface friction coefficient corresponding to the current weather type, the third mapping relationship table comprises preset road surface friction coefficients corresponding to different weather types.

4. The method according to any of claims 1 to 3, characterized in that, The generation of the current visibility of the target road according to the current weather type comprises the following sub-steps: when the current weather type is a rain day, obtaining a target precipitation intensity corresponding to the real-time environment images, and generating a current visibility of the target road according to the target precipitation intensity; When the current weather type is a non-rainy day, a preset foggy day detection method based on image color space features is used to classify the current weather, a foggy day classification result is generated, and the current visibility of the target road is generated based on the foggy day classification result.

5. The method of claim 4, wherein, The foggy day classification result includes non-foggy day, light foggy day, and heavy foggy day. If the current weather is a heavy foggy day, generating the current visibility of the target road includes the following sub-steps: Using a preset horizon detection algorithm to mark at least one initial horizon in the real-time environment image; Performing binaryzation processing on the real-time environment image, and performing secondary processing on the binaryzation processing result based on a preset region growing method to mark at least one initial foggy ground boundary line in the real-time environment image; Using a preset lane line detection algorithm to mark lane lines in the real-time environment image, and setting at least one decision line at the end of the lane lines, at least one of the decision lines forming a decision line area; Obtaining a target horizon and a target foggy ground boundary line in the decision line area from at least one of the initial horizons and at least one of the initial foggy ground boundary lines; Generating the current visibility of the target road according to a first pixel number of the target horizon and a second pixel number of the target foggy ground boundary line.

6. The method of claim 4, wherein, The real-time driving data at least includes a first real-time driving speed, a vehicle body length, and a first coordinate position of a nearest vehicle in front, and a second real-time driving speed and a second coordinate position of a target vehicle; The step of generating the optimal safe speed limit of the target vehicle according to the current road surface friction coefficient, the current visibility, and the real-time driving data includes the following steps: Calling a first preset formula and calculating an initial safe speed limit of the target vehicle based on the current road surface friction coefficient and the current visibility; Calling a second preset formula and calculating a potential conflict time of the target vehicle colliding with the nearest vehicle in front according to the real-time driving data and the initial safe speed limit; Determining whether the potential conflict time meets a preset high-risk condition. If not, the initial safe speed limit is used as the optimal safe speed limit. If yes, a third preset formula is called, and a candidate safe speed limit of the target vehicle is calculated based on the real-time driving data and a preset reaction time, and the candidate safe speed limit is used as the optimal safe speed limit.

7. The method of claim 6, wherein, The step of generating a safe driving warning scheme of the target vehicle on the target road according to the optimal safe speed limit specifically includes: Obtaining a second real-time driving speed corresponding to the target vehicle. When the second real-time driving speed is greater than the optimal safe speed limit, a first warning signal corresponding to the target vehicle is generated and displayed by sound and light. And / or obtaining a real-time yaw rate of the target vehicle. When the real-time yaw rate is greater than a preset angular velocity threshold, a second warning signal corresponding to the target vehicle is generated and displayed by sound and light.

8. A device for early warning of safe driving of a vehicle in a bad environment, characterized in that, The system includes a collection module, a first generation module, a second generation module, an analysis module, and a warning module, The collection module is used to obtain real-time environment images and real-time driving data of a target road through a vehicle-mounted device of a target vehicle; The first generation module is used to call a pre-trained weather classification model and generate a current weather type of the target road based on the real-time environment images; The second generation module is configured to generate a current road surface friction coefficient and a current visibility of a target road according to the current weather type; The analysis module is configured to generate an optimal safe speed limit of the target vehicle according to the current road surface friction coefficient, the current visibility and the real-time driving data; The early warning module is configured to generate a safe driving early warning scheme of the target vehicle on the target road according to the optimal safe speed limit. 9.A computer readable storage medium, storing a computer program, wherein the computer program is executed by a processor to implement the steps of the vehicle safe driving early warning method in a poor environment according to any one of claims 1-7. 10.A vehicle safe driving early warning device in a poor environment, comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the vehicle safe driving early warning method in a poor environment when executing the computer program.

Images