Expressway fog early warning system and early warning method

Abstract

The application discloses a highway fog group early warning system and early warning method, the highway fog group early warning system comprises an input end, a sensing controller and an output end; the input end is used for collecting the fog group outside the vehicle and road condition information, and sending the fog group and road condition information to the sensing controller; the sensing controller is used for carrying out data processing, model calculation and result fusion on the information acquired by the input end, obtaining a final sensing judgment result and sending an alarm signal to the output end; the output end is used for carrying out fog group early warning on the driver and controlling the vehicle to carry out emergency avoidance. The application provides a highway fog group early warning system and early warning method, which can make the vehicle itself have fog group road condition sensing capability, and has the characteristics of high recognition rate, low cost and easy popularization.

Description

Technical Field

[0001] This invention relates to a highway fog warning system and warning method. Background Technology

[0002] Dense fog is a type of fog with higher concentration and lower visibility within a localized area of ​​tens to hundreds of meters, influenced by local microclimates. The main causes of dense fog are significant temperature and humidity differences. Therefore, dense fog often occurs between nighttime and 8:00 AM the following day in autumn and winter. It is also more likely to occur near water resources, especially on highways. Dense fog can cause sudden changes in visibility, posing a significant threat to highway traffic safety and easily leading to serious traffic accidents.

[0003] Currently, common solutions often employ IoT technology to transmit alarm information to vehicles behind via wireless communication. However, accidents involving fog on highways typically result in severe vehicle damage, posing a risk that IoT devices might malfunction and fail to transmit alarm signals in a timely manner. Other solutions involve using the scattered laser signals emitted by specialized vehicle-mounted LiDAR to determine the presence of fog ahead; however, LiDAR itself is expensive and cannot yet be widely adopted.

[0004] Therefore, a system solution that enables vehicles to perceive fog conditions while being cost-effective and easy to implement is an urgent problem to be solved. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to overcome the shortcomings of the prior art and provide a highway fog warning system and warning method that enables vehicles to have the ability to perceive fog conditions, and also has the characteristics of high recognition rate, low cost and easy popularization.

[0006] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows:

[0007] This invention provides a highway fog warning system, which includes an input terminal, a sensing controller, and an output terminal;

[0008] The input terminal is used to collect information on fog patches and road conditions outside the vehicle, and send the fog patches and road conditions information to the perception controller;

[0009] The perception controller is used to process the information acquired from the input terminal, perform model calculations and result fusion, obtain the final perception judgment result, and send the alarm signal to the output terminal.

[0010] The output terminal is used to provide fog warnings to the driver and control the vehicle to take emergency evasive action.

[0011] Furthermore, the input terminal includes a temperature and humidity sensor, a front-facing camera, and a forward-facing millimeter-wave radar;

[0012] The temperature and humidity sensor is connected to the sensing controller and is used to acquire external temperature and humidity information with time information.

[0013] The front-facing camera is connected to the perception controller via a MIPI signal line to obtain information about the road conditions ahead of the vehicle.

[0014] The forward-facing millimeter-wave radar is connected to the perception controller via a CAN bus to obtain information on the position and distance of vehicles ahead.

[0015] Furthermore, the temperature and humidity sensor is installed inside the front grille of the vehicle.

[0016] Furthermore, the front-facing camera is installed behind the vehicle's windshield.

[0017] Furthermore, the forward-facing millimeter-wave radar is installed at the front of the vehicle, and the forward-facing millimeter-wave radar operates at a frequency of 77 GHz.

[0018] Furthermore, the road conditions ahead of the vehicle include the vehicles ahead and lane markings.

[0019] Furthermore, the output terminal includes an alarm module and a vehicle control module;

[0020] The alarm module uses sound, light, and electricity to alert the driver to fog formations.

[0021] The vehicle control module is used to control the vehicle to decelerate and turn on the hazard lights and fog lights.

[0022] Another aspect of the present invention provides a warning method for the aforementioned highway fog warning system, comprising:

[0023] Step S1: The perception controller preprocesses the acquired input image stream data, and then sends each single frame image into a deep convolutional neural network model for lane line recognition and object recognition, and locally extracts the lane lines, vehicles ahead, and traffic signs in the single frame image to obtain images of lane lines, vehicles ahead, and traffic signs.

[0024] Step S2: The perception controller analyzes the acquired input radar signal to obtain the position and distance information of the vehicle ahead.

[0025] Step S3: The sensing controller compares the real-time temperature value acquired from the input terminal with the temperature value collected according to the preset time point. When the temperature difference is greater than the preset difference, it indicates that the temperature difference is too large and the sensing controller outputs "1". Otherwise, it outputs "0".

[0026] The sensor controller compares the real-time humidity value acquired from the input terminal with the pre-stored humidity value. When the humidity difference is greater than the preset difference, it indicates that the humidity is too high and outputs "1", otherwise it outputs "0".

[0027] Step S4: The environmental processing module of the sensing controller multiplies the comparison results of temperature and humidity by weights and adds them together, as shown below:

[0028] R 环 =R T *w T +R H *w H

[0029] Among them, R T For the results of temperature comparison, w T As a weight for temperature, R H For the results of humidity comparison, w H The weighting is based on humidity.

[0030] Step S5: The comprehensive judgment module of the perception controller obtains the final perception judgment result by weighting and summing each input element.

[0031] Furthermore, step S1 also includes:

[0032] The images of the lane lines and vehicles are processed by a post-processing model to obtain the state information of the lane lines and the state information of the vehicles ahead.

[0033] The image of the traffic sign is processed by a semantic recognition model to further identify whether the traffic sign contains the semantic meaning of "caution for fog".

[0034] The status information of the lane lines includes the visible lengths of multiple lane lines;

[0035] The status information of the vehicle ahead mainly includes the distance information of the vehicle ahead and the position information of the vehicle ahead.

[0036] Furthermore, in step S5, the comprehensive judgment module of the perception controller obtains the final perception judgment result by weighting and summing each input element, specifically including:

[0037] Step S51: The comprehensive judgment module obtains lane line information. When the visible distance of the lane line is <30m, then R L =1, otherwise R L =0.

[0038] Step S52: The integrated judgment module of the perception controller obtains the information of the vehicles ahead from the image and the radar.

[0039] If the front-facing camera does not detect a vehicle ahead, but the millimeter-wave radar does, then R S =1;

[0040] If neither the front-facing camera nor the millimeter-wave radar detects a vehicle ahead, then R S =0;

[0041] When both the front-facing camera and millimeter-wave radar detect a vehicle ahead, then R S =0.2;

[0042] If the front-facing camera detects a vehicle ahead but the millimeter-wave radar does not, then R S =0.5;

[0043] Step S53, w L =0.2, w S =0.3, w 环 =0.5, then:

[0044] R 综 =R L *w L +R S *w S +R 环 *w 环

[0045] When R 综 When the value is ≥0.9, the alarm module at the output end will alarm the driver through various means such as sound, light, and electricity, while the vehicle control module at the output end will control the vehicle to decelerate.

[0046] Among them, R L The result of the lane marking judgment;

[0047] R S The result of the sensor's judgment;

[0048] W L The weight of the lane lines;

[0049] L S The weights of the sensors;

[0050] W 环 The weight of environmental factors;

[0051] R 综 This is the result of a comprehensive judgment.

[0052] By employing the above technical solution, this invention uses various algorithm models in the perception controller to perform actions such as forward vehicle detection, lane line detection, and traffic sign semantic recognition on the image stream signal sent by the front-facing camera. It then fuses the image perception results with signals from millimeter-wave radar and temperature and humidity differences to obtain a more accurate assessment of the road conditions ahead, especially in foggy conditions. Finally, the perception controller sends the final perception judgment results to the alarm module and vehicle control module in real time for alerting and emergency vehicle avoidance. This invention can be widely used in intelligent assisted driving vehicles, real-time road monitoring, and other applications. Attached Figure Description

[0053] Figure 1 This is a system architecture diagram of the highway fog warning system of the present invention;

[0054] Figure 2 This is the execution logic diagram of the perception controller of the present invention. Detailed Implementation

[0055] To make the content of this invention easier to understand, the invention will be further described in detail below with reference to specific embodiments and accompanying drawings.

[0056] Example 1

[0057] like Figure 1 As shown, this embodiment provides a highway fog warning system, which includes an input terminal, a sensing controller, and an output terminal;

[0058] The input terminal is used to collect information on fog patches and road conditions outside the vehicle, and send the fog patches and road conditions information to the perception controller;

[0059] The perception controller is used to process the information acquired from the input terminal, perform model calculations and result fusion, obtain the final perception judgment result, and send the alarm signal to the output terminal.

[0060] The output terminal is used to provide fog warnings to the driver and control the vehicle to take emergency evasive action.

[0061] like Figure 1 As shown, the input terminal in this embodiment includes a temperature and humidity sensor, a front-facing camera, and a forward-facing millimeter-wave radar.

[0062] The temperature and humidity sensor, connected to the sensing controller and installed inside the front grille of the vehicle, acquires external temperature and humidity information with time information. The sensor includes a time synchronization module and provides the sensing controller with real-time external temperature and humidity signals. The sensing controller automatically records daytime and nighttime temperature information and compares the temperature difference between different time periods to see if it is excessive. Simultaneously, the sensing controller compares its internally preset humidity value with the real-time humidity value to determine if the external ambient humidity is excessive. The magnitude of the temperature and humidity difference serves as one of the conditions for the sensing controller to determine the presence of fog.

[0063] The front-facing camera is connected to the perception controller via a MIPI signal line. The front-facing camera is mounted behind the vehicle's windshield to acquire information about the road ahead. It collects road information in real time and sends it to the perception controller as an image stream. Preferably, the camera has 8 megapixels, a frame rate of 30fps, and a field of view of 120° horizontally and 55° vertically.

[0064] The forward-facing millimeter-wave radar is mounted at the front of the vehicle and is connected to the perception controller via a CAN bus to acquire information on the position and distance of vehicles ahead. The forward-facing millimeter-wave radar uses a 77GHz long-range millimeter-wave radar with a maximum detection range of approximately 250m. A preferred architecture is an 8-transmitter, 8-receiver configuration, capable of real-time monitoring of up to 64 targets ahead of the vehicle, acquiring information on their distance, relative speed, azimuth, and altitude.

[0065] like Figure 1 As shown, the output terminal in this embodiment includes an alarm module and a vehicle control module.

[0066] The alarm module provides fog warnings and signal reminders to the driver through sound, light, and electricity, and can provide multiple alarm and reminder modes based on the actual module functions.

[0067] The vehicle control module is either the vehicle's BCM or central controller module. The vehicle control module is used to control the vehicle's deceleration and turn on the hazard lights and fog lights to play an emergency avoidance role.

[0068] like Figure 1As shown, the perception controller in this embodiment, as the core component of the system, employs a deep convolutional neural network model, a post-processing model, and a semantic recognition model to perform tasks such as forward vehicle detection, lane line detection, and traffic sign semantic recognition on the image stream signal sent by the front-facing camera. It then fuses the image perception results with signals from millimeter-wave radar and temperature / humidity differences to obtain a more accurate picture of the road ahead. The post-processing model further calculates the objects identified in the image, including the position, distance, and size of the vehicle in front; the output of the deep convolutional neural network model does not include this information. Similarly, the semantic recognition model identifies traffic signs, but the content and meaning of the text on the signs require the semantic recognition model to interpret.

[0069] Finally, the comprehensive judgment module adds up the weights of each factor and sends the final perception judgment result to the alarm module and vehicle control module in real time.

[0070] Example 2

[0071] like Figure 2 As shown in the figure, this embodiment provides a warning method for a highway fog warning system, which includes:

[0072] Step S1: The perception controller preprocesses the acquired input image stream data, including data cleaning and data augmentation. Then, each single-frame image is fed into a deep convolutional neural network model for lane line recognition and object recognition, and local extraction of lane lines, vehicles ahead, and traffic signs is performed on each single-frame image to obtain images of lane lines, vehicles ahead, and traffic signs, wherein:

[0073] The images of the lane lines and vehicles are processed by a post-processing model to obtain the state information of the lane lines and the state information of the vehicles ahead.

[0074] The image of the traffic sign is processed by a semantic recognition model to further identify whether the traffic sign contains the semantic meaning of "caution for fog".

[0075] The status information of the lane lines includes the visible lengths of multiple lane lines;

[0076] The status information of the vehicle ahead mainly includes the distance information of the vehicle ahead and the position information of the vehicle ahead.

[0077] Step S2: The perception controller analyzes the acquired input radar signal to obtain the position and distance information of the vehicle ahead.

[0078] Step S3: The sensing controller compares the real-time temperature value acquired from the input terminal with the temperature value collected according to the preset time point. When the temperature difference is greater than the preset difference, it indicates that the temperature difference is too large and the sensing controller outputs "1". Otherwise, it outputs "0".

[0079] The sensor controller compares the real-time humidity value acquired from the input terminal with the pre-stored humidity value. When the humidity difference is greater than the preset difference, it indicates that the humidity is too high and outputs "1", otherwise it outputs "0".

[0080] Step S4: The environmental processing module of the sensing controller multiplies the comparison results of temperature and humidity by weights and adds them together, as shown below:

[0081] R 环 =R T *w T +R H *w H

[0082] Among them, R T For the results of temperature comparison, w T As a weight for temperature, R H For the results of humidity comparison, w H For humidity weighting, preferably, w T =0.35, w H =0.65.

[0083] Step S5: The comprehensive judgment module of the perception controller calculates the final perception judgment result by weighting and summing the input elements, specifically including:

[0084] Step S51: The comprehensive judgment module obtains lane line information. When the visible distance of the lane line is <30m, then R L =1, otherwise R L =0.

[0085] Step S52: The integrated judgment module of the perception controller obtains the information of the vehicles ahead from the image and the radar.

[0086] If the front-facing camera does not detect a vehicle ahead, but the millimeter-wave radar does, then R S =1;

[0087] If neither the front-facing camera nor the millimeter-wave radar detects a vehicle ahead, then R S =0;

[0088] When both the front-facing camera and millimeter-wave radar detect a vehicle ahead, then R S =0.2;

[0089] If the front-facing camera detects a vehicle ahead but the millimeter-wave radar does not, then R S =0.5;

[0090] Step S53, w L =0.2, w S =0.3, w 环 =0.5, then:

[0091] R 综 =R L *w L +R S *w S +R 环 *w 环

[0092] When R 综 When the value is ≥0.9, the alarm module at the output end will alarm the driver through various means such as sound, light, and electricity, while the vehicle control module at the output end will control the vehicle to decelerate.

[0093] Among them, R L The result of the lane marking judgment;

[0094] R S The result of the sensor's judgment;

[0095] W L The weight of the lane lines;

[0096] L S The weights of the sensors;

[0097] W 环 The weight of environmental factors;

[0098] R 综 This is the result of a comprehensive judgment.

[0099] The specific embodiments described above further illustrate the technical problems, technical solutions, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the present invention. 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 highway fog warning system, characterized in that: It includes an input terminal, a sensing controller, and an output terminal; The input terminal is used to collect information on fog patches and road conditions outside the vehicle, and send the fog patches and road conditions information to the perception controller; The perception controller is used to process the information acquired from the input terminal, perform model calculations and result fusion, obtain the final perception judgment result, and send the alarm signal to the output terminal. The output terminal is used to provide the driver with fog warnings and control the vehicle to take emergency evasive action. The early warning method of the highway fog warning system includes: Step S1: The perception controller preprocesses the acquired input image stream data, and then sends each single frame image into a deep convolutional neural network model for lane line recognition and object recognition, and locally extracts the lane lines, vehicles ahead, and traffic signs in the single frame image to obtain images of lane lines, vehicles ahead, and traffic signs. Step S2: The perception controller analyzes the acquired input radar signal to obtain the position and distance information of the vehicle ahead. Step S3: The sensing controller compares the real-time temperature value acquired from the input terminal with the temperature value collected according to the preset time point. When the temperature difference is greater than the preset difference, it indicates that "the temperature difference is too large" and the sensing controller outputs "1"; otherwise, it outputs "0". The sensor controller compares the real-time humidity value acquired from the input terminal with the pre-stored humidity value. When the humidity difference is greater than the preset difference, it indicates that the humidity is too high and outputs "1", otherwise it outputs "0". Step S4: The environmental processing module of the sensing controller multiplies the comparison results of temperature and humidity by weights and adds them together, as shown below: ； in, The results are for temperature comparison. Weighted by temperature, The results are for humidity comparison. Weighted by humidity; Step S5: The comprehensive judgment module of the perception controller obtains the final perception judgment result by weighting and summing each input element; Step S1 further includes: The images of the lane lines and vehicles are processed by a post-processing model to obtain the state information of the lane lines and the state information of the vehicles ahead. The images of the traffic signs are processed by a semantic recognition model to further identify whether the traffic signs contain the semantic meaning of "caution: fog". The status information of the lane lines includes the visible lengths of multiple lane lines; The status information of the vehicle ahead includes the distance information of the vehicle ahead and the position information of the vehicle ahead. In step S5, the comprehensive judgment module of the perception controller obtains the final perception judgment result by weighting and summing each input element. Specifically, this includes: Step S51: The comprehensive judgment module obtains lane line information. When the visible distance of the lane line is <30m, then... Conversely ; Step S52: The integrated judgment module of the perception controller obtains the information of the vehicles ahead from the image and the radar. If the front-facing camera does not detect a vehicle ahead, but the millimeter-wave radar does, then ; If neither the front-facing camera nor the millimeter-wave radar detects a vehicle ahead, then ; When both the front-facing camera and millimeter-wave radar detect a vehicle ahead, then ; If the front-facing camera detects a vehicle ahead but the millimeter-wave radar does not, then ; Step S53 , , Then we have: ； when At this time, the alarm module at the output end will alarm the driver through various forms of sound, light, and electricity, while the vehicle control module at the output end will control the vehicle to decelerate. Among them, R L The result of the lane marking judgment; R S The result of the sensor's judgment; W L The weight of the lane lines; L S The weights of the sensors; W 环 The weight of environmental factors; R 综 This is the result of a comprehensive judgment.

2. The highway fog warning system according to claim 1, characterized in that: The input terminal includes a temperature and humidity sensor, a front-facing camera, and a forward-facing millimeter-wave radar. The temperature and humidity sensor is connected to the sensing controller and is used to acquire external temperature and humidity information with time information. The front-facing camera is connected to the perception controller via a MIPI signal line to obtain information about the road conditions ahead of the vehicle. The forward-facing millimeter-wave radar is connected to the perception controller via a CAN bus to obtain information on the position and distance of vehicles ahead.

3. The highway fog warning system according to claim 2, characterized in that: The temperature and humidity sensor is installed inside the front grille of the vehicle.

4. The highway fog warning system according to claim 2, characterized in that: The front-facing camera is installed behind the vehicle's windshield.

5. The highway fog warning system according to claim 2, characterized in that: The forward-facing millimeter-wave radar is installed at the front of the vehicle, and the operating frequency of the forward-facing millimeter-wave radar is 77 GHz.

6. The highway fog warning system according to claim 2, characterized in that: The road conditions ahead of the vehicle include the vehicles ahead and lane markings.

7. The highway fog warning system according to claim 2, characterized in that: The output terminal includes an alarm module and a vehicle control module; The alarm module uses sound, light, and electricity to alert the driver to fog formations. The vehicle control module is used to control the vehicle to decelerate and turn on the hazard lights and fog lights.

Images