Vehicle wading early warning system

Abstract

The invention relates to a vehicle wading early warning system, which captures a real-time video stream of a road surface around a vehicle through a camera module, and extracts road conditions, waterlogging depth, weather changes and topographic features from the video stream in combination with an intelligent video analysis module based on a convolutional neural network (CNN) and a Transform architecture. The system queries historical data and real-time weather information through networking, analyzes the distance between a vehicle chassis and the ground and tire pressure changes, performs real-time wading risk assessment in combination with the accumulated water depth, can predict the change trend of the accumulated water depth through an accumulated water dynamic change prediction module, provides an early warning function, and improves the wading risk of a vehicle. The cross-motorcade wading early warning module enables multiple vehicles to share real-time wading information based on V2V communication and improves the collaborative early warning capability of the motorcade, and the satellite communication module sends compressed pictures, GPS position information and distress signals to a rescue center at regular time when the vehicles are in danger and synchronizes data to a cloud end during normal operation to provide support for subsequent rescue.

Description

Technical Field

[0001] The invention relates to a vehicle wading warning system, in particular to a vehicle wading warning system applied in the technical field of video warning. Background Art

[0002] With the development of automobile technology and the complexity of road environment, vehicle wading has become one of the important issues affecting driving safety. Traditional wading warning systems mainly rely on sensor detection. However, these technologies often cannot provide accurate warnings, especially in complex terrain, dynamic changes in water accumulation or sparsely populated wilderness. Misjudgments or delayed warnings are prone to occur. Currently, some wading warning systems based on visual perception and sensors have been applied in practice, but there are still many technical limitations.

[0003] Chinese invention patent CN115973034B discloses a water wading warning method and system integrating visual perception. The system monitors in real time whether the vehicle enters a water-wading section, obtains the water wading depth and road slope conditions on both sides of the vehicle body, and determines whether the vehicle faces a water wading risk. Although this design can reduce false alarms and improve the recognition accuracy of slope roads, it still relies on water wading depth monitoring under static conditions and cannot handle scenes with dynamic changes in water accumulation. At the same time, the system can only play a good warning effect when the vehicle enters a downhill section, and may have greater limitations on other complex terrains.

[0004] Chinese invention patent CN114906085B discloses a vehicle wading processing method, which monitors the water pressure signal at a certain position of the vehicle through a water pressure sensor to control the vehicle to enter the wading mode and perform corresponding anti-wading processing. Although the design can monitor the vehicle wading situation in real time through the water pressure sensor, the technology is limited to the acquisition of local water pressure signals, and cannot perform a global assessment of the depth of water accumulation, nor can it predict the future water accumulation dynamics. In addition, the water pressure sensor is easily interfered in a complex environment, resulting in delayed or invalid warning signals, and can only handle local wading situations, and cannot coordinate multiple workshops to share warning information.

[0005] The above designs have improved the accuracy of water wading warnings to a certain extent through sensors or visual perception, but there are still the following limitations. The existing technology lacks the ability to predict the dynamic changes of water accumulation, and cannot assess the risk of water accumulation and issue warnings in advance based on real-time data; it relies on the information source of a single vehicle and cannot share water wading information through collaboration between fleets, resulting in limited warning range and accuracy; the monitoring of vehicle status such as vehicle tire pressure and chassis height is relatively limited, and the vehicle's actual wading ability cannot be fully evaluated; under complex terrain or meteorological conditions, it is impossible to effectively respond to sudden expansion or deterioration of water accumulation. Summary of the invention

[0006] In view of the above-mentioned existing technologies, the technical problem to be solved by the present invention is how to evaluate the dynamic changes of water accumulation and the vehicle's wading ability in real time based on video stream analysis, dynamic prediction and cross-fleet collaboration, and improve the early warning effect through fleet collaborative communication.

[0007] To solve the above problems, the present invention provides a vehicle wading warning system, comprising a camera module mounted on a vehicle, for capturing a video stream of the road surface in front of the vehicle and the surrounding road surface;

[0008] A computing unit, used to process the video stream captured by the camera module, the computing unit comprising an intelligent video analysis module based on the Transformer architecture, the module obtaining road conditions, water depth, weather changes, terrain features and vehicle status from the video stream, and combining historical data queried online to assess the risk of wading ahead;

[0009] The video stream analysis module is used to analyze the distance between the vehicle chassis and the ground through the video stream, monitor the tire pressure changes, and evaluate the vehicle's wading ability in combination with the road conditions;

[0010] The terrain analysis module analyzes the current road terrain based on camera data and combines historical data obtained through network queries to evaluate the probability of wading in the current environment;

[0011] The dynamic change prediction module of water accumulation is based on the video stream collected by the camera, real-time weather information and terrain analysis, combined with historical water accumulation data, to predict the change of water accumulation on the road ahead and determine whether the water accumulation will deepen or subside;

[0012] The cross-fleet flood warning module uses the V2V communication system to automatically transmit flood information to adjacent vehicles when one vehicle detects flooding, enhancing the fleet's overall perception of the flooded area ahead.

[0013] Satellite communication module, used to send compressed images and distress signals to the rescue center at regular intervals when the vehicle is in distress, and synchronize the data to the cloud during normal operation to provide clues for subsequent rescue;

[0014] An alarm module that issues visual and audible alerts to the driver when the system detects a risk of wading;

[0015] The communication module is used to obtain historical water accumulation and road information through network query, and transmit road surface video stream and water wading warning information to external devices, such as other vehicles or traffic management centers.

[0016] In the above-mentioned vehicle wading warning system, a high-definition camera module is installed on the vehicle, combined with an intelligent video analysis module based on a convolutional neural network (CNN) and a Transformer architecture, to analyze the real-time captured road video stream, extract information such as road water depth, weather conditions, and vehicle status, and combine historical data and real-time meteorological information through a computing unit. The system can dynamically predict the changing trend of water depth. The system also analyzes the distance between the vehicle chassis and the ground and tire pressure changes through video streams, evaluates the vehicle's wading ability, and issues real-time warnings. In addition, the system realizes multi-vehicle collaborative warnings through a V2V communication module, and vehicles can share real-time information. Road water accumulation information ensures that all vehicles in the fleet are informed of water risks in a timely manner. At the same time, the system is equipped with satellite communication function, which can send distress signals and compressed pictures to the rescue center at a regular time when the vehicle is in water difficulties, and synchronize driving data to the cloud to provide support for subsequent rescue. The application scenarios of the present invention are wide, especially suitable for areas with frequent rainy seasons and complex terrain. It can significantly improve the safety of vehicle wading and reduce traffic accidents and vehicle damage caused by wading. Economically, through intelligent early warning and multi-vehicle collaboration functions, the present invention can reduce misjudgment and accident rates, reduce vehicle maintenance costs and rescue costs, and at the same time improve the overall road driving safety and traffic efficiency.

[0017] As a further improvement of the present application, the intelligent video analysis module includes a convolutional neural network (CNN) for extracting spatial features from the video stream, and the spatial features are further passed into the Transformer model for temporal feature analysis and prediction of water accumulation changes.

[0018] As a further improvement of the present application, the Transformer architecture embeds time information into image features in the video stream through position encoding, which is used to perform time series analysis on image features at different time points to accurately determine the dynamic changes of water accumulation and water wading risks.

[0019] As a further improvement of the present application, the system obtains real-time and historical meteorological data, traffic flow data and geographic information through network query via the communication module, and combines the data with real-time image data in the video stream to improve the accuracy of water-related warnings.

[0020] As another improvement of the present application, the video stream analysis module is used to monitor tire pressure changes in real time by analyzing image features of vehicle tires contacting the ground, determine whether the tire pressure is abnormal due to wading, and perform wading risk assessment based on the depth of water accumulation.

[0021] As another improved supplement of the present application, the video stream analysis module is used to calculate the distance between the vehicle chassis and the ground from the video stream, and to evaluate the wading capability in combination with the depth of water accumulation, and to provide corresponding wading warnings.

[0022] As another improved supplement to the present application, the cross-fleet water wading warning module enables multiple vehicles to share road environment information in the video stream and historical data of online query through the V2V communication protocol, so as to collaboratively improve the accuracy and response speed of water wading warnings.

[0023] As another improvement of the present application, the satellite communication module periodically sends compressed images, GPS location information and vehicle status data to the rescue center when the vehicle is in distress, and synchronizes driving data to the cloud during normal operation to provide data support for subsequent rescue.

[0024] A vehicle wading warning system, the working method of which comprises the following steps:

[0025] S1: Captures real-time video streams of the road ahead and around the vehicle through a camera module installed on the vehicle;

[0026] S2: In the computing unit, the video stream is processed using an intelligent video analysis module based on the Transformer architecture to obtain information about road conditions, water depth, weather changes, terrain features, and vehicle status;

[0027] S3: Combined with historical data and real-time meteorological information from online queries, the distance between the vehicle chassis and the ground is calculated through the video stream analysis module, and the tire pressure changes are monitored to determine the risk of vehicle wading;

[0028] S4: Through the waterlogging dynamic change prediction module, based on the real-time data collected by the camera, historical waterlogging information, and terrain analysis results, the dynamic change of waterlogging is predicted to determine whether the waterlogging will deepen or subside;

[0029] S5: When the vehicle detects that it is in distress due to water, it will send compressed images and distress signals to the rescue center through the satellite communication module at regular intervals, and synchronize real-time data to the cloud to provide clues for subsequent rescue;

[0030] S6: During normal use, the V2V communication module is used to share water-crossing information with other vehicles, improving the water-crossing warning capability of multi-vehicle collaboration, and synchronizing data to the cloud or traffic management center to enhance overall road safety.

[0031] In summary, this application has the following beneficial effects:

[0032] 1. By introducing an intelligent video analysis module based on convolutional neural network (CNN) and Transformer architecture, key data such as road conditions, water depth, weather changes and terrain features can be automatically extracted from the real-time video streams in front of and around the vehicle. This intelligent analysis method enables the system to not only accurately judge the current water risk, but also predict the dynamic changes of water through time series analysis of the video stream, helping the driver to make driving decisions in advance. Compared with traditional water warning systems that rely on a single sensor or static data, the present invention has significantly improved the depth and real-time performance of data processing, ensuring the accuracy of warnings in complex environments.

[0033] 2. It combines vehicle video streams, historical data from online queries and real-time meteorological information, analyzes the distance between the vehicle chassis and the ground through video streams, and monitors changes in tire pressure. This combination of multi-dimensional data enables the system to not only provide basic wading risk assessments, but also conduct in-depth analysis of the vehicle's wading capability and tire status to avoid vehicle failures and safety hazards caused by wading. Unlike traditional early warning systems, the present invention enables vehicles to respond more easily in complex water-logged environments through real-time analysis of chassis height and tire pressure, reducing the possibility of accidents during wading.

[0034] 3. The V2V communication module further improves the safety and efficiency of flood warning through fleet collaboration. Multiple vehicles can share real-time road environment data to ensure that each vehicle can promptly learn about the situation in the flooded area ahead. Compared with a single-vehicle flood warning system, fleet collaboration can cover a wider area and avoid chain problems among multiple vehicles in the same flooded area. This fleet collaboration model not only improves driving safety and reduces the occurrence of accidents between vehicles, but also provides a collaborative reference for future intelligent transportation.

[0035] 4. The dynamic change prediction module of water accumulation can predict the changing trend of water accumulation depth based on the real-time data collected by the camera, historical water accumulation information and terrain analysis results. This dynamic prediction function enables the system to not only detect the current water accumulation situation, but also to foresee in advance whether the water accumulation will deepen or expand in the future. The early warning function ensures that the driver can take preventive measures before water accumulation occurs to reduce the risk of wading, which is especially suitable for scenes of severe weather or sudden road water accumulation.

[0036] 5. When the vehicle encounters a water-related predicament, the satellite communication module of this patent can maintain a stable distress signal sending function in areas where the communication signal is weak or there is no signal. When the vehicle is in trouble, the system will regularly send compressed images, GPS location information and vehicle status data to the rescue center to ensure that rescue personnel can grasp the on-site situation in a timely manner. In addition, the system will synchronize driving data and environmental information to the cloud during normal operation to provide data support for possible subsequent rescue operations. This comprehensive rescue support system greatly improves the rescue efficiency in emergency situations and ensures the safety of vehicles and drivers.

[0037] 6. Through intelligent water-crossing warning, dynamic prediction and collaborative communication functions, it not only provides drivers with more intuitive water-crossing information, but also reminds drivers of potential dangers through visual and sound alarms, enhancing the user's driving experience. When facing complex road conditions, drivers can rely on the system's precise analysis and warnings to make the best driving decisions, greatly reducing vehicle damage and traffic accidents caused by water-crossing. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 For the overall interaction of this application Figure 1 ;

[0039] Figure 2 A partial interactive diagram of this application;

[0040] Figure 3 For the overall interaction of this application Figure 2 . DETAILED DESCRIPTION

[0041] Three embodiments of the present application are described in detail below with reference to the accompanying drawings.

[0042] Example 1

[0043] like Figures 1 to 3 As shown, this embodiment relates to a vehicle wading warning system based on intelligent video analysis, which captures real-time video streams of the road in front of and around the vehicle through a camera, processes the video streams using a computing unit, and combines real-time weather data, historical geographic information, and traffic data to predict the risk of road waterlogging, issue warning information, and improve the overall warning effect through a fleet collaborative communication system.

[0044] The camera module is a high-definition waterproof camera installed on the front and side of the vehicle to capture real-time video streams of the road surface in front of and around the vehicle. The camera module can be set as a 360-degree panoramic camera to ensure coverage of a wide monitoring area and capture complete road information.

[0045] The computing unit embedded processor is responsible for receiving the video stream from the camera module and performing image processing in combination with the intelligent video analysis module based on the Transformer architecture. The computing unit is also responsible for connecting with the network communication module to obtain real-time and historical meteorological data, traffic data and geographic information, and combining it with the vehicle's local video stream for analysis.

[0046] The intelligent video analysis module is based on the collaborative work of convolutional neural networks (CNN) and Transformer models. CNN is used to extract spatial features from the video stream, such as waterlogged areas and road obstacles. The Transformer model embeds time information into video features through position encoding, performs time series analysis, and predicts the dynamic changes of waterlogging and potential water-related risks.

[0047] The video stream analysis module uses video stream data to analyze the distance between the vehicle and the ground and monitor the contact status of the tires to infer changes in tire pressure. It determines the height of the vehicle chassis and the current water accumulation through video data to evaluate the vehicle's wading ability.

[0048] The terrain analysis module obtains the terrain features of the road (such as slope and curve) from the video stream, and combines it with historical data from online queries to evaluate the water risk in the terrain area. The historical data includes information such as the frequency and depth of water accumulation in the area, which helps to better predict the possibility of water accumulation.

[0049] The dynamic change prediction module of water accumulation is based on the video stream collected by the camera, real-time weather data and terrain analysis results, combined with historical water accumulation data, to predict the dynamic change of water accumulation and determine whether the water accumulation on the road ahead will deepen or subside, thereby warning the driver in advance.

[0050] The cross-fleet water-wading warning module uses the V2V (vehicle-to-vehicle) communication protocol, allowing vehicles to share current road water-wading information. If a vehicle detects a water-wading risk, the system will transmit the information to other nearby vehicles through V2V communication, thereby enhancing the overall perception of the entire fleet of water-wading areas ahead.

[0051] When the vehicle is in distress, the satellite communication module will regularly send compressed images and distress signals to the rescue center. When the system is operating normally, it will automatically synchronize driving data, geographic location and environmental information to the cloud, so as to provide effective clues and data support in emergency situations.

[0052] When the system detects a risk of wading, the alarm module issues visual and audible alarms to the driver, prompting the driver to take action.

[0053] The communication module queries real-time and historical meteorological data, traffic flow data and geographic information through the Internet, and combines road video streams with historical data to improve the accuracy of water wading warnings. The communication module is also responsible for transmitting real-time road condition videos to other vehicles and traffic management centers.

[0054] Working process

[0055] The camera module captures the road video stream in front of the vehicle in real time, and the video stream is transmitted to the intelligent video analysis module through the computing unit. First, CNN extracts the spatial features of the waterlogged area from the video stream, and performs temporal feature analysis on the video stream through the Transformer model. Through position encoding, the system can correlate image features at different times and analyze the changing trend of water depth.

[0056] The communication module obtains meteorological, traffic and geographic data in real time through the network, and performs comprehensive analysis based on the image features in the video stream. For example, when the system detects water accumulation on the road ahead and finds historical data indicating that the area is prone to water accumulation, the system will increase the warning level to alert the driver.

[0057] The video stream analysis module calculates the distance between the vehicle chassis and the ground, the tire pressure, and combines the depth of the water to determine in real time whether the vehicle is suitable for passing through the flooded area ahead. The terrain analysis module uses camera data to analyze the slope of the road, road curvature and other features to assess the difficulty of wading.

[0058] The dynamic change prediction module of water accumulation predicts the changing trend of water accumulation based on the water depth on the road ahead and weather changes, and determines whether the water accumulation will deepen. If the prediction results indicate that the water accumulation may increase, the system will issue an emergency alarm to prompt the driver to take measures as soon as possible.

[0059] If a vehicle encounters difficulties in wading, the system will transmit the information to neighboring vehicles through the cross-fleet water warning module to alert other drivers. In addition, the satellite communication module will send a distress signal at regular intervals when in distress, provide the vehicle's location and compressed images to assist in rescue.

[0060] Through convolutional neural networks and Transformer architecture, the system can quickly identify waterlogged areas from video streams and accurately predict water changes through time series analysis. This intelligent analysis method greatly improves the accuracy of water-related warnings.

[0061] The system obtains real-time meteorological, historical traffic and geographic data through online queries, and combines this data with vehicle video streams to update waterlogging warning information in real time. This approach effectively avoids the problem of inaccurate warnings that may be caused by relying solely on local data.

[0062] By analyzing the vehicle's chassis height and tire pressure through video streams, the system can determine the vehicle's wading ability in real time, helping the driver make timely judgments and avoid vehicle damage caused by wading.

[0063] The cross-fleet water-fording warning function enables vehicles to collaboratively share water-fording information, improving the safety of the entire fleet, while the satellite communication function ensures that when a vehicle is in distress, distress signals and data can be transmitted in a timely manner to facilitate rescue operations.

[0064] Example 2

[0065] like Figures 1 to 3 As shown, based on Example 1, this embodiment further expands the system's water wading capability assessment function and cross-fleet collaborative warning function. The system monitors the vehicle tire pressure and chassis height in real time through a video stream analysis module, and performs water wading risk assessment based on the depth of water accumulation. At the same time, satellite communication and V2V (vehicle-to-vehicle) communication functions are used to achieve rescue in water distress and collaborative sharing of information by fleets.

[0066] The video stream analysis module is not only used to analyze water accumulation and terrain on the road, but also can capture the image features of the vehicle tires contacting the ground through the camera, monitor the pressure changes of the tires in real time, and infer whether the tire pressure is abnormal due to wading by analyzing the contact status of the tires with the ground. Combined with the water depth data, the vehicle's wading ability is evaluated.

[0067] The video stream analysis module uses a camera to capture the distance data between the vehicle chassis and the ground, and combines it with the depth of the water to assess whether the vehicle can safely pass through the flooded area. When the distance to the vehicle chassis is less than the depth of the water, the system will automatically issue a water risk warning and suggest that the driver choose to detour or slow down.

[0068] The cross-fleet water wading warning module is based on the V2V communication protocol, which enables nearby vehicles to share water wading information. Through V2V communication, vehicles can send and receive environmental information of the road ahead in real time, including water depth, terrain data and weather changes, thereby improving the accuracy and response speed of the overall water wading warning.

[0069] When the vehicle encounters water risks or gets into trouble, the satellite communication module will regularly send compressed images, GPS location information and vehicle status data to the rescue center to ensure that the vehicle's location information and environmental data are provided in a timely manner. In addition, during normal operation, the system will synchronize the vehicle's driving data, geographic location information and environmental information to the cloud to provide support for possible subsequent rescue operations.

[0070] Working process

[0071] The camera module continuously captures video streams on the road, and the video stream analysis module automatically analyzes the contact status of the tire with the ground and identifies the deformation of the tire through image features. When the system detects that the tire contact area is too large or the deformation is abnormal, it determines that the tire pressure may be abnormal, and prompts the driver that there is a risk of wading.

[0072] The camera module is also used to capture the distance between the vehicle chassis and the ground in real time. The system compares and analyzes the chassis height and the depth of water. When the depth of water is greater than the vehicle chassis height, the system will issue an alarm to remind the driver that the risk of wading is high and recommends detour or slow down. The wading ability assessment is achieved in this way to ensure that the driver can make timely decisions.

[0073] When a vehicle detects a risk of flooding, the cross-fleet flood warning module will send the information to other nearby vehicles through V2V communication. In this way, all vehicles in the fleet can share real-time information on the road ahead, including water depth, road conditions and weather data, thereby improving the overall warning effect and preventing other vehicles from entering dangerous areas.

[0074] When the vehicle is in trouble due to water or a serious risk of water is detected, the satellite communication module will automatically send a distress signal to the rescue center. The signal includes a compressed image of the vehicle, GPS location information and vehicle status data, which allows rescue personnel to quickly locate the vehicle and take rescue actions. In addition, when the system is operating normally, the vehicle's driving data and environmental information will be synchronized to the cloud to provide support in similar situations in the future.

[0075] By analyzing the contact between the tire and the ground through video stream, the system can detect abnormal changes in tire pressure in time. Especially when wading, abnormal tire pressure may cause vehicle loss of control or tire damage. Real-time monitoring helps to detect potential problems in time and reduce the risk of accidents.

[0076] By capturing the distance between the vehicle chassis and the ground through a camera, the system can accurately determine whether the vehicle has the ability to pass through flooded areas. This automated wading capability assessment enables the driver to make decisions in advance and avoid getting stuck in dangerous flooded areas.

[0077] Through V2V communication, each vehicle in the fleet can share water-crossing information, greatly improving the scope and accuracy of early warning. Compared with single-vehicle early warning, fleet collaborative communication can cover a wider area, prevent more vehicles from falling into the same water-crossing predicament, and improve road driving safety.

[0078] When a vehicle encounters serious water-related difficulties, the satellite communication module ensures that the vehicle can send out a distress signal in an environment with poor communication networks, and regularly sends compressed images and GPS location information to help rescue personnel quickly grasp the situation on site and provide data support for subsequent rescue. At the same time, the data synchronized to the cloud during normal operation of the system can provide valuable historical basis for future water-related warnings.

[0079] By expanding the video stream analysis module, Example 2 realizes more intelligent tire pressure monitoring and chassis height analysis, which can effectively improve the accuracy of vehicle wading ability assessment. In addition, the introduction of V2V communication module and satellite communication module enhances the fleet's collaborative early warning capability and the reliability of distress signal sending, further improving the overall safety and emergency response capabilities of the system.

[0080] Example 3

[0081] like Figures 1 to 3 As shown, this embodiment is based on Embodiment 1 and Embodiment 2, and describes in detail a working method of a vehicle wading warning system. The system processes the real-time video stream in front of the vehicle through an intelligent video analysis module, and analyzes the vehicle's wading risk in combination with real-time and historical data. When a wading risk is detected, the system can send a distress signal via satellite communication, and share wading information through V2V communication, thereby improving the wading warning capability of the entire fleet.

[0082] The working method of the vehicle wading warning system described in this embodiment includes the following steps:

[0083] S1: Video Stream Capture

[0084] High-definition waterproof cameras mounted on the front and sides of the vehicle capture real-time video streams of the road ahead and around the vehicle. The camera modules can be set to wide-angle or panoramic mode to ensure complete coverage of the area ahead of the vehicle's path. The video stream includes current road conditions, waterlogged areas and other relevant environmental features.

[0085] S2: Intelligent Video Analysis

[0086] In the computing unit, the video stream is processed by an intelligent video analysis module based on the Transformer architecture. The system first uses a convolutional neural network (CNN) to extract road spatial features from the video stream, including information such as the depth of water on the road, obstacles, and changes in road conditions. Subsequently, the Transformer model embeds time information into video features through position encoding, analyzes dynamic information such as water depth, road terrain, and weather changes, and assesses the vehicle's current water risk.

[0087] S3: Ride height and tire pressure analysis

[0088] The system combines historical data from online queries and real-time meteorological information, and calculates the distance between the vehicle chassis and the ground through the video stream analysis module. At the same time, the system analyzes the contact status of the tires with the ground and monitors changes in tire pressure. By comprehensively analyzing the chassis height, tire status and water depth, the system can evaluate the vehicle's wading ability in real time and determine whether there is a wading risk.

[0089] S4: Prediction of dynamic changes of water accumulation

[0090] The system uses a dynamic change prediction module for water accumulation to predict the dynamic changes of water accumulation on the road ahead based on real-time data captured by the camera, historical water accumulation information and terrain analysis results. For example, the system can predict whether the water accumulation will deepen or spread in the future based on real-time meteorological information (such as rainfall, wind speed) and historical data. Through this dynamic analysis, the system helps drivers predict water wading situations and avoid entering areas with deepening water accumulation.

[0091] S5: Satellite communication and distress signal transmission

[0092] When the vehicle detects that it is in distress (for example, the vehicle is stuck in water or the tires lose grip), the satellite communication module will regularly send compressed images of the vehicle, GPS location information and distress signals to the rescue center. At the same time, the system will synchronously upload the real-time captured environmental information (such as water depth, vehicle status, etc.) to the cloud to ensure that rescue personnel can obtain the latest environmental data in a timely manner and provide effective clues for subsequent rescue operations.

[0093] S6: V2V communication and multi-vehicle collaboration

[0094] During normal operation, the system shares water-related information with other vehicles through the V2V communication module. The current vehicle sends water depth, road conditions and weather data to neighboring vehicles through V2V communication, thereby improving the multi-vehicle collaborative water warning capabilities. In addition, the system synchronizes this data to the cloud or traffic management center to enhance the overall road safety. By collaboratively sharing information, multiple vehicles can avoid water risk areas in advance and prevent danger.

[0095] Through the camera module, the system continuously captures video streams of the road in front of and around the vehicle, and extracts spatial features related to water accumulation, roads and weather through the intelligent video analysis module.

[0096] Combining real-time meteorological data, historical water accumulation records and terrain data obtained through online query, the system dynamically analyzes the current water depth and the vehicle's wading ability, analyzes the distance between the chassis and the ground through video streams, and monitors tire pressure in real time to help the system determine the current water wading risk.

[0097] The system predicts the dynamic changes in the depth and range of water accumulation, and adjusts the water wading warning level based on the prediction results. When the water accumulation is likely to deepen, the system will issue a warning in advance to prevent the driver from entering the high-risk area.

[0098] Through V2V communication, the system will share water-crossing information with neighboring vehicles to improve the overall safety of the fleet. The system will also synchronize vehicle and environmental data to the cloud to ensure that there is sufficient data to support rescue operations in an emergency.

[0099] When a vehicle is in distress, the system automatically sends compressed images, location information and distress signals through the satellite communication module to ensure that rescuers can quickly locate and take measures. During the rescue process, the cloud data synchronization function further supports real-time decision-making.

[0100] By capturing road information in real time through video streaming and combining historical and meteorological data, the system can intelligently assess water risks, help drivers make more accurate judgments, and reduce water risks.

[0101] The dynamic change prediction function of water accumulation enables the system to predict the changing trend of water accumulation in advance, and provide early warning based on this to prevent vehicles from entering high-risk areas.

[0102] Through V2V communication, vehicles can share road condition information with other vehicles, forming a cross-fleet collaborative warning system to improve the driving safety of the entire fleet. This collaborative method is more effective than single-vehicle warning and can cover a wider area.

[0103] When a vehicle is stuck in water, the system can send real-time location information and environmental data via satellite communications to ensure that rescue personnel can quickly grasp the situation on site. In addition, the cloud synchronization function can provide data support for subsequent rescue operations, further improving rescue efficiency.

[0104] Through the working method of this embodiment, the vehicle wading warning system realizes a complete workflow from real-time video capture to intelligent analysis, dynamic prediction, cross-vehicle collaboration and distress communication. The system can not only intelligently assess the vehicle's wading risk, but also provide more efficient collaborative warning and rescue support through satellite communication and V2V communication.

[0105] In view of current practical needs, the above-mentioned implementation mode adopted in this application is not limited to the scope of protection. Various changes made within the knowledge scope of technical personnel in this field without departing from the concept of this application still fall within the scope of protection of the present invention.

Claims

1. A vehicle wading warning system, characterized by: A camera module mounted on a vehicle to capture a video stream of a road surface in front of and around the vehicle; A computing unit, used to process the video stream captured by the camera module, the computing unit comprising an intelligent video analysis module based on the Transformer architecture, the module obtaining road conditions, water depth, weather changes, terrain features and vehicle status from the video stream, and combining historical data queried online to assess the risk of wading ahead; The video stream analysis module is used to analyze the distance between the vehicle chassis and the ground through the video stream, monitor the tire pressure changes, and evaluate the vehicle's wading ability in combination with the road conditions; The terrain analysis module analyzes the current road terrain based on camera data and combines historical data obtained through network queries to evaluate the probability of wading in the current environment; The dynamic change prediction module of water accumulation is based on the video stream collected by the camera, real-time weather information and terrain analysis, combined with historical water accumulation data, to predict the change of water accumulation on the road ahead and determine whether the water accumulation will deepen or subside; The cross-fleet flood warning module uses the V2V communication system to automatically transmit flood information to adjacent vehicles when one vehicle detects flooding, enhancing the fleet's overall perception of the flooded area ahead. Satellite communication module, used to send compressed images and distress signals to the rescue center at regular intervals when the vehicle is in distress, and synchronize the data to the cloud during normal operation to provide clues for subsequent rescue; An alarm module that issues visual and audible alerts to the driver when the system detects a risk of wading; The communication module is used to obtain historical water accumulation and road information through network query, and transmit road surface video stream and water wading warning information to external devices, such as other vehicles or traffic management centers.

2. A vehicle wading warning system according to claim 1, characterized in that: The intelligent video analysis module includes a convolutional neural network (CNN) for extracting spatial features from the video stream, and the spatial features are further passed into the Transformer model for temporal feature analysis and prediction of water accumulation changes.

3. A vehicle wading warning system according to claim 1, characterized in that: The Transformer architecture embeds time information into image features in the video stream through position encoding, which is used to perform time series analysis on image features at different time points to accurately judge the dynamic changes of water accumulation and water wading risks.

4. A vehicle wading warning system according to claim 1, characterized in that: The system obtains real-time and historical meteorological data, traffic flow data and geographic information through online query by the communication module, and combines the data with real-time image data in the video stream to improve the accuracy of water-related warning.

5. A vehicle wading warning system according to claim 1, characterized in that: The video stream analysis module is used to monitor tire pressure changes in real time by analyzing image features of vehicle tires contacting the ground, determine whether tire pressure is abnormal due to wading, and conduct wading risk assessment based on the depth of water accumulation.

6. A vehicle wading warning system according to claim 1, characterized in that: The video stream analysis module is used to calculate the distance between the vehicle chassis and the ground from the video stream, and to evaluate the wading capability in combination with the depth of the accumulated water, and to provide corresponding wading warnings.

7. A vehicle wading warning system according to claim 1, characterized in that: The cross-fleet water wading warning module uses the V2V communication protocol to enable multiple vehicles to share road environment information in the video stream and historical data of network query, so as to collaboratively improve the accuracy and response speed of water wading warning.

8. The vehicle wading warning system according to claim 1, characterized in that: The satellite communication module regularly sends compressed images, GPS location information and vehicle status data to the rescue center when the vehicle is in distress, and synchronizes driving data to the cloud during normal operation to provide data support for subsequent rescue.

9. A vehicle wading warning system according to claim 1, characterized in that: Its working method includes the following steps: S1: Captures real-time video streams of the road ahead and around the vehicle through a camera module installed on the vehicle; S2: In the computing unit, the video stream is processed using an intelligent video analysis module based on the Transformer architecture to obtain information on road conditions, water depth, weather changes, terrain features, and vehicle status; S3: Combined with historical data and real-time meteorological information from online queries, the distance between the vehicle chassis and the ground is calculated through the video stream analysis module, and the tire pressure changes are monitored to determine the risk of vehicle wading; S4: Through the waterlogging dynamic change prediction module, based on the real-time data collected by the camera, historical waterlogging information, and terrain analysis results, the dynamic change of waterlogging is predicted to determine whether the waterlogging will deepen or subside; S5: When the vehicle detects that it is in distress due to water, it will send compressed images and distress signals to the rescue center through the satellite communication module at regular intervals, and synchronize real-time data to the cloud to provide clues for subsequent rescue; S6: During normal use, the V2V communication module is used to share water-related information with other vehicles, improving the water-related warning capabilities of multi-vehicle collaboration, and synchronizing data to the cloud or traffic management center to enhance overall road safety.

Images