Early warning method and vehicle

Abstract

This application provides a warning method and a vehicle. The warning method includes: acquiring vehicle driving information and image information of its surrounding environment; in response to detecting the presence of a target animal in the image information, determining the state information of the target animal based on the image information; determining a collision risk coefficient based on the state information of the target animal and the vehicle driving information; determining the collision risk level between the vehicle and the target animal based on the collision risk coefficient, and determining the warning strategy to be executed based on the collision risk level. This warning method can improve the accuracy and timeliness of collision accident prediction, and dynamically adjust the warning strategy according to the degree of collision risk, avoiding ineffective or excessive warnings while ensuring driving safety, thereby enhancing the driver's trust in the system and the overall reliability.

Description

Technical Field

[0001] This application relates to the field of driving safety warning technology, and in particular to a warning method and vehicle. Background Technology

[0002] When driving at night or in low light conditions, drivers may have difficulty spotting animals on the road in advance, leading to significant safety hazards.

[0003] The vehicles in the relevant technologies have limited ability to analyze animal behavior and cannot accurately predict the animal's movement trends. This makes it difficult for the vehicles to issue effective collision warnings to the drivers and also makes it difficult to provide reasonable avoidance measures to guide the drivers to adjust their driving behavior in a timely manner, increasing the safety risks during driving. Summary of the Invention

[0004] In view of this, the purpose of this application is to propose a warning method and vehicle to solve the technical problem that the vehicle cannot issue a collision warning to the driver and provide reasonable avoidance measures.

[0005] To achieve the above objectives, this application provides an early warning method, comprising:

[0006] Acquire vehicle driving information and image information of its surrounding environment; In response to detecting the presence of a target animal in the image information, the state information of the target animal is determined based on the image information; The collision risk coefficient is determined based on the state information of the target animal and the driving information of the vehicle. The collision risk level between the vehicle and the target animal is determined based on the collision risk coefficient, and a warning strategy to be implemented is determined based on the collision risk level. The early warning strategy includes generating a prompt message to alert the user and / or enabling the vehicle to avoid the target animal.

[0007] Furthermore, the image information includes multiple image frames, and the target animal's state information includes the target animal's predicted trajectory; The step of responding to the detection of a target animal in the image information and determining the state information of the target animal based on the image information includes: In response to detecting the presence of the target animal in the image frame at the current moment, among multiple consecutive image frames acquired after the current moment, the image frame containing the target animal is taken as the target image frame; In response to determining that the ratio between the target image frame and the plurality of consecutive image frames is greater than a preset percentage, the position of the target animal in each target image frame is determined; Predict the expected trajectory of the target animal based on its location.

[0008] Furthermore, the driving information includes driving trajectory, vehicle position, and vehicle speed; the target animal's status information also includes migration speed and animal type; The step of determining the collision risk coefficient based on the state information of the target animal and the driving information of the vehicle includes: In response to determining that a predicted collision point exists between the driving trajectory and the expected driving trajectory, a predicted collision distance between the predicted collision point and the vehicle position is determined; A first risk coefficient is determined based on the predicted collision distance and the vehicle speed; A second risk factor is determined based on the migration speed and the vehicle speed; A third risk factor is determined based on the animal type; The collision risk coefficient is determined based on at least one of the first risk coefficient, the second risk coefficient, and the third risk coefficient.

[0009] Further, determining the first risk coefficient based on the predicted collision distance and the vehicle speed includes: In response to determining that the predicted collision distance is less than or equal to a safe distance threshold, a first risk coefficient is determined based on the predicted collision distance and the vehicle speed; In response to determining that the predicted collision distance is greater than a safe distance threshold, the first risk coefficient is zero.

[0010] Further, determining the second risk coefficient based on the migration speed and the vehicle speed includes: The relative velocity component at the predicted collision point is determined based on the migration speed and the vehicle speed; The second risk coefficient is determined based on the relative velocity component and a preset risk reference threshold.

[0011] Further, determining the collision risk level between the vehicle and the target animal based on the collision risk coefficient, and determining the early warning strategy to be implemented based on the collision risk level, includes: In response to determining that the collision risk coefficient is less than a preset first coefficient threshold, the collision risk level is a low risk level, and the vehicle is controlled to display the status information of the target animal to notify the user; In response to determining that the collision risk coefficient is greater than or equal to a preset first coefficient threshold and less than a preset second coefficient threshold, the collision risk level is a medium risk level, and the vehicle is controlled to send a warning message to the user to alert the user. In response to determining that the collision risk coefficient is greater than or equal to a preset second coefficient threshold, the collision risk level is a high risk level, the vehicle is controlled to pretension the seat belts, and the vehicle steering avoidance function or emergency braking function is activated.

[0012] Furthermore, the state information of the target animal includes the target animal's posture; The response to determining that the collision risk coefficient is greater than or equal to a preset first coefficient threshold and less than a preset second coefficient threshold further includes: In response to determining that the target animal's posture is an active posture, the vehicle is controlled to generate a shooing signal corresponding to the collision risk coefficient to shoo away the target animal.

[0013] Furthermore, the image information includes multiple image frames; The response to the presence of a target animal in the image information includes: The image frame is culled to obtain the central region of interest of the image frame; In response to the detection of a complete animal image in the central area of ​​interest, the confidence level and size ratio of the animal image are determined based on the image frame; In response to determining that the confidence level is greater than a preset confidence threshold and the size ratio conforms to a preset ratio threshold range, the animal image is selected as the target animal.

[0014] Furthermore, the early warning method also includes: Based on the collision risk level, determine the display style and display area of ​​the target animal on the display interface; The target animal is displayed based on the display style and the display area.

[0015] Based on the same inventive concept, this application also provides a vehicle including an electronic device, the electronic device including a memory, a processor and a computer program stored in the memory and running on the processor, wherein the processor executes the program to implement the method described above.

[0016] As described above, the early warning method and vehicle provided in this application can provide data support for the accurate execution of target animal status detection and early warning strategies by acquiring vehicle driving information and image information of the surrounding environment. When a target animal is detected in the image information, its status information can be determined based on the image information, which can enrich the data dimensions used to assess the possibility of collision between the vehicle and the target animal, help extract the behavioral characteristics of the target animal, and improve the accuracy of predicting its movement trend. At the same time, by combining the target animal's status information and the vehicle's driving information, the corresponding collision risk coefficient can be determined to quantify the possibility of a collision in the current scenario. The collision risk coefficient can be used to classify the collision risk level, and an early warning strategy suitable for the current collision risk level can be applied accordingly to improve driving safety and active avoidance capabilities. By implementing this early warning method, the accuracy and timeliness of collision accident prediction can be improved, and the early warning strategy can be dynamically adjusted according to the degree of collision risk, avoiding ineffective or excessive warnings while ensuring driving safety, thereby enhancing the driver's trust in the system and the overall reliability. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in this application or related technologies, the drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the drawings described below are only embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a flowchart of the early warning method in the embodiments of this application; Figure 2 This is a flowchart illustrating the method for determining the expected trajectory of the target animal in this embodiment of the application. Figure 3 This is a flowchart illustrating the method for determining the collision risk coefficient in the embodiments of this application; Figure 4 This is a flowchart illustrating the method for determining the first risk coefficient in an embodiment of this application. Figure 5 This is a flowchart illustrating the method for determining the second risk coefficient in an embodiment of this application. Figure 6 This is a flowchart illustrating the method for collision risk level classification and early warning strategy execution in this application embodiment; Figure 7 This is a flowchart of the image frame noise reduction method in the embodiments of this application; Figure 8 This is a flowchart illustrating the display method of the target animal in the embodiments of this application; Figure 9 This is a structural block diagram of the early warning device in the embodiments of this application; Figure 10 This is a structural block diagram of the electronic device in the embodiments of this application. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with specific embodiments and the accompanying drawings.

[0020] It should be noted that, unless otherwise defined, the technical or scientific terms used in the embodiments of this application should have the ordinary meaning understood by one of ordinary skill in the art to which this application pertains. The terms "first," "second," and similar terms used in the embodiments of this application do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed after the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are only used to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0021] When driving at night or in low-light conditions, insufficient light makes it difficult for drivers to spot animals on the road in time, increasing the risk of collisions between vehicles and animals.

[0022] While some vehicles have incorporated animal detection capabilities, significant limitations remain in practical applications. On one hand, due to limitations in sensor performance and algorithm accuracy, the accuracy of animal identification is relatively low, leading to missed detections or false alarms. On the other hand, even when an animal is successfully detected, the warning information issued by the vehicle does not accurately reflect the actual risk level. This can result in frequent false alarms due to oversensitivity, missed opportunities for optimal intervention due to slow response, and inappropriate avoidance strategies due to incomplete perception of the animal's characteristics. Therefore, current warning strategies suffer from limitations such as simplistic detection methods and insufficient reliability. The persistent presence of these problems can lead to driver fatigue and potentially cause them to ignore or even resist vehicle warnings, thereby weakening the overall effectiveness of the vehicle's warning system.

[0023] Furthermore, the relevant technologies have relatively weak analytical capabilities for animal behavior, making it difficult to accurately predict animal movement trends and thus unable to provide forward-looking collision warnings. Moreover, due to a lack of in-depth understanding of animal behavior, vehicles also struggle to generate accurate and reasonable avoidance strategies, thereby failing to effectively guide drivers to take appropriate countermeasures in a timely manner, which increases safety risks during driving.

[0024] This application provides a warning method, and the warning method can be executed by a vehicle's warning controller or other control unit as the executing entity. Figure 1 As shown, the early warning method may include: S100: Acquire vehicle driving information and image information of its surrounding environment; In this step, the vehicle's driving information is generated during driving and reflects the vehicle's own state and operating status. This driving information can include vehicle speed, acceleration, steering angle, braking status, gear status, vehicle position, and heading angle. Utilizing this driving information allows for the understanding of the vehicle's dynamic behavior, providing necessary data support for vehicle control, driver assistance, and safety systems, ensuring the vehicle's stable and safe operation. This driving information can be measured and collected through onboard sensors.

[0025] Image information about the vehicle's surrounding environment is external scene data acquired through image acquisition devices. This includes road conditions, driving environment, pedestrian status, and the location and behavior of other vehicles and animals, presented in image form. The image information is used to provide relevant data for the vehicle to perceive and understand its external environment, and can support assisted driving functions such as target detection, obstacle recognition, path planning, and risk warning.

[0026] It should be noted that different types of image acquisition devices can be selected depending on the specific type of image information required about the vehicle's surrounding environment and the application scenario. For example, in environments with low visibility or insufficient light, to effectively detect targets such as animals or pedestrians, night vision camera modules or infrared thermal imaging camera modules can be used to accurately capture target information in front of and around the vehicle, reducing the interference of low-light environments on the device's perception performance. Furthermore, to further improve the system's reliability under complex weather conditions, the image acquisition device can also integrate auxiliary sensors such as millimeter-wave radar to supplement, measure, and cross-validate the acquired image information in adverse weather conditions such as dense fog, rain, and snow, enhancing the robustness and accuracy of the overall vehicle perception.

[0027] In practice, the early warning controller can collect vehicle driving information through onboard sensors and acquire image information of the surrounding environment by controlling the onboard image acquisition device. After collecting the driving information and the image information of the surrounding environment, the onboard sensors and image acquisition device can transmit the collected vehicle driving information and image information to the early warning controller through the transmission bus and data interface. Performing this step can provide corresponding data support for subsequent environmental perception, risk assessment and vehicle control adjustment.

[0028] S200: In response to detecting the presence of a target animal in the image information, determine the state information of the target animal based on the image information; In this step, the target animal is an animal in the environment surrounding the vehicle that can be identified by the vehicle's image acquisition device. Specifically, the presence of the target animal may pose a potential threat to driving safety, especially when the target animal migrates into the driving path in front of the vehicle, which may easily lead to a collision accident. In the image information, the target animal can be presented in the form of image data for subsequent analysis and utilization.

[0029] Furthermore, the target animal's state information is data that reflects the target animal's dynamic characteristics and may include the target animal's position, direction of movement, migration speed, and posture. It can be used to determine its behavioral trends and serve as a basis for assessing whether it poses a collision risk to the vehicle.

[0030] In practice, after the warning controller acquires image information of the vehicle's surrounding environment, it can analyze the image based on preset recognition rules (such as detecting whether there are organisms with heat source characteristics in the image) to determine whether a target animal is present in the image. When a target animal is detected in the image, the warning controller analyzes both the image information and the target animal to determine the target animal's status information based on the image information. This step confirms the presence of the target animal and provides relevant data for subsequent assessment of the potential collision risk between the target animal and the vehicle.

[0031] S300: Determine the collision risk factor based on the target animal's status information and the vehicle's driving information; In this step, the collision risk coefficient is a data indicator used to quantify the probability of a collision between a vehicle and a target animal. The collision risk coefficient can be determined based on three dimensions: spatial, temporal, and object attribute. Specifically, the spatial dimension reflects whether the vehicle's trajectory overlaps with the target animal's trajectory, determining if they intersect. The temporal dimension assesses their temporal proximity by predicting when they will arrive at the potential intersection point within a future time window, thus reflecting the urgency of the collision and determining whether they arrive at the intersection point synchronously. The object attribute dimension considers the impact of animal type (e.g., animal size, behavioral characteristics, and posture) on the degree of collision risk. Therefore, the collision risk coefficient derived by integrating these three dimensions more closely reflects real-world scenarios, dynamically characterizes the potential danger level in the current environment, and provides a reliable basis for early warning triggering and early warning strategy generation.

[0032] In practice, after acquiring the target animal's state information based on image information of the vehicle's surrounding environment, the warning controller combines this information with the vehicle's driving information to determine a corresponding collision risk coefficient. This coefficient quantifies the likelihood of a collision between the vehicle and the target animal in the current scenario, providing a quantitative basis for subsequent vehicle warnings. This step improves the accuracy of collision prediction, thereby increasing driver confidence in the warning strategy.

[0033] S400: Determine the collision risk level between the vehicle and the target animal based on the collision risk coefficient, and determine the warning strategy to be implemented based on the collision risk level; wherein, the warning strategy includes generating a prompt message to alert the user, and / or enabling the vehicle to avoid the target animal.

[0034] In this step, the collision risk level can be used to characterize the likelihood and urgency of a collision between the vehicle and the target animal. Essentially, it transforms the quantitative results of collision risk into a classification of the degree of danger during vehicle operation, so as to provide a clear and accurate basis for judgment for early warning strategies.

[0035] For example, collision risk levels can be divided into multiple levels such as low risk level, medium risk level and high risk level according to the magnitude of the collision risk coefficient.

[0036] The warning strategy refers to differentiated response measures preset for different collision risk levels. Specifically, using the example above as an illustration, when the risk level is low, the warning controller can control the dashboard or central control screen to display an icon or related information of the target animal for a mild warning. When the collision risk level rises to medium risk, the warning controller can issue a relatively obvious warning using at least one of the following methods: voice alarm, text prompt, flashing lights, or seat / steering wheel vibration. In the case of high risk, a combination of warnings is triggered, and the vehicle's active safety functions, such as seatbelt pretensioning, automatic emergency braking, or assisted avoidance control, are activated simultaneously to proactively avoid potential collisions. By implementing graded and appropriate intervention measures based on the severity of the risk, driving safety can be effectively ensured while avoiding driver fatigue or neglect due to frequent or excessive warnings, thereby improving the timeliness, rationality, and human-machine collaboration efficiency of the warnings.

[0037] In practice, after determining the collision risk coefficient based on the vehicle's driving information and the target animal's status information, the warning controller can classify the collision risk level between the vehicle and the target animal according to the collision risk coefficient, thereby accurately determining the degree of danger of a collision and improving the accuracy of predicting potential collision events. After determining the collision risk coefficient, the warning controller can generate corresponding warning strategies based on the determined collision risk level. Specifically, the warning strategies include generating prompts to alert the user, such as informing the driver of the presence of a target animal around the vehicle through visual, auditory, or tactile means, enabling the driver to be aware of the risk in advance and react in time. Simultaneously, when the detected status information of the target animal indicates that it interferes with vehicle driving or that the collision risk is relatively high, the warning controller can autonomously activate the vehicle's avoidance functions, such as pre-tensioning seat belts, emergency braking, or steering avoidance, to proactively avoid a collision if the driver is unable to react in time. This step enhances the targeting and timeliness of the warning, improving driving safety and proactive avoidance capabilities.

[0038] The early warning method provided in this application, by acquiring vehicle driving information and image information of its surrounding environment, can provide data support for the accurate execution of target animal status detection and early warning strategies. When a target animal is detected in the image information, its status information can be determined based on the image information, enriching the data dimensions used to assess the probability of collision between the vehicle and the target animal, helping to extract the behavioral characteristics of the target animal, and improving the accuracy of predicting its movement trend. At the same time, by combining the target animal's status information with the vehicle's driving information, a corresponding collision risk coefficient can be determined to quantify the probability of a collision in the current scenario. Using the collision risk coefficient, collision risk levels can be classified, and early warning strategies applicable to the current collision risk level can be applied accordingly to improve driving safety and active avoidance capabilities. By implementing this early warning method, the accuracy and timeliness of collision accident prediction can be improved, and the early warning strategy can be dynamically adjusted according to the degree of collision risk, avoiding ineffective or excessive warnings while ensuring driving safety, thereby enhancing the driver's trust in the system and overall reliability.

[0039] In some embodiments, the image information includes multiple image frames, and the target animal's state information includes the target animal's predicted trajectory. The image frames are continuous static images captured by the image acquisition device at consecutive time points, and each image frame reflects the visual data of the vehicle's surrounding environment at the corresponding moment. Multiple image frames are temporally correlated, and these temporally correlated image frames provide the basis for dynamic environmental perception for the early warning controller, supporting continuous detection, tracking, and behavioral analysis of the target animal.

[0040] The predicted trajectory of the target animal is based on a pre-set trajectory analysis model. It uses the positional changes of the target animal in multiple temporally related image frames to predict the trajectory that the target animal may move in the future time window. The trajectory that the target animal may move can reflect the animal's movement trend and can be used as the predicted trajectory of the target animal. It can provide data basis for determining whether the target animal may enter the vehicle's driving route, assessing the collision risk level, and formulating corresponding early warning strategies.

[0041] Based on the content described in S200, such as Figure 2 As shown, in response to detecting the presence of a target animal in the image information, the state information of the target animal is determined based on the image information, including: S210: In response to detecting that there is a target animal in the image frame at the current time, among the multiple consecutive image frames acquired after the current time, the image frame containing the target animal is taken as the target image frame; In this step, the continuous image frame is a series of image frames captured sequentially by the image acquisition device after the current moment, used to reflect the dynamic process of the vehicle's surrounding environment changing over time; the target image frame is a valid image frame among these continuous image frames that has been identified and confirmed to include the target animal, which can be used for subsequent tracking and analysis of the target animal's position, movement status, and expected trajectory.

[0042] In practice, the early warning controller detects whether a target animal exists in the currently acquired image frame to determine if the image frame can be used as a target image frame. Once a target animal is identified in the current image frame, the controller continuously receives multiple subsequent image frames and performs target matching and tracking detection on each frame to determine if the target animal is present. For image frames where a target animal is confirmed, the controller marks and stores them as target image frames, providing data support for determining the target animal's predicted trajectory. This step, by filtering out valid image frames related to the target animal, constructs a temporally correlated target image sequence, thus laying the data foundation for analyzing its movement trajectory and behavioral state based on multi-frame information.

[0043] S220: In response to determining that the ratio between the target image frame and multiple consecutive image frames is greater than a preset percentage, determine the position of the target animal in each target image frame; In this step, the preset percentage is a pre-set comparison threshold used to measure whether the proportion of image frames containing the target animal in multiple continuously acquired image frames is sufficient to support the trajectory analysis model in accurately determining the target animal's expected trajectory. This preset percentage ensures that the target animal continues to exist in the environment around the vehicle, thereby guaranteeing the reliability of subsequent trajectory analysis.

[0044] The location of the target animal is the specific spatial coordinates of the target animal relative to the vehicle or image coordinate system in each target image frame. It can be represented by pixel coordinates or transformed world coordinates, which can be used to accurately describe the spatial distribution of the target animal at the current moment. The location of the target animal can serve as the data basis for calculating the target animal's direction of movement, speed and its expected trajectory.

[0045] In practice, after acquiring multiple consecutive image frames, the early warning controller records the total number of these frames and counts the number of target image frames from which the target animal can be identified. Then, the controller calculates the ratio between the number of target image frames and the total number of consecutive image frames, comparing this ratio to a preset percentage to obtain a comparison result. When the ratio is greater than the preset percentage, it indicates that the target animal has sufficient visibility and persistence during the current observation period, providing a sufficient data foundation for the trajectory analysis model. At this point, the controller can process each target image frame, extracting the spatial coordinates of the target animal within the corresponding frame based on a preset image processing algorithm, thereby determining the target animal's location.

[0046] Considering the complexity and variability of the vehicle's surrounding environment, obstacles such as trees, guardrails, or other vehicles may obstruct the target animal, causing it to be undetectable in some image frames. Therefore, using a preset percentage as the evaluation criterion can effectively eliminate false positives caused by temporary occlusion or accidental missed detections. By performing this step, the extracted location data has good temporal continuity, which helps to accurately predict the target animal's trajectory.

[0047] S230: Predict the expected trajectory of the target animal based on its location.

[0048] In practice, after determining the position of the target animal in each target image frame in chronological order, the early warning controller can construct a spatial position sequence based on the determined positions of the target animal in multiple target image frames in chronological order. This spatial position sequence is then input into a preset trajectory analysis model. Using the trajectory analysis model, by fitting the changing trends of multiple position points and combining kinematic constraints and behavioral patterns, the potential travel path of the target animal within a preset future time window is calculated, forming a predicted travel trajectory. The predicted travel trajectory derived from this trajectory analysis model reflects the target animal's direction of movement, speed changes, and potential turning intentions, providing crucial input for subsequent assessment of its interaction with the vehicle's travel path and collision risk. By performing this step, the predicted travel trajectory of the target animal can be accurately predicted, improving the ability to predict animal dynamic behavior and providing reliable support for collision risk assessment and early warning strategies.

[0049] It should be noted that trajectory analysis models can be implemented using various algorithms, including but not limited to motion prediction models based on Kalman filtering or particle filtering, time-series prediction models based on recurrent neural networks (RNNs) or long short-term memory networks (LSTMs), and trajectory prediction models based on the Transformer architecture. The models mentioned above can effectively infer the movement trend of the target animal and generate a reasonable predicted trajectory based on the position sequence of the target animal in consecutive image frames.

[0050] In some embodiments, driving information includes driving trajectory, vehicle position, and vehicle speed; the target animal's state information also includes migration speed and animal type; wherein, the vehicle's driving trajectory is the path traversed by the vehicle during its journey, which can reflect the vehicle's historical stage, current moment, and planning stage of its driving route. For example, the vehicle's driving trajectory is provided by the vehicle's navigation system; the vehicle position can be the vehicle's specific coordinates in geographic space, which can be obtained by the vehicle's positioning system; the vehicle speed can refer to the speed at which the vehicle travels along its driving trajectory, used to characterize the speed of the vehicle's movement.

[0051] The migration speed of a target animal refers to the distance and direction it moves per unit of time, reflecting the speed and trend of its movement. Animal type refers to the species or category of the target animal, such as small animals like cats, dogs, and foxes; medium-sized animals like wolves and sheep; and large animals like bears, camels, deer, cattle, and horses. Different types of animals exhibit significant differences in size, behavior, and sudden reaction capabilities, which can directly affect the potential threat posed by the target animal to driving safety.

[0052] Based on the content described in S300, such as Figure 3 As shown, the collision risk coefficient is determined based on the target animal's status information and the vehicle's driving information, including: S310: In response to the determination that a predicted collision point exists between the vehicle trajectory and the expected trajectory, the predicted collision distance between the predicted collision point and the vehicle position is determined. In this step, the predicted collision point is the location in space where the vehicle's trajectory and the target animal's expected trajectory may intersect, reflecting the possibility of path overlap and potential contact between the two in the future; the predicted collision point is used to determine whether the vehicle and the target animal are likely to meet at the same location.

[0053] Predicted collision distance is the distance traveled from the vehicle's current position to the predicted collision point. It reflects the time and space margin required for the vehicle to reach the potential collision location, thus highlighting the urgency of a collision. Furthermore, by combining vehicle speed and other driving information, predicted collision distance can quantify the time window for collision risk, providing data for early warning and avoidance strategies.

[0054] In practice, after obtaining the vehicle's driving trajectory and the target animal's expected trajectory, the early warning controller performs spatial intersection analysis on the driving trajectory and the expected trajectory to determine whether there is an intersection point. When one or more intersection points are detected, the intersection point closest to the current vehicle position can be determined as the predicted collision point, indicating that there is a potential collision risk between the vehicle and the target animal in the spatial dimension.

[0055] Furthermore, based on the path between the current vehicle position and the predicted collision point, the warning controller can calculate the actual distance between the vehicle along its trajectory and the predicted collision point to obtain the predicted collision distance. Since the predicted collision distance reflects the distance the vehicle needs to travel from its current position to the predicted collision point at the potential collision location, combining this with vehicle speed can further estimate the time to reach the predicted collision point, providing parameters for determining the subsequent collision risk coefficient. By performing this step, potential predicted collision points in space can be effectively identified, the probability of the vehicle approaching the predicted collision point can be quantified, and the accuracy and timeliness of collision risk assessment can be improved.

[0056] S320: Determine the first risk factor based on the predicted collision distance and vehicle speed; In this step, the first risk coefficient is a quantitative value calculated based on the predicted collision distance and the current vehicle speed, used to characterize the urgency of the time required for the vehicle to reach the predicted collision point. The premise of the first risk coefficient is that a predicted collision point exists between the vehicle's trajectory and the target animal's expected trajectory. Based on this, the first risk coefficient reflects the probability that the vehicle and the target animal will be close in time. The shorter the predicted collision distance and the higher the vehicle speed, the shorter the time it takes for the vehicle to reach the predicted collision point, and the larger the first risk coefficient, indicating a more imminent collision risk; conversely, the risk is relatively milder. Therefore, the value of the first risk coefficient directly reflects the urgency of the collision event in the time dimension.

[0057] In practice, after determining the predicted collision point and the corresponding predicted collision distance, the warning controller, combined with the vehicle's current speed, can determine the predicted time required for the vehicle to travel from its current position to the predicted collision point. Subsequently, based on this predicted time, or directly using a preset functional relationship between the predicted collision distance and vehicle speed, a first risk coefficient can be generated. This first risk coefficient quantifies the urgency of the vehicle's approach to the predicted collision point, assuming the vehicle's path intersects with that of the target animal, thus providing a basis for subsequent collision risk level assessment. This step transforms the intersection information of the vehicle's trajectory and the target animal into a first risk coefficient reflecting the urgency of the collision, providing relevant support for subsequent comprehensive assessment of the collision risk level.

[0058] S330: Determine the second risk factor based on migration speed and vehicle speed; In this step, the second risk coefficient is a quantitative value derived from the migration speed of the target animal and the vehicle speed, used to characterize the approach trend of the vehicle and the target animal under relative motion. Specifically, the second risk coefficient reflects the dynamic interaction between the vehicle and the target animal over time: when the target animal moves rapidly towards the vehicle, or when the vehicle approaches a stationary or slowly moving animal at high speed, the relative speed is greater, the probability of a potential collision is higher, and the second risk coefficient increases accordingly; conversely, the second risk coefficient decreases. Therefore, the second risk coefficient can assess the degree of dynamic risk arising from relative motion, helping to determine the likelihood and urgency of a collision.

[0059] In practice, after acquiring the migration speed of the target animal and the current vehicle speed, the warning controller calculates the relative speed between the two in their respective directions of motion and generates a second risk coefficient based on this relative speed. This second risk coefficient characterizes the dynamic approach trend between the vehicle and the target animal due to their relative motion: when the target animal moves rapidly in the direction of the vehicle's travel, or when the vehicle approaches a stationary, slow-moving, or reverse-moving animal at high speed, the relative speed is higher, and the second risk coefficient increases accordingly, indicating a relatively high probability of collision between the vehicle and the target animal; conversely, the coefficient is lower. Through this step, the speed difference between the vehicle and the target animal is transformed into a quantifiable second risk coefficient, providing a dynamic dimension to support the comprehensive assessment of collision risk.

[0060] S340: Determine the third risk factor based on animal type; In this step, the third risk coefficient is a quantitative value determined based on the type of target animal. It reflects the differentiated impact of different animal species on driving safety due to factors such as their size, behavioral characteristics, and sudden reaction capabilities. For example, large animals (such as deer, cattle, and bears) have a higher risk coefficient due to their large mass and severe impact consequences; while small animals (such as cats, dogs, and foxes) may trigger evasive maneuvers, but the direct collision hazard is relatively low, resulting in a correspondingly lower risk coefficient. Therefore, the third risk coefficient reflects the corrective effect of the object attribute dimension on collision risk, making the risk assessment more closely reflect the actual degree of harm.

[0061] In practice, after identifying the type of target animal, the warning controller queries a preset risk mapping relationship to obtain the risk value corresponding to the target animal, and uses this risk value as a third risk coefficient. Different animal types pose varying degrees of threat to driving safety due to differences in size, behavior patterns, and sudden reaction capabilities. Specifically, when the identified target animal is large, the consequences of a collision with a vehicle are severe, thus the assigned third risk coefficient is relatively high; conversely, when the identified target animal is small, the assigned third risk coefficient is relatively low. This step transforms the attribute characteristics of the target animal type into a quantifiable third risk coefficient, which is used to reflect the impact of object attribute dimensions in comprehensive collision risk assessment, making the collision risk assessment closer to real-world collision scenarios.

[0062] For example, the pre-defined risk mapping relationship can be determined by querying Table 1; Table 1

[0063] Furthermore, considering the vehicle's driving environment, when the vehicle is driving in scenarios such as the wild, the type of the target animal can be comprehensively judged in conjunction with its protection level; the higher the protection level of the animal, the greater the third risk coefficient, and vice versa. This mechanism takes into account ecological protection factors when assessing collision risk, so that the early warning strategy not only reflects the degree of safety threat, but also reflects the priority avoidance consideration for rare or protected species.

[0064] S350: The collision risk coefficient is determined based on at least one of the first risk coefficient, the second risk coefficient, and the third risk coefficient.

[0065] In practice, after obtaining the first risk coefficient, the second risk coefficient, and the third risk coefficient, the early warning controller can process the three risk coefficients based on a preset fusion processing method to generate a collision risk coefficient. The preset fusion processing method can adopt a preset nonlinear combination function, and the weight of each risk coefficient can be dynamically adjusted according to the actual application scenario, vehicle status, or environmental conditions.

[0066] Furthermore, since the first risk coefficient reflects the urgency of time, the second risk coefficient characterizes the dynamic approach trend brought about by relative motion, and the third risk coefficient reflects the impact of animal type and its protection level on risk, the collision risk coefficient determined by combining one or more of the above risk coefficients can more comprehensively and accurately quantify the possibility of a collision between a vehicle and a target animal, providing a reliable basis for subsequent risk level classification and early warning strategy implementation.

[0067] For example, after obtaining the first risk coefficient, the second risk coefficient, and the third risk coefficient from the early warning controller, the collision risk coefficient can be calculated according to the following formula; (1); (2); In formulas (1) and (2), Indicates the collision risk factor; This can represent the first risk coefficient; It can represent the first weighting coefficient; This can represent the second risk coefficient; It can represent the second weighting coefficient; It can represent the third risk coefficient; The third weighting coefficient can be represented; that is, the collision risk coefficient can be determined based on formula (1) and formula (2), so formula (1) and formula (2) can be used as a preset fusion processing method.

[0068] It should be noted that, to ensure the rationality of the collision risk coefficient, the warning controller uses a weighted approach when combining the various risk coefficients, combining the first, second, and third risk coefficients, with the corresponding weights decreasing sequentially; that is, the first weighted coefficient is greater than the second weighted coefficient, and the second weighted coefficient is greater than the third weighted coefficient. Specifically, in the collision risk assessment process, the urgency of time (represented by the first risk coefficient) has the most critical direct impact on the occurrence of a collision, followed by the relative motion trend (reflected by the second risk coefficient), while object attribute factors such as animal type and protection level (reflected by the third risk coefficient) serve as auxiliary correction items. This setting highlights dynamic collision risks while also taking into account object characteristics, making the final determined collision risk coefficient more consistent with actual driving safety logic.

[0069] For example, the first weighting coefficient can be 0.5, the second weighting coefficient can be 0.3, and the third weighting coefficient can be 0.2.

[0070] In some embodiments, based on the content described in S320, such as Figure 4 As shown, the first risk factor is determined based on the predicted collision distance and vehicle speed, including: S321: In response to determining that the predicted collision distance is less than or equal to a safe distance threshold, a first risk factor is determined based on the predicted collision distance and vehicle speed; In this step, the safe distance threshold is a pre-set distance threshold used to characterize the minimum safe distance that a vehicle must maintain to avoid colliding with a target animal under the current driving conditions. This safe distance threshold can be flexibly set according to factors such as the vehicle's braking performance, driver reaction time, and road conditions. For example, the vehicle's braking performance can be negatively correlated with the safe distance threshold. Or, for example, the higher the driver's fatigue level, the longer their reaction time, so the driver's reaction time can be positively correlated with the safe distance threshold.

[0071] In practice, after obtaining the predicted collision distance, the warning controller compares it with a preset safe distance threshold. When the warning controller determines that the predicted collision distance is less than or equal to the preset safe distance threshold, it indicates that the vehicle is close to the predicted collision point and there is a potential collision risk. At this point, the warning controller determines the expected collision time based on the predicted collision distance and the current vehicle speed, and generates a first risk coefficient based on the expected collision time and predicted collision distance, combined with a preset first functional relationship. This step ensures that the first risk coefficient with significant impact is output only when there is an actual imminent risk to the vehicle, thereby avoiding over-response to distant targets or low-speed scenarios and improving the pertinence and rationality of collision risk assessment.

[0072] For example, the preset first functional relationship can be an inverse proportional or exponential decay function, so that the shorter the predicted collision distance and the higher the vehicle speed, the greater the first risk coefficient; wherein, the first risk coefficient can be determined according to the following formula to obtain the first risk coefficient; (3) in, This indicates the first risk factor. Indicates the estimated collision time. This represents a time variable (i.e., a time slice), with a value range of [value range missing]. Used to represent the time from the current moment to Any time within; Indicates the safe distance threshold; Indicates at time The minimum Euclidean distance between the vehicle's trajectory and the target animal's predicted trajectory (i.e., the distance between the closest points of the two trajectories); therefore, This indicates the entire expected collision time period. The minimum trajectory distance among all time variables is the predicted collision distance. In other words, when the predicted collision distance is less than or equal to the safe distance threshold, the corresponding first risk coefficient can be determined based on formula (3). Therefore, formula (3) can be used as a preset first functional relationship.

[0073] S322: In response to determining that the predicted collision distance is greater than the safe distance threshold, the first risk factor is zero.

[0074] In practice, after obtaining the predicted collision distance, the warning controller compares it with a preset safe distance threshold. When the warning controller determines that the predicted collision distance is greater than the preset safe distance threshold, it indicates that the vehicle is relatively far from the predicted collision point, and there is no potential collision risk or the urgency of the collision is low. Therefore, the warning controller sets the first risk coefficient to zero, indicating that there is no immediate risk of an emergency collision in the time dimension. Applying this step can effectively filter out distant target animals, prevent the vehicle from triggering unnecessary warnings due to premature or excessive responses, and help improve the stability and practicality of the vehicle's overall warning logic.

[0075] In some embodiments, based on the content described in S330, such as Figure 5 As shown, the second risk factor is determined based on the migration speed and vehicle speed, including: S331: Determine the relative velocity component at the predicted collision point based on the migration speed and vehicle speed; In this step, the relative velocity component at the predicted collision point is the projected difference in velocity of the vehicle and the target animal in the direction of common action when they move in the same or opposite directions at the intersection, which is used to reflect the dynamic trend of the two approaching or moving away from each other near the predicted collision point.

[0076] Furthermore, the relative velocity component, combined with the angle between the vehicle's direction of travel and the target animal's direction of migration, can more accurately reflect the actual approach rate at the moment when the vehicle's trajectory and the target animal's expected trajectory intersect in space; that is, the larger the relative velocity component, the more intense the dynamic conflict between the two at the predicted collision point, and the higher the potential collision risk; conversely, the risk is lower.

[0077] In practice, after determining the migration speed of the target animal and the vehicle speed, the early warning controller combines their respective movement directions at the predicted collision point, projects their vector velocities onto the direction of the line connecting them or the direction of their combined action, and calculates the velocity difference to obtain the relative velocity component at the predicted collision point. This relative velocity component at the predicted collision point reflects the actual dynamic trend of the vehicle and target animal approaching or moving away from each other near the intersection, thus assessing whether a conflict will occur in the near future. Extracting this component effectively eliminates interference from motion in vertical or irrelevant directions, allowing the calculation of the second risk coefficient to focus more on the motion components that truly affect the probability of a collision, thereby improving the accuracy of risk assessment.

[0078] S332: Determine the second risk coefficient based on the relative velocity component and the preset risk reference threshold.

[0079] In this step, the preset risk reference threshold is a reference relative speed value used to characterize a medium-risk level, such as 50 km / h, corresponding to the typical approach speed at which a vehicle and a target animal may pose a collision risk in the direction of the predicted collision point. This preset risk reference threshold serves as a benchmark parameter in the calculation of the relative speed factor, reflecting the risk classification of dynamic conflict intensity by the warning controller: when the relative speed component equals the preset risk reference threshold, the second risk coefficient is at a medium risk; when the relative speed component is higher than the preset risk reference threshold, it indicates a shorter reaction time for the driver or system, thus a higher collision risk, and the second risk coefficient increases accordingly; conversely, the collision risk is lower, and the second risk coefficient decreases accordingly. Therefore, the preset risk reference threshold plays a role in anchoring the risk level in the preset second functional relationship, enabling the second risk coefficient to continuously and reasonably reflect the time urgency and collision probability at different relative speeds.

[0080] In practice, after obtaining the relative speed component, the warning controller substitutes the relative speed component and a preset risk reference threshold into a preset second functional relationship to generate a second risk coefficient. The preset second functional relationship can be a normalized function and can take the form of an S-shaped function, using the risk reference threshold as the center point corresponding to medium risk. Thus, when the relative speed component equals this threshold, the second risk coefficient is at an intermediate level; as the relative speed component increases, the function output approaches 1, indicating an increased risk; when the relative speed component decreases, the output approaches 0, indicating a decreased risk. Therefore, by executing this step, the second risk coefficient can continuously and smoothly reflect the dynamic risk level caused by the relative motion between the vehicle and the target animal at the predicted collision point, allowing the driver or vehicle sufficient reaction time.

[0081] For example, the pre-defined second functional relationship can be determined according to the following formula to obtain the second risk coefficient: (4); in, This indicates the second risk factor; This represents the relative velocity components between the vehicle and the target animal in the predicted collision direction; This indicates a preset risk reference threshold; This represents the preset slope in the preset second functional relationship. In other words, when the relative velocity component and the preset risk reference threshold are obtained, the corresponding second risk coefficient can be determined based on formula (4). Therefore, formula (4) can be used as a preset second functional relationship.

[0082] In some embodiments, based on the content described in S400, such as Figure 6As shown, the collision risk level between the vehicle and the target animal is determined based on the collision risk coefficient, and the early warning strategy to be implemented is determined based on the collision risk level, including: S410: In response to determining that the collision risk coefficient is less than a preset first coefficient threshold, the collision risk level is low risk level, and the vehicle is controlled to display the target animal's status information to notify the user. In this step, the first coefficient threshold is a pre-set lower limit critical value for the collision risk coefficient, which can be used as the boundary between low-risk levels and other higher-risk levels. The first coefficient threshold reflects the minimum judgment standard of the warning controller for the acceptable risk level: when the collision risk coefficient is lower than this value, it indicates that the possibility of a collision between the vehicle and the target animal is small, there is sufficient time or the consequences are minor, the collision risk is at a low level, there is no need to initiate active intervention measures, and it is only necessary to inform the driver through information prompts.

[0083] In practice, after obtaining the collision risk coefficient, the warning controller compares it with a preset first threshold. When the collision risk coefficient is less than the first threshold, it indicates that the probability of a collision between the vehicle and the target animal in the current scenario is low, and the collision risk is within a controllable range. At this time, the warning controller does not trigger emergency intervention measures. Instead, it presents the target animal's status information to the driver through the in-vehicle display device, including its location, type, migration speed, and expected trajectory, providing a gentle informational prompt. This step enhances the driver's perception of the surrounding environment, avoids unnecessary alarm interference caused by low-risk events, helps maintain the driver's trust and attention in the warning system, and ensures driving safety.

[0084] S420: In response to determining that the collision risk coefficient is greater than or equal to a preset first coefficient threshold and less than a preset second coefficient threshold, the collision risk level is medium risk level, and the vehicle is controlled to send a warning message to the user to warn the user. In this step, the second coefficient threshold is a pre-set upper limit critical value for the collision risk coefficient, used to define the boundary between medium-risk and high-risk levels. The second coefficient threshold reflects the warning controller's judgment standard for the risk level that requires active warning but not yet emergency intervention.

[0085] In practice, after obtaining the collision risk coefficient, the warning controller compares it with a preset first coefficient threshold. When the collision risk coefficient is greater than or equal to the first coefficient threshold but less than a second coefficient threshold, the current scenario is determined to be of medium risk. At this point, there is a potential collision risk between the vehicle and the target animal. Even if it doesn't reach the level requiring emergency avoidance, it exceeds the warning range for low-risk scenarios. Therefore, the warning controller can send warning information to the driver through sound, light, or touch. For example, it can highlight the target animal's location on the dashboard or head-up display with an audible alert to attract the driver's attention and prompt them to take evasive action such as slowing down or steering. This warning strategy, without excessively interfering with driving, can improve the timeliness of response to medium-risk events and enhance driving safety.

[0086] S430: In response to determining that the collision risk coefficient is greater than or equal to a preset second coefficient threshold, the collision risk level is high risk level, the vehicle is controlled to pretension the seat belts, and the vehicle steering avoidance function or emergency braking function is activated.

[0087] In practice, after obtaining the collision risk coefficient, the warning controller compares it with a preset first and second threshold values. When the collision risk coefficient is greater than or equal to the second threshold value, it determines that there is a highly imminent collision risk between the vehicle and the target animal, requiring active intervention. At this point, the warning controller can control the vehicle to perform at least one avoidance function. For example, it can control the seatbelt pretensioner to tighten the occupants' seatbelts to reduce displacement and injury during a collision. Alternatively, based on the vehicle's current driving status, the surrounding environment, and system capabilities, it can automatically activate steering avoidance or emergency braking to actively avoid a collision or minimize the collision speed. By executing this step, passive safety preparation and active avoidance actions can be integrated in critical scenarios, effectively improving occupant protection and overall vehicle safety, thus reliably responding to serious collisions between the vehicle and the target animal.

[0088] In some embodiments, the state information of the target animal includes the target animal's posture; wherein, the target animal's posture is the body state presented by the target animal in an image or scene. Specifically, the target animal's posture may include an active posture and a lying posture; an active posture may indicate that the target animal is in a state of movement such as standing, walking, or running, indicating that the target animal has the ability to move actively or has the potential to move; a lying posture may indicate that the animal is stationary on the ground, possibly in a state of rest, injury, or inability to move. Based on the analysis of the target animal's posture, it is helpful to determine the potential behavioral trends of the target animal and the degree of impact on driving safety.

[0089] The step of responding to determining that the collision risk coefficient is greater than or equal to a preset first coefficient threshold and less than a preset second coefficient threshold further includes: responding to determining that the target animal's posture is an active posture, controlling the vehicle to generate a shooing signal corresponding to the collision risk coefficient to shoo away the target animal.

[0090] Among them, the driving signals used to drive away the target animal can be warning signals emitted by the vehicle to attract the attention of the target animal and prompt it to leave the driving path, such as sound waves of a specific frequency, flashing lights, or directional sound and light combinations, so as to stimulate the animal's auditory or visual perception and induce it to produce avoidance behavior, thereby reducing the risk of collision.

[0091] In practice, after the warning controller determines that the collision risk coefficient is greater than or equal to the first coefficient threshold and less than the second coefficient threshold, it indicates that the collision risk level between the vehicle and the target animal is medium or high, thus posing a potential threat to driving safety. At this point, to further reduce the collision risk with the target animal, the animal's posture information can be further analyzed. Specifically, if it is confirmed that the animal is in an active posture, it indicates that the target animal has the ability to move autonomously. Based on the current collision risk coefficient, a preset mapping strategy is invoked to generate a repelling signal of corresponding intensity and type. The repelling signal can include sound waves of a specific frequency, directional flashes, or a combination of sound and light, with its intensity increasing as the collision risk coefficient increases, thereby prompting the target animal to actively leave the vehicle's path. This step introduces active intervention in medium-risk or high-risk scenarios, reducing the potential collision probability by utilizing the animal's natural response to environmental stimuli without triggering emergency braking or steering, thus improving driving safety and eco-friendliness.

[0092] In some embodiments, based on the content described in S200, such as Figure 7 As shown, the image information includes multiple image frames; in response to the presence of a target animal in the image information, it includes: S240: Perform a culling process on the image frame to obtain the central area of ​​interest of the image frame; In this step, the removal process involves cropping, masking, or suppressing regions in the original image frame that are irrelevant to or significantly interfere with the detection of the target animal. This reduces the impact of background noise, occlusions, or edge distortion on the accuracy of animal recognition, thereby improving the efficiency and accuracy of subsequent target animal detection and localization.

[0093] The central region of interest (CRI) of an image frame is the core area retained after the removal process. The CRI of an image frame is located in the center of the image frame or is dynamically adjusted according to the vehicle's driving direction. The CRI reflects the relatively important environmental range of the vehicle in the current driving state, which helps to improve the real-time performance and robustness of target animal detection and trajectory analysis.

[0094] In practice, after acquiring multiple image frames, the early warning controller performs a removal process for each image frame to improve the accuracy and consistency of target animal identification. This involves removing interfering regions such as edges far from the road, areas with severe distortion, occlusion, or low information content, based on preset rules or the vehicle's current driving state. The controller retains the core region located in the center of the image frame or dynamically adjusted along the vehicle's direction of travel as the central focus area. This step, by removing redundant or interfering areas, reduces the computational burden of subsequent image processing and improves the accuracy and reliability of target animal detection and localization.

[0095] S250: In response to detecting a complete animal image in the central area of ​​interest, determine the confidence level and size scale of the animal image based on the image frame; In this step, the confidence level of the animal image is based on the degree of confidence that the identified image region belongs to the target animal, according to the preset target detection algorithm. Its value ranges from 0 to 1. The higher the confidence level, the greater the probability that the identified animal image is a real animal, which can effectively reflect the reliability of the recognition result and help filter out false detections or noise interference.

[0096] The size ratio of an animal image refers to the ratio of the length to the width of the area occupied by the image in the image frame. It is used to characterize the morphological features and relative scale of the target animal in the field of vision, and to help determine whether the displayed content conforms to the typical appearance features of the real animal, thereby improving the reliability of animal recognition results.

[0097] In practice, the early warning controller runs a target detection algorithm within the central area of ​​interest (CIFO) of the image frame. Once a complete and clearly defined animal image is detected within CIFO, the confidence level and size ratio of the animal image are determined based on the image frame. The authenticity of the animal image is then verified based on these factors. Confidence level reflects the reliability of the recognition result, while size ratio characterizes the morphological features of the target in the field of view. The effective combination of these two factors can accurately distinguish between real animals and external interference factors. This step improves the accuracy of subsequent target animal confirmation and collision risk assessment.

[0098] S260: In response to determining that the confidence level is greater than a preset confidence threshold and the size ratio conforms to a preset ratio threshold range, the animal image is used as the target animal.

[0099] In this step, the preset confidence threshold is a pre-defined confidence limit used to determine whether the target detection result is reliable enough to be identified as a real animal. The preset confidence threshold reflects the minimum requirement of the warning controller for recognition accuracy. Therefore, only when the confidence is higher than this value can the detection result of the animal image be considered to have sufficient credibility, avoiding misjudging noise, shadows, or blurry objects as target animals. For example, the preset confidence threshold can be 0.01.

[0100] The preset ratio threshold range is a reasonable range set based on the typical aspect ratios that real animals may appear in images. This range is used to reflect whether the target's shape conforms to the morphological characteristics of common animals, excluding abnormal size ratios caused by occlusion, distortion, or non-animal objects. By simultaneously meeting the judgment conditions of confidence level and size ratio, the accuracy and robustness of target animal screening are improved. For example, the preset ratio threshold range is 1:4 to 4:1.

[0101] In practice, after obtaining the confidence level and size ratio of an animal image, the warning controller compares these values ​​with preset confidence and size ratio thresholds, respectively. Specifically, if the confidence level is greater than or equal to the confidence threshold, it indicates that the detection result for the animal image is sufficiently reliable. Simultaneously, if the size ratio of the animal image falls within the preset size ratio threshold range, it indicates that the animal's physical characteristics conform to the typical form of a real animal. Therefore, when these two conditions are met, the warning controller can determine that the animal image is the target animal. This step, by combining recognition confidence and geometric plausibility, filters out false positives and anomaly detection results, ensuring that only highly reliable, morphologically plausible animal objects are used for driving safety decisions, thereby improving the accuracy and stability of the entire warning system.

[0102] In some embodiments, such as Figure 8 As shown, the early warning methods also include: S510: Determine the display style and display area of ​​the target animal on the display interface based on the collision risk level; In this step, the display style of the target animal refers to the visual presentation of the target animal in the human-computer interaction interface, including color, icon type, border style, flashing status or additional markings, etc. Different display styles can directly reflect the current collision risk level. For example, a green static icon is used for a low risk level, a yellow highlighted or flashing icon is used for a medium risk level, and a red dynamic warning symbol is used for a high risk level.

[0103] The display area of ​​the target animal refers to the size of the pixel area occupied by the target animal on the display interface; the display area of ​​the target animal can be dynamically adjusted according to the relative distance between the target animal and the vehicle or the urgency of the risk. Specifically, the higher the collision risk level or the closer the target is, the larger the display area of ​​the target animal, thereby enhancing the driver's attention to and perception priority of key targets.

[0104] In practice, after determining the collision risk level, the warning controller can dynamically configure the presentation of the target animal on the in-vehicle display interface according to preset mapping rules. For example, for low-risk levels, the target animal is displayed with a small display area and a green static icon; for medium-risk levels, the display area of ​​the target animal is moderately increased, and a yellow highlighted or flashing icon is used to attract attention; for high-risk levels, the display area of ​​the target animal is further increased, and a red dynamic warning symbol, a bold border, or additional warning signs are used for prominent indication, thereby transforming the collision risk level into an intuitive visual parameter. This step enables the driver to quickly identify the severity and urgency of potential threats, improves the clarity and response efficiency of human-machine interaction, avoids information overload, and ensures that key risk information is effectively conveyed.

[0105] S520: Displays the target animal based on the display style and display area.

[0106] In practice, the warning controller applies the predetermined display style and area to the target animal graphic rendering process in the in-vehicle human-machine interface. Following specified colors, icon types, border features, dynamic effects, and pixel dimensions, it draws a visual identifier of the target animal at the corresponding position on the display screen. This identifier is overlaid on a real-time or simulated road scene image, ensuring its position matches the actual spatial orientation of the target animal. By strictly adhering to visual parameters determined by the collision risk level, the warning controller can provide differentiated and hierarchical visual feedback at different collision risk levels, enabling the driver to quickly perceive the presence, distance, and degree of danger of the target, thereby assisting them in making timely and reasonable driving decisions.

[0107] It should be noted that some publicly disclosed vehicle warning methods based on animal collisions identify and track animal targets by collecting image data in real time during vehicle operation, analyzing the animal's movement state based on the predicted collision time, and issuing different levels of warnings accordingly to improve the accuracy of animal collision warnings and driving safety. Some related technologies disclose a method for animal risk warning when driving on grassland roads. In the grassland road scenario, infrared cameras and preset classification models are used to identify and determine the type of animals. The collision time is calculated by combining radar or ultrasonic ranging. When the collision time is less than the preset takeover request time, an avoidance warning is issued through tactile and auditory means. At the same time, vehicle speed warning is provided to improve the driver's ability to react in advance and take over effectively.

[0108] The early warning method provided in this application has the following differences and advantages compared to the aforementioned related technologies: First, the early warning method provided in this application can identify the presence and status information of the target animal, and can construct a more comprehensive and reliable collision risk assessment system to enhance the accuracy of early warning strategy execution. Second, by introducing different collision risk coefficients to determine the corresponding collision risk levels, the collision risk between the vehicle and the target animal can be quantified and layered, making the collision risk judgment more logical and adaptable. In addition, differentiated early warning strategies can be dynamically matched according to different collision risk levels, which can improve safety while avoiding excessive intervention, and enhance system reliability and human-machine collaboration efficiency.

[0109] It should be noted that the method in this embodiment can be executed by a single device, such as a computer or server. The method can also be applied in a distributed scenario, where multiple devices cooperate to complete the task. In such a distributed scenario, one of these devices may execute only one or more steps of the method in this embodiment, and the multiple devices will interact with each other to complete the method described.

[0110] It should be noted that the above description describes some embodiments of this application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in a different order than that shown in the above embodiments and still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0111] Based on the same inventive concept, corresponding to any of the above embodiments, this application also provides an early warning device.

[0112] refer to Figure 9 The early warning device includes an information acquisition module 11, a detection and processing module 12, a coefficient determination module 13, and an early warning control module 14. The information acquisition module is configured to acquire vehicle driving information and image information of its surrounding environment; The detection processing module is configured to determine the state information of the target animal based on the image information in response to detecting the presence of a target animal in the image information; The coefficient determination module is configured to determine the collision risk coefficient based on the state information of the target animal and the driving information of the vehicle. The early warning control module is configured to determine the collision risk level between the vehicle and the target animal based on the collision risk coefficient, and to determine the early warning strategy to be executed based on the collision risk level. The early warning strategy includes generating a prompt message to alert the user and / or enabling the vehicle to avoid the target animal.

[0113] For ease of description, the above devices are described in terms of function, divided into various modules. Of course, in implementing this application, the functions of each module can be implemented in one or more software and / or hardware.

[0114] The apparatus of the above embodiments is used to implement the corresponding early warning method in any of the foregoing embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0115] Based on the same inventive concept, corresponding to the methods of any of the above embodiments, this application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the early warning method described in any of the above embodiments.

[0116] Figure 10 This embodiment illustrates a more specific hardware structure of an electronic device. The device may include a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, memory 1020, input / output interface 1030, and communication interface 1040 are interconnected internally via the bus 1050.

[0117] The processor 1010 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.

[0118] The memory 1020 can be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 1020 can store the operating system and other applications. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program code is stored in the memory 1020 and is called and executed by the processor 1010.

[0119] The input / output interface 1030 is used to connect input / output modules to realize information input and output. Input / output modules can be configured as components within the device (not shown in the figure) or externally connected to the device to provide corresponding functions. Input devices may include keyboards, mice, touchscreens, microphones, various sensors, etc., while output devices may include displays, speakers, vibrators, indicator lights, etc.

[0120] The communication interface 1040 is used to connect a communication module (not shown in the figure) to enable communication between this device and other devices. The communication module can communicate via wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).

[0121] Bus 1050 includes a pathway for transmitting information between various components of the device, such as processor 1010, memory 1020, input / output interface 1030, and communication interface 1040.

[0122] It should be noted that although the above-described device only shows the processor 1010, memory 1020, input / output interface 1030, communication interface 1040, and bus 1050, in specific implementations, the device may also include other components necessary for normal operation. Furthermore, those skilled in the art will understand that the above-described device may only include the components necessary for implementing the embodiments of this specification, and not necessarily all the components shown in the figures.

[0123] The electronic devices described above are used to implement the corresponding early warning methods in any of the foregoing embodiments and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0124] Based on the same inventive concept, corresponding to the methods of any of the above embodiments, this application also provides a vehicle, which includes the electronic equipment described in any of the above embodiments and possesses all the advantages and beneficial effects of the electronic equipment.

[0125] Based on the same inventive concept, corresponding to the methods of any of the above embodiments, this application also provides a non-transitory computer-readable storage medium that stores computer instructions for causing the computer to execute the early warning method as described in any of the above embodiments.

[0126] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transfer medium that can be used to store information accessible by a computing device.

[0127] The computer instructions stored in the storage medium of the above embodiments are used to cause the computer to execute the early warning method as described in any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0128] Based on the same concept, corresponding to any of the above embodiments, this application also provides a computer program product, including computer program instructions, which, when run on a computer, cause the computer to perform the method described in any of the above embodiments, and have the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0129] It is understood that before using the technical solutions of the various embodiments in this application, users will be informed of the type, scope of use, and usage scenarios of the personal information involved in an appropriate manner, and user authorization will be obtained.

[0130] For example, upon receiving a user's active request, a prompt message is sent to the user to explicitly inform them that the requested operation will require the acquisition and use of the user's personal information. This allows the user to independently choose, based on the prompt message, whether to provide personal information to the software or hardware such as electronic devices, applications, servers, or storage media performing the operations described in this application.

[0131] As an optional but not limited implementation, in response to a user's active request, sending a prompt message to the user can be done via a pop-up window, where the prompt message can be presented in text format. Furthermore, the pop-up window can also include a selection control allowing the user to choose "agree" or "disagree" to provide personal information to the electronic device.

[0132] It is understood that the above notification and user authorization process is merely illustrative and does not limit the implementation of this application. Other methods that comply with relevant laws and regulations may also be applied to the implementation of this application.

[0133] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of this application is limited to these examples; under the concept of this application, the technical features of the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations of different aspects of the embodiments of this application as described above, which are not provided in detail for the sake of brevity.

[0134] Additionally, to simplify the description and discussion, and to avoid obscuring the embodiments of this application, the well-known power / ground connections to integrated circuit (IC) chips and other components may or may not be shown in the provided drawings. Furthermore, the apparatus may be shown in block diagram form to avoid obscuring the embodiments of this application, and this also takes into account the fact that the details of the implementation of these block diagram apparatuses are highly dependent on the platform on which the embodiments of this application will be implemented (i.e., these details should be fully understood by those skilled in the art). While specific details (e.g., circuits) have been set forth to describe exemplary embodiments of this application, it will be apparent to those skilled in the art that the embodiments of this application can be implemented without these specific details or with variations thereof. Therefore, these descriptions should be considered illustrative rather than restrictive.

[0135] Although this application has been described in conjunction with specific embodiments thereof, many substitutions, modifications, and variations of these embodiments will be apparent to those skilled in the art from the foregoing description. For example, other memory architectures (e.g., dynamic RAM (DRAM)) may be used with the embodiments discussed.

[0136] The embodiments of this application are intended to cover all such substitutions, modifications, and variations that fall within the broad scope of the claims of this application. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the embodiments of this application should be included within the protection scope of this application.

Claims

1. An early warning method, characterized in that, include: Acquire vehicle driving information and image information of its surrounding environment; In response to detecting the presence of a target animal in the image information, the state information of the target animal is determined based on the image information; The collision risk coefficient is determined based on the state information of the target animal and the driving information of the vehicle. The collision risk level between the vehicle and the target animal is determined based on the collision risk coefficient, and a warning strategy to be implemented is determined based on the collision risk level. The early warning strategy includes generating a prompt message to alert the user and / or enabling the vehicle to avoid the target animal.

2. The early warning method according to claim 1, characterized in that, The image information includes multiple image frames, and the target animal's state information includes the target animal's predicted trajectory. The step of responding to the detection of a target animal in the image information and determining the state information of the target animal based on the image information includes: In response to detecting the presence of the target animal in the image frame at the current moment, among multiple consecutive image frames acquired after the current moment, the image frame containing the target animal is taken as the target image frame; In response to determining that the ratio between the target image frame and the plurality of consecutive image frames is greater than a preset percentage, the position of the target animal in each target image frame is determined; Predict the expected trajectory of the target animal based on its location.

3. The early warning method according to claim 2, characterized in that, The driving information includes the driving trajectory, vehicle location, and vehicle speed; the target animal's status information also includes migration speed and animal type. The step of determining the collision risk coefficient based on the state information of the target animal and the driving information of the vehicle includes: In response to determining that a predicted collision point exists between the driving trajectory and the expected driving trajectory, a predicted collision distance between the predicted collision point and the vehicle position is determined; A first risk coefficient is determined based on the predicted collision distance and the vehicle speed; A second risk factor is determined based on the migration speed and the vehicle speed; A third risk factor is determined based on the animal type; The collision risk coefficient is determined based on at least one of the first risk coefficient, the second risk coefficient, and the third risk coefficient.

4. The early warning method according to claim 3, characterized in that, Determining the first risk coefficient based on the predicted collision distance and the vehicle speed includes: In response to determining that the predicted collision distance is less than or equal to a safe distance threshold, a first risk coefficient is determined based on the predicted collision distance and the vehicle speed; In response to determining that the predicted collision distance is greater than a safe distance threshold, the first risk coefficient is zero.

5. The early warning method according to claim 3, characterized in that, The determination of the second risk coefficient based on the migration speed and the vehicle speed includes: The relative velocity component at the predicted collision point is determined based on the migration speed and the vehicle speed; The second risk coefficient is determined based on the relative velocity component and a preset risk reference threshold.

6. The early warning method according to claim 1, characterized in that, The step of determining the collision risk level between the vehicle and the target animal based on the collision risk coefficient, and determining the early warning strategy to be implemented based on the collision risk level, includes: In response to determining that the collision risk coefficient is less than a preset first coefficient threshold, the collision risk level is a low risk level, and the vehicle is controlled to display the status information of the target animal to notify the user; In response to determining that the collision risk coefficient is greater than or equal to a preset first coefficient threshold and less than a preset second coefficient threshold, the collision risk level is a medium risk level, and the vehicle is controlled to send a warning message to the user to alert the user. In response to determining that the collision risk coefficient is greater than or equal to a preset second coefficient threshold, the collision risk level is a high risk level, the vehicle is controlled to pretension the seat belts, and the vehicle steering avoidance function or emergency braking function is activated.

7. The early warning method according to claim 6, characterized in that, The target animal's state information includes its posture; The response to determining that the collision risk coefficient is greater than or equal to a preset first coefficient threshold and less than a preset second coefficient threshold further includes: In response to determining that the target animal's posture is an active posture, the vehicle is controlled to generate a shooing signal corresponding to the collision risk coefficient to shoo away the target animal.

8. The early warning method according to claim 1, characterized in that, The image information includes multiple image frames; The response to the presence of a target animal in the image information includes: The image frame is culled to obtain the central region of interest of the image frame; In response to the detection of a complete animal image in the central area of ​​interest, the confidence level and size ratio of the animal image are determined based on the image frame; In response to determining that the confidence level is greater than a preset confidence threshold and the size ratio conforms to a preset ratio threshold range, the animal image is selected as the target animal.

9. The early warning method according to claim 1, characterized in that, Also includes: Based on the collision risk level, determine the display style and display area of ​​the target animal on the display interface; The target animal is displayed based on the display style and the display area.

10. A vehicle, characterized in that, The device includes an electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor, when executing the program, implements the method as claimed in any one of claims 1 to 9.

Images