Passenger car blind area early warning method, system and equipment and storage medium

Abstract

The invention belongs to the field of passenger car blind area early warning, and particularly discloses a passenger car blind area early warning method, system and device and a storage medium, a visual field missing area caused by vehicle body posture change is recognized by comparing visual field coverage areas of a vehicle under a straight-going working condition and a turning working condition, a dynamic blind area can be accurately recognized, and the passenger car blind area early warning efficiency is improved. The problem of blind area omission caused by static modeling is avoided, the comprehensiveness and accuracy of blind area recognition are ensured, and meanwhile, when the visual field missing area is recognized, real-time recognition and motion trail capture of a dynamic target are carried out in an adjacent visible area focused on the periphery of the visual field missing area. And early warning information is generated in time based on potential conflict analysis between the captured target movement track and the vehicle driving track, so that the early-stage discovery capability of potential conflicts is improved, a driver has more time to take danger avoiding measures, traffic accidents caused by blind areas are effectively prevented, and the safety of pedestrians, non-motor vehicles and other vehicles is guaranteed.

Description

Technical Field

[0001] The invention belongs to the field of blind spot warning for passenger buses, and specifically discloses a blind spot warning method, system, equipment and storage medium for passenger buses. Background Art

[0002] With the advancement of urbanization and the continuous development of public transportation, buses, as a key means of urban travel, are responsible for transporting a large number of people. However, due to their large body structure and limited driver's field of view, there are significant blind spots around the vehicle, especially when turning. This can easily lead to a loss of perception of pedestrians, non-motorized vehicles, and other traffic participants, posing a safety hazard. Therefore, monitoring and early warning of blind spots on buses is extremely necessary.

[0003] There are also technical solutions for blind spot warning of passenger buses in the existing technology. For example, the Chinese invention patent with publication number CN113744532A proposes a blind spot warning method and device for urban traffic passenger buses based on vehicle-road collaboration. By obtaining the vehicle's position information and the vehicle's inherent visual blind spot parameters, the vehicle's coordinate system is constructed, and the regional representation function of each sub-blind spot is established based on the blind spot geometric characteristics; further combined with the vehicle-mounted and roadside sensing equipment, the position information of surrounding traffic participants is obtained and converted into the vehicle's coordinate system; finally, according to the vehicle's current driving behavior, the target position is substituted into the corresponding blind spot function model to realize blind spot risk identification and warning.

[0004] The above scheme effectively utilizes vehicle-road cooperative technology to enhance the perception capability of blind spot targets in complex urban traffic environments. However, this scheme still has certain limitations in practical applications: first, its blind spot modeling is mainly based on static geometric parameters, and does not fully consider the dynamic migration and deformation of the blind spot range caused by changes in vehicle body posture under turning conditions. This static modeling method is prone to cause deviations between the detection of the blind spot area and the actual visible range when the vehicle is in a non-straight driving state, thereby affecting the accuracy of risk identification; secondly, the scheme lacks effective prediction of the motion trend of dynamic targets in target behavior analysis, and fails to establish a dynamic assessment of the potential conflict between the blind spot target entry path and the vehicle driving trajectory. Therefore, it is difficult to identify potential threats in a timely manner when facing complex traffic behaviors such as sudden approach and cross-crossing, resulting in delayed warning generation, affecting the timeliness and reliability of the warning. Summary of the Invention

[0005] To this end, one purpose of the embodiments of the present application is to provide a passenger bus blind spot warning method, system, equipment and storage medium. By focusing on the blind spots generated by the passenger bus under turning conditions, the movement trends and trajectories of dynamic targets in the blind spots are identified and predicted in real time, thereby achieving accurate warning of potential conflict risks and effectively solving the problems existing in the existing technology.

[0006] The purpose of the present invention can be achieved through the following technical solutions: A passenger bus blind spot warning method, comprising the following steps: S1. A wide-angle camera group deployed around the vehicle body generates a first panoramic field of view coverage area under straight-ahead conditions, and generates a second panoramic field of view coverage area under turning conditions.

[0007] S2. Spatially superimpose the first panoramic field of view coverage area and the second panoramic field of view coverage area, and mark the field of view missing area of ​​the second panoramic field of view coverage area relative to the first panoramic field of view coverage area.

[0008] S3. Perform an adjacent area search operation within the second panoramic view coverage area based on the spatial position of the missing view area and its outline boundary information to determine a visible area adjacent to the missing view area.

[0009] S4. Perform dynamic target detection on the visible area adjacent to the missing field of view area, and record the last spatial coordinates and motion vector of the target at the moment of disappearance.

[0010] S5. Using the final spatial coordinate as the starting point, generate a predicted trajectory line according to the motion vector, and calculate the conflict time window between the predicted trajectory and the vehicle body trajectory based on the real-time motion trajectory of the bus.

[0011] S6. Mark and warn the blind spots in the areas where vision is lost based on the conflict time window.

[0012] Combining all the above technical solutions, the positive effects of the present invention are as follows: 1. By comparing the field of view coverage of the vehicle under straight-moving conditions and turning conditions, the present invention identifies the field of view loss areas caused by changes in vehicle body posture, can accurately identify dynamic blind spots, avoid blind spot omissions caused by static modeling, and ensure the comprehensiveness and accuracy of blind spot identification.

[0013] 2. When the present invention identifies an area with lost vision, it focuses on the adjacent visible area around the area to perform real-time recognition of dynamic targets and capture their motion trajectories. Furthermore, based on the potential conflict analysis between the captured target motion trajectory and the vehicle's driving trajectory, early warning information is generated. This improves the ability to detect potential conflicts early, giving drivers more time to take risk avoidance measures, effectively preventing traffic accidents caused by blind spots, and ensuring the safety of pedestrians, non-motor vehicles, and other vehicles. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The present invention is further described with reference to the accompanying drawings. However, the embodiments in the accompanying drawings do not constitute any limitation to the present invention. A person skilled in the art can obtain other drawings based on the following drawings without creative effort.

[0015] Figure 1 This is a diagram of the implementation steps of a blind spot warning method for a passenger car in the present invention.

[0016] Figure 2 This is a flow chart of the conflict time window operation for analyzing the target predicted trajectory and the real-time motion trajectory of the bus in the present invention.

[0017] Figure 3 This is a schematic diagram of the blind area marking and warning operation for the visual field loss area in the present invention.

[0018] Figure 4 This is a module connection diagram of a blind spot warning system for passenger vehicles in the present invention. DETAILED DESCRIPTION

[0019] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0020] Example 1

[0021] See also Figure 1 As shown, the present invention proposes a blind spot warning method for a passenger car, comprising the following steps: S1. generating a first panoramic field of view coverage area under a straight-ahead condition and a second panoramic field of view coverage area under a turning condition through a wide-angle camera group deployed circumferentially of the vehicle body.

[0022] As an implementable method of the above steps, the specific generation process of the first panoramic field of view coverage area is as follows: when the vehicle is in a straight-ahead condition, the wide-angle camera group deployed around the vehicle body is activated, and each camera synchronously captures images.

[0023] In the aforementioned image acquisition process, wide-angle cameras installed around the vehicle capture 360-degree information about the surrounding environment. Each camera monitors a specific direction, ensuring there are no blind spots.

[0024] A rectangular projection coordinate system parallel to the ground is constructed with the longitudinal axis of the vehicle body as the axis of symmetry.

[0025] In a preferred embodiment, the boundary of the rectangular plane is determined by the length and width of the vehicle body, and is symmetrically extended with the longitudinal axis of the vehicle body as the axis of symmetry, forming a rectangular area whose center is symmetrical to the longitudinal axis of the vehicle body, and its range covers the visible area within a certain distance around the vehicle. Specifically, the rectangular projection coordinate system is established as follows: the geometric center of the vehicle is used as the origin of the coordinate system, which is usually the reference base point of the vehicle coordinate system, so as to facilitate consistency with the vehicle kinematic model.

[0026] Set the X-axis direction: Extending forward along the longitudinal axis of the vehicle body, indicating the vehicle's forward direction.

[0027] The Y-axis is perpendicular to the longitudinal axis of the vehicle body and extends horizontally to the right, indicating the lateral direction of the vehicle. The Z-axis is perpendicular to the ground and extends upward, indicating the height direction.

[0028] It should be understood that the above-mentioned rectangular projection coordinate system is constructed with the longitudinal axis of the vehicle body because the longitudinal axis of the vehicle body is the main direction of vehicle movement and is also the reference base for the driver to perceive the environment, steer and avoid obstacles. Establishing a projection coordinate system with this axis as the axis of symmetry can ensure that the constructed panoramic image is consistent with the direction of movement of the vehicle, thereby providing a unified geometric reference framework for subsequent field of view coverage analysis, target trajectory prediction and conflict judgment. In addition, if the projection coordinate system is not aligned with the vehicle body structure during the multi-camera image stitching process, it is easy to cause geometric distortion, perspective misalignment or ghosting of the image during the stitching process, affecting the integrity and accuracy of the panoramic field of view. Establishing a rectangular projection coordinate system with the longitudinal axis of the vehicle body as the axis of symmetry ensures that the images of each camera are spatially aligned and coordinately transformed in a unified coordinate system, thereby improving the accuracy and visual consistency of image stitching and reducing stitching errors.

[0029] The original wide-angle images output by each camera are subjected to distortion correction and coordinate transformation, and each image is projected from its original viewing angle into the above-mentioned rectangular projection coordinate system.

[0030] The distortion correction mentioned above is due to the optical distortion problems such as barrel distortion in wide-angle lenses. First, the original images obtained by each camera need to be corrected for distortion.

[0031] The rectified images are stitched into a continuous rectangular field of view through image stitching.

[0032] The above-mentioned pixel-level fusion of the corrected images through image stitching eliminates the visual differences and geometric misalignment in the overlapping areas of the images, and generates a continuous, seamless rectangular panoramic image, which is used to represent the surrounding environment range that the driver can effectively perceive through the visual system when the vehicle is in a straight-ahead state.

[0033] As another possible implementation method of the above steps, when it is detected that the turn signal is activated or the steering wheel angle exceeds a preset threshold, the wide-angle camera group located around the vehicle body is synchronously activated to start real-time image acquisition.

[0034] In the aforementioned operations, triggering image acquisition when the vehicle is about to or in the process of turning helps promptly capture newly lost areas of vision due to changes in vehicle posture, thereby improving the real-time and dynamic adaptability of blind spot recognition. A preset threshold for steering wheel angle is used to identify whether the vehicle is entering a turning condition with a significant risk of blind spot changes. This threshold reflects the critical point at which the vehicle's steering behavior significantly affects the blind spot range. Specifically, data on common vehicle steering behaviors at intersections and mountain roads is collected, and the distribution of common steering angles is calculated. This allows the identification of the minimum angle threshold that effectively represents turning behavior without affecting normal driving operations.

[0035] The projection coordinate system with the longitudinal axis of the vehicle body as the axis of symmetry is dynamically adjusted according to the steering angle of the vehicle's front wheels.

[0036] Applied to the above operation, when the vehicle is moving straight ahead, the projected coordinate system is symmetrically distributed around the longitudinal axis of the vehicle, forming a centrally symmetrical rectangular projection area. However, when the vehicle turns, the originally symmetrical blind spot area shifts due to the change in the vehicle's posture, with the blind spot primarily concentrated on the inside of the turn.

[0037] To more accurately map the visual coverage area at this point, the front wheel steering angle is obtained during a turn. This angle reflects the vehicle's current steering range, and the projection coordinate system is dynamically offset based on the front wheel steering angle. Specifically, the center point of the projection area is shifted a certain distance inward of the turn based on the steering direction and angle, bringing the projection area closer to the vehicle's actual visual range. The X-axis of the projection coordinate system is aligned with the vehicle's current direction of travel, ensuring geometric consistency of the image projection.

[0038] This operation takes into account the changes in the vehicle's posture during the turn, ensuring that the coordinate system accurately reflects the vehicle's actual position and orientation relative to the ground.

[0039] The original images output by each wide-angle camera are subjected to distortion correction and coordinate transformation and then mapped into a projection coordinate system that is dynamically adjusted according to the steering angle of the vehicle's front wheels.

[0040] The corrected images are stitched into a continuous curved edge field of view through image stitching.

[0041] S2. Spatially superimpose the first panoramic field of view coverage area and the second panoramic field of view coverage area, and mark the field of view missing area of ​​the second panoramic field of view coverage area relative to the first panoramic field of view coverage area.

[0042] The specific implementation process of the above steps is as follows: for the first panoramic field of view coverage area and the second panoramic field of view coverage area, the closed envelope boundaries of their outer contours are extracted respectively.

[0043] It should be pointed out that the closed envelope boundaries of the above two visual field coverage areas are used as spatial geometric boundary representations of their respective visible areas.

[0044] The closed envelope boundary of the first panoramic field of view coverage area is used as a reference benchmark to geometrically superimpose the closed envelope boundary of the second panoramic field of view coverage area with it, and the first panoramic field of view coverage area that is not covered by the second panoramic field of view coverage area is extracted as the field of view missing area.

[0045] It should be added that before geometrically superimposing the closed envelope boundaries of the two covered domains, the two closed envelope boundaries need to be uniformly mapped to the same reference coordinate system and the coordinates need to be normalized or aligned to eliminate the geometric errors caused by coordinate offset or rotation.

[0046] It should be understood that the aforementioned lost field of view area represents the visual range lost when the vehicle switches from a straight-moving state to a turning state.

[0047] S3. Perform an adjacent area search operation within the second panoramic view coverage area based on the spatial position of the missing view area and its outline boundary information to determine a visible area adjacent to the missing view area.

[0048] As an optional implementation of the above steps, the outline boundary of the field of view loss area is marked in the second panoramic field of view coverage area based on the spatial position of the field of view loss area and its outline boundary information.

[0049] Boundary adjacency analysis is performed based on the contour boundaries of the lost field of view area to identify image areas that share common boundaries with the lost field of view area. The boundary features of these image areas are further traced until continuous image areas with no image stitching are identified. These continuous areas that are not disturbed by stitching and share edges with the lost field of view area are defined as adjacent visible ranges.

[0050] It should be understood that the image area that shares a common boundary with the area with lost vision is regarded as an adjacent visible area because the image area with a common boundary is directly adjacent to the area with lost vision in space, and belongs to the spatial neighborhood of the blind spot change area. The target in this area has a high probability of being visible before the blind spot is formed, or re-entering the field of vision after the blind spot disappears. Therefore, it is a key area for dynamic target recognition and trajectory prediction. The target usually appears in the adjacent visible area before entering or leaving the blind spot. Through continuous monitoring of this area, the system can capture the target's entry trajectory, movement direction and speed information, and provide a data basis for subsequent trajectory prediction and conflict warning of blind spot targets. It only focuses on monitoring the area adjacent to the blind spot, rather than performing indiscriminate analysis of the entire panoramic field of view, which helps to reduce unnecessary image processing calculations and improve system response efficiency.

[0051] In particular, the image stitching process may introduce visual discontinuities, ghosting, or mismatches in the stitched area due to factors such as differences in field of view between cameras, projection transformation errors, or uneven lighting. Therefore, when identifying an area that shares a boundary with the missing field of view area, the area surrounding the missing field of view without stitching is selected as the adjacent visible range to ensure that the extracted adjacent visible range has spatial geometric consistency and image integrity.

[0052] The adjacent visible range is divided according to the image acquisition source to which it belongs to generate visible sub-areas.

[0053] The image acquisition source mentioned above corresponds to the wide-angle camera.

[0054] S4. Perform dynamic target detection on the visible area adjacent to the missing field of view area, and record the last spatial coordinates and motion vector of the target at the moment of disappearance.

[0055] As an innovative implementation of the above solution, moving objects entering the area are identified in real time in the adjacent visible sub-areas around the area without vision. Once a dynamic target is detected, its motion trajectory is continuously tracked.

[0056] During the tracking process, if a dynamic target disappears from its visible sub-region in the current image frame, a target matching search operation is performed in the spatially adjacent regions of the visible sub-region within a set time window to determine whether the target has migrated to an adjacent field of view area. If the target cannot be detected again in the adjacent region, the moment when the target disappears is recorded, and the adjacent visible sub-region where it disappeared is located.

[0057] A set number of historical image frames are traced back from the disappearance moment to extract the spatial position information of the target in the continuous frames.

[0058] The displacement vector of the center of mass of the target between adjacent frames is calculated based on the position coordinates of the target in consecutive frames. The displacement vector reflects the movement trend and direction characteristics of the target before it disappears.

[0059] Applied to the above scheme, the displacement vector expression of the center of mass between adjacent frames is ,in and Respectively represent the target in Frame and The coordinates of the center of mass of the frame.

[0060] See also Figure 2 As shown, S5. Taking the last spatial coordinate as the starting point, a predicted trajectory line is generated according to the motion vector, and the conflict time window between the predicted trajectory and the vehicle body trajectory is calculated in combination with the real-time motion trajectory of the bus.

[0061] As an implementation method of the above steps: based on the direction of the motion vector of the dynamic target when it disappears in the adjacent visible sub-area, it is determined whether it is pointing to the lost field of view area. If it is pointing to the lost field of view area, it means that the dynamic target may be moving towards the blind spot of the vehicle. At this time, linear extrapolation is used to generate a straight line prediction trajectory along the direction of its motion vector with the last visible position of the target as the starting point. It is used to simulate the potential movement path of the dynamic target after entering the blind spot.

[0062] It should be understood that linear extrapolation means assuming that the target's motion trend remains unchanged under the premise of known motion states such as position, speed and direction, and extending its motion path along its current motion direction through mathematical modeling methods to predict its position or trajectory at a certain moment in the future.

[0063] In one embodiment, the specific implementation process of linear extrapolation is as follows:

[0064] Input conditions: The location coordinates of the target at the last visible moment are known , the target's motion vector including velocity is known and direction angle .

[0065] Assume that the target keeps moving in a straight line at a uniform speed after entering the blind spot, along the direction of its motion vector. As the starting point, a straight line prediction trajectory is generated, and any point on the trajectory It can be expressed as ,in is a time variable, which indicates the time elapsed after the target enters the blind spot.

[0066] The real-time bus trajectory is determined based on the vehicle steering wheel angle and vehicle speed.

[0067] In the specific implementation of the above scheme, in order to achieve high-precision calculation of the real-time motion trajectory of the bus, the real-time rotation speed of the four wheels and the front wheel steering angle of the vehicle are collected, and the vehicle wheel speed is used to estimate the longitudinal speed of the vehicle. ,in Indicates the The angular velocity of the wheels, Indicates the The tire radius of each wheel, Indicates the wheel number, .

[0068] The front wheel steering angle is used to calculate the vehicle's heading angle. The specific heading angle expression is, in represents the front wheel steering angle, Indicates wheelbase.

[0069] Vehicle position update ,in Indicates that the vehicle is The travel position coordinates at the moment, Indicates the moment, represents the sampling time interval, Indicates that the vehicle is The heading angle at the moment.

[0070] The real-time bus movement trajectory can be obtained based on the real-time updated vehicle movement position.

[0071] It's important to note that the above analysis of the bus's trajectory uses the vehicle's own sensor data, such as wheel speed and steering angle, to infer the vehicle's trajectory. This is a form of dead reckoning that does not rely on an external positioning system and achieves high-precision calculation of the bus's trajectory.

[0072] The predicted trajectory of the dynamic target entering the lost field of view area is geometrically compared with the movement trajectory of the bus to determine whether the two trajectories intersect. If there is an intersection, it means that the dynamic target and the vehicle may appear in the same spatial position at some point in the future, which means there is a potential collision risk. At this time, the shortest spatial distance between the two trajectories is calculated and compared with the set safety threshold. If the shortest spatial distance is less than the safety threshold, the conflict time window is calculated based on the shortest spatial distance and the relative speed between the target and the vehicle body. When there is no intersection, it means that the dynamic target and the vehicle are unlikely to appear in the same spatial position in a certain period of time in the future, and the probability of a collision is small. The shortest spatial distance between the two trajectories is also calculated. If the shortest spatial distance is less than the safety threshold, the spatial distance between the two trajectories continues to be updated. If the distance is detected to be continuously shortened in multiple consecutive time steps, the conflict time window calculation is retriggered.

[0073] The shortest spatial distance mentioned above refers to the Euclidean distance between the two nearest points on the paths of the predicted trajectory of the dynamic target and the moving trajectory of the bus. This distance is used to measure the spatial proximity between the target and the vehicle.

[0074] The safety threshold above sets a critical distance. When the shortest spatial distance between a dynamic object and a vehicle is less than this threshold, a high collision risk is considered. For example, statistical analysis of a large amount of traffic accident data can be used to identify the typical minimum distance between vehicles and objects before an accident occurs, which can serve as a reference for the safety threshold.

[0075] The conflict time window (CTW) is calculated by dividing the shortest spatial distance by the relative velocity between the target and the vehicle. This CTW reflects the time interval between the current moment and the predicted occurrence of a potential conflict and quantifies the response time available for the driver or the autonomous driving system to take evasive action. A shorter CTW indicates a faster approaching speed between the dynamic target and the vehicle, or a shorter remaining safety time. This means the system faces a higher time-criticality risk and needs to respond effectively in a shorter time to avoid a potential collision.

[0076] It should be understood that when two trajectories intersect, the shortest spatial distance usually occurs near the intersection. Using the shortest spatial distance for calculation can accurately reflect the spatial interval between the dynamic target and the vehicle at which a potential collision is about to occur. This helps the system issue a timely warning before a potential collision occurs, providing the driver with sufficient reaction time to take emergency braking or other evasive actions, thereby effectively avoiding the occurrence of a collision.

[0077] When two trajectories do not intersect, the shortest spatial distance refers to the distance between the two nearest points on the two non-intersecting trajectories. If this distance is greater than the set safety threshold, it is considered that there is no direct collision risk. If it is less than the safety threshold, there may still be potential danger, and it is necessary to continue monitoring the relative position change trend of the two. If it is found that the shortest spatial distance between the two is continuously decreasing, the conflict time window must still be calculated based on the shortest spatial distance to assess the risk level of potential collision and determine whether an early warning is needed. In this way, the early warning strategy can be dynamically adjusted to ensure that timely and accurate risk warnings are provided in any situation, helping the driver or the autonomous driving system make the best decision.

[0078] When the target is not pointing to the area without vision, it means that the target is not moving towards the blind spot of the vehicle, but may be turning to other visible sub-areas. At this time, a target search is performed in the other visible sub-areas pointed by its motion vector. If the target is detected again in the other visible sub-areas, continuous tracking is resumed, and its trajectory and motion vector are updated to repeat the judgment of whether it is pointing to the area without vision. If the target is not detected, regional monitoring and re-identification attempts are performed based on the target's last movement trend until it is confirmed that the target leaves the system's field of view or reappears.

[0079] See also Figure 3 As shown, S6. According to the conflict time window, the blind area is marked and the warning is issued for the vision loss area.

[0080] Specifically, the vision loss area is divided into the following types according to the conflict time window:

[0081] High-risk sub-area: The conflict time window is less than the lower limit of the warning range, and it is determined that there is a high collision risk between the target and the vehicle.

[0082] Medium-risk sub-area: the conflict time window is within the warning range;

[0083] Low-risk sub-area: The conflict time window is greater than the upper limit of the warning range, indicating that there is a low possibility of collision between the target and the vehicle.

[0084] The warning range within the conflict time window is a range used to define different risk levels within the conflict time window. The design of this warning range must comprehensively consider the average reaction time of the driver or autonomous driving system and the braking distance required for the vehicle at its current speed. For example, at highway speeds, due to the longer braking distance, the warning range should be appropriately increased to provide sufficient warning time.

[0085] Based on this consideration, when setting the warning range, the average reaction time of the driver or the automatic driving system and the time required for the vehicle to brake at the current speed are first accumulated as the lower limit of the warning range. This value reflects the shortest time required for the system to detect the potential collision risk and take effective avoidance measures, ensuring that there is enough time to respond and brake in an emergency. The upper limit can be set to a certain proportion larger than the lower limit to provide a certain safety margin. This value provides additional time margin, allowing the system to issue a warning at an earlier time point, thereby improving the timeliness and reliability of the warning.

[0086] An acoustic and visual directional alarm is triggered for high-risk sub-areas, and a directional sound beam is generated through the built-in phased array speaker in the vehicle door. The focus of the sound beam locks the coordinate direction where the target disappears, realizing spatial perception-guided auditory warning. When necessary, it can be combined with visual prompts such as red flashing warnings to enhance the driver's perception of high-risk areas in blind spots.

[0087] The system triggers enhanced display of local images in the dangerous sub-area, automatically retrieves the camera image in the direction where the target disappears, and displays the original view and wide-angle expanded view on the on-board display screen in split screen, providing a wider field of view coverage and spatial perception assistance, helping the driver to further confirm the blind spot status.

[0088] Only continuous background monitoring is carried out on low-risk sub-areas, and no early warning is triggered actively.

[0089] The present invention divides the area of ​​lost vision into high-risk sub-areas, medium-risk sub-areas and low-risk sub-areas by dynamically dividing the risk area based on the conflict time window, and implements differentiated warning strategies according to the risk level of each sub-area, thereby achieving refined hierarchical management of potential collision risks. This method can not only dynamically assess the threat level of the target based on the spatiotemporal relationship between the target and the vehicle, but also adopt corresponding warning forms such as visual prompts and auditory alarms for different risk levels, thereby improving the response accuracy of the warning system and driving safety. Through this mechanism, the system can adaptively identify and gradedly respond to the potential collision risks of blind spot targets in complex traffic environments, ensuring that timely, reliable, and clearly graded risk prompts can be provided to the driver or the automatic driving system in different driving scenarios, assisting them in making scientific and efficient driving decisions.

[0090] Example 2

[0091] See also Figure 4 As shown, the present invention proposes a blind spot warning system for passenger buses, comprising the following modules: a field of view generation module: a wide-angle camera group deployed circumferentially of the vehicle body generates a first panoramic field of view coverage area under straight-ahead conditions, and generates a second panoramic field of view coverage area under turning conditions.

[0092] Field of view loss identification module: connected to the field of view generation module, used to spatially superimpose the first panoramic field of view coverage area with the second panoramic field of view coverage area, and mark the field of view loss area of ​​the second panoramic field of view coverage area relative to the first panoramic field of view coverage area.

[0093] Dynamic target detection module: connected to the field of view loss recognition module, it performs an adjacent area search operation within the second panoramic field of view coverage area based on the spatial position and contour boundary information of the field of view loss area to determine the visible area adjacent to the field of view loss area, and then performs dynamic target detection on the visible area adjacent to the field of view loss area, recording the last spatial coordinates and motion vector of the target at the moment of disappearance.

[0094] Blind spot risk analysis module: connected to the dynamic target detection module, it is used to generate a predicted trajectory line according to the motion vector based on the last spatial coordinate as the starting point, and calculate the conflict time window between the predicted trajectory and the vehicle body trajectory based on the real-time motion trajectory of the bus.

[0095] Blind spot risk warning module: connected to the blind spot risk analysis module, used to mark and warn the blind spot area of ​​the lost vision area according to the conflict time window.

[0096] Example 3

[0097] The present invention proposes a device, which includes a processor, and a memory and a network interface connected to the processor; the network interface is connected to a non-volatile memory in a server; when the processor is running, it calls a computer program from the non-volatile memory through the network interface and runs the computer program through the memory to execute a passenger bus blind spot warning method described in the present invention.

[0098] Example 4

[0099] The present invention provides a storage medium having a computer program burned therein. When the computer program is run in the memory of a server, the method for early warning of a blind spot of a passenger vehicle according to the present invention is implemented.

[0100] The above embodiments may be implemented in whole or in part through software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments may be implemented in whole or in part in the form of a computer program product.

[0101] Those skilled in the art will appreciate that the algorithmic steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0102] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0103] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0104] Finally, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A blind spot warning method for a passenger car, characterized in that: The following steps are involved: S1. Using a wide-angle camera array deployed around the vehicle body, a first panoramic field of view is generated during straight driving, and a second panoramic field of view is generated during turning driving. S2. spatially superimpose the first panoramic view coverage area and the second panoramic view coverage area, and mark the missing area of ​​the second panoramic view coverage area relative to the first panoramic view coverage area; S3. Performing an adjacent area search operation within the second panoramic field of view coverage area based on the spatial position of the missing field of view area and its contour boundary information to determine a visible area adjacent to the missing field of view; S4. Perform dynamic target detection in the visible area adjacent to the missing field of view, and record the last spatial coordinates and motion vector of the target at the time of disappearance; S5. Using the final spatial coordinate as the starting point, generate a predicted trajectory based on the motion vector, and calculate the conflict time window between the predicted trajectory and the vehicle body trajectory based on the real-time motion trajectory of the bus; S6. Mark and warn the blind spots in the areas where vision is lost based on the conflict time window.

2. A passenger vehicle blind spot warning method according to claim 1, characterized in that: The S1 is specifically implemented as follows: When the vehicle is moving straight ahead, the wide-angle cameras deployed around the vehicle body are activated, and each camera collects images synchronously. Construct a rectangular projection coordinate system parallel to the ground with the longitudinal axis of the vehicle body as the axis of symmetry; Perform distortion correction and coordinate transformation on the original wide-angle images output by each camera, and project each image from its original viewing angle into the above-mentioned rectangular projection coordinate system; The corrected images are stitched into a continuous rectangular field of view through image stitching; When the turn signal is activated or the steering wheel angle exceeds the preset threshold, the wide-angle cameras located around the vehicle body are synchronously activated to start real-time image acquisition; Dynamically adjust the projection coordinate system with the longitudinal axis of the vehicle body as the axis of symmetry according to the steering angle of the vehicle's front wheels; The original images output by each wide-angle camera are subjected to distortion correction and coordinate transformation, and then mapped into a projection coordinate system that is dynamically adjusted according to the steering angle of the vehicle's front wheels; The corrected images are stitched into a continuous curved edge field of view through image stitching.

3. The blind spot warning method for a passenger vehicle according to claim 1, characterized in that: The S2 includes the following contents: Extracting closed envelope boundaries of the outer contours of the first panoramic field of view coverage area and the second panoramic field of view coverage area respectively; The closed envelope boundary of the first panoramic field of view coverage area is used as a reference benchmark to geometrically superimpose the closed envelope boundary of the second panoramic field of view coverage area with it, and the first panoramic field of view coverage area that is not covered by the second panoramic field of view coverage area is extracted as the field of view missing area.

4. The blind spot warning method for a passenger vehicle according to claim 1, characterized in that: The S3 includes the following: Marking the outline boundary of the visual field loss area within the second panoramic visual field coverage area based on the spatial position and outline boundary information of the visual field loss area; Based on the contour boundary of the missing field of view area, a boundary adjacency analysis is performed to identify image regions that share a common boundary with the missing field of view area. The boundary features of these image regions are further traced back until continuous image regions without image splicing are identified. These continuous regions that are not disturbed by splicing and share a common boundary with the missing field of view area are defined as the adjacent visible range. The adjacent visible range is divided according to the image acquisition source to which it belongs, and adjacent visible sub-areas are generated.

5. The blind spot warning method for a passenger vehicle according to claim 1, characterized in that: The S4 includes the following: Real-time recognition of moving objects entering the area within the adjacent visible sub-area surrounding the lost field of view. Once a dynamic target is detected, its motion trajectory is continuously tracked. During the tracking process, if a dynamic target disappears from its visible sub-region in the current image frame, a target matching search operation is performed in the spatially adjacent regions of the visible sub-region within a set time window to determine whether the target has moved to an adjacent field of view area. If the target cannot be detected again in the adjacent region, the time when the target disappears is recorded, and the adjacent visible sub-region where it disappeared is located; Go back a set number of historical image frames from the disappearance moment to extract the spatial position information of the target in the consecutive frames; The displacement vector of the center of mass of the target between adjacent frames is calculated based on the position coordinates of the target in consecutive frames.

6. A passenger vehicle blind spot warning method according to claim 5, characterized in that: The S5 specifically includes the following contents: Based on the direction of the motion vector of the dynamic target when it disappears in the adjacent visible sub-region, it is determined whether it points to the missing view area. If it points to the missing view area, linear extrapolation is used to generate a linear prediction trajectory along the direction of its motion vector starting from the last visible position of the target; Determine the real-time bus trajectory based on the vehicle steering wheel angle and vehicle speed; The predicted trajectory of the dynamic target entering the lost field of view area is geometrically compared with the bus's trajectory to determine whether the two trajectories intersect. If so, the shortest spatial distance between the two trajectories is calculated and compared with the set safety threshold. If the shortest spatial distance is less than the safety threshold, the conflict time window is calculated based on the shortest spatial distance and the relative speed between the target and the bus. When there is no intersection, the shortest spatial distance between the two trajectories is also calculated. If the shortest spatial distance is less than the safety threshold, the spatial distance between the two trajectories is continuously updated. If the distance is detected to be continuously shortened in multiple consecutive time steps, the conflict time window calculation is retriggered. When it is not pointing to the area where the field of view is missing, a target search is performed in other visible sub-areas pointed by its motion vector direction. If the target is detected again in other visible sub-areas, continuous tracking is resumed and its trajectory and motion vector are updated. If the target is not detected, regional monitoring and re-identification attempts are performed based on the target's last motion trend until it is confirmed that the target has left the system's field of view or reappeared.

7. The blind spot warning method for a passenger vehicle according to claim 1, characterized in that: The S6 includes the following: The vision loss area is divided into the following areas according to the conflict time window: High-risk sub-area: The conflict time window is less than the lower limit of the warning range; Medium-risk sub-area: the conflict time window is within the warning range; Low-risk sub-area: The conflict time window is greater than the upper limit of the warning range; Triggering an acoustic and visual directional alarm in high-risk sub-areas, a directional sound beam is generated through the built-in phased array speaker in the vehicle door, and the focus of the sound beam locks on the coordinate direction where the target disappears; The system triggers enhanced display of local images in the mid-risk sub-area, automatically retrieves the camera image at the direction where the target disappears, and displays the original view and wide-angle expanded view on the vehicle display screen; Only continuous background monitoring is carried out on low-risk sub-areas, and no early warning is triggered actively.

8. A blind spot warning system for passenger vehicles, characterized in that: Includes the following modules: Field of view generation module: This module uses a wide-angle camera array deployed around the vehicle body to generate a first panoramic field of view coverage area during straight driving conditions and a second panoramic field of view coverage area during cornering conditions. A vision loss identification module: spatially superimposes the first panoramic vision coverage area and the second panoramic vision coverage area, and marks the vision loss area of ​​the second panoramic vision coverage area relative to the first panoramic vision coverage area; Dynamic target detection module: Based on the spatial position and outline boundary information of the missing area, it performs an adjacent area search operation within the second panoramic field of view to determine the visible area adjacent to the missing area. It then performs dynamic target detection in the visible area adjacent to the missing area and records the last spatial coordinates and motion vector of the target at the moment of disappearance. Blind spot risk analysis module: Starting from the last spatial coordinate, it generates a predicted trajectory according to the motion vector and calculates the conflict time window between the predicted trajectory and the vehicle body trajectory based on the real-time motion trajectory of the bus. Blind spot risk warning module: Blind spot marking and warning of areas with lost vision are carried out according to the conflict time window.

9. A device, characterized in that: The device includes a processor, and a memory and a network interface connected to the processor; the network interface is connected to a non-volatile memory in a server; when running, the processor retrieves a computer program from the non-volatile memory through the network interface, and runs the computer program through the memory to execute the method described in any one of claims 1 to 7 above.

10. A storage medium, characterized in that: The storage medium is burned with a computer program, and when the computer program is run in the memory of the server, the method according to any one of claims 1 to 7 is implemented.

Images