Vehicle control methods and related equipment

Abstract

This application discloses a vehicle control method and related equipment, relating to the field of intelligent driving technology. The method includes: acquiring motion state information of a vehicle ahead in the same lane as a target vehicle; identifying abnormal color block regions in the driving lane based on a road surface image of the target vehicle's driving lane; determining the type of anomaly on the road surface ahead of the target vehicle based on the motion state information and the abnormal color block region image; and controlling the driving process of the target vehicle based on the anomaly type. This application achieves early perception and accurate location of road surface anomalies by fusing the motion state information of the preceding vehicle with the abnormal color block information in the road surface image, and accordingly performs targeted control of the vehicle's driving process, thereby improving the vehicle's ability to identify and respond to road surface anomalies.

Description

Technical Field

[0001] This application relates to the field of intelligent driving technology, and more specifically, to a vehicle control method and related equipment. Background Technology

[0002] With the continuous development of intelligent driving and advanced driver assistance technologies, vehicles face higher demands on their perception and decision-making capabilities regarding the driving environment. In particular, the ability to identify and respond to road anomalies under complex road conditions directly impacts driving safety and passenger comfort. Currently, vehicles inevitably encounter various road anomalies such as potholes, bumps, and damage during operation. Therefore, accurately identifying these road anomalies and implementing appropriate control accordingly has become a crucial research direction in vehicle control technology.

[0003] In related technologies, the identification of road surface anomalies typically relies on the vehicle's own sensors to directly detect the road surface, such as by recognizing road features from images or analyzing road conditions based on sensor data. However, these methods struggle to simultaneously consider both the spatial location of the anomaly and its actual impact on vehicle movement. For example, while image recognition can detect changes in road color or texture, it's difficult to determine whether the anomaly will have a substantial impact on the vehicle. Relying solely on changes in vehicle motion lacks the ability to accurately locate the anomaly area. Furthermore, in complex lighting, occlusion, or long-distance scenarios, the accuracy and stability of these technologies remain insufficient, thus affecting the reliability of vehicle control decisions. In other words, these technologies suffer from insufficient accuracy in identifying road surface anomalies and inaccurate control responses. Summary of the Invention

[0004] In the summary section of this application, the relevant technical solutions are described in general terms, and a series of simplified concepts are introduced. These concepts will be further elaborated in the detailed embodiments section. This summary section should not be construed as limiting the key or essential technical features of the claimed solutions, nor is it intended to limit the scope of protection of the claimed solutions.

[0005] The vehicle control method and related equipment provided in this application can achieve early perception and accurate location of abnormalities on the road surface by fusing the motion state information of the preceding vehicle with the abnormal color block information in the road surface image, and can then perform targeted control on the vehicle's driving process accordingly, thereby improving the vehicle's ability to identify and respond to road surface abnormalities.

[0006] In a first aspect, this application provides a vehicle control method applied to a target vehicle, comprising: acquiring motion state information of a vehicle ahead in the same lane as the target vehicle; identifying an abnormal color block region image of the driving lane based on a road surface image of the driving lane of the target vehicle; determining the abnormal type of the road surface ahead of the target vehicle based on the motion state information and the abnormal color block region image; and controlling the driving process of the target vehicle based on the abnormal type.

[0007] In some embodiments, identifying abnormal color block regions in the driving lane based on the road surface image of the target vehicle's driving lane includes: performing lane line detection on the road surface image to obtain a lane region image of the driving lane; performing color histogram statistical analysis on the lane region image to obtain a color segmentation threshold; performing binarization processing on the lane region image based on the color segmentation threshold to obtain a binarized image; and performing morphological processing on the binarized image to obtain the abnormal color block region image.

[0008] In some implementations, the motion state information includes driving trajectory information and vehicle posture information; determining the anomaly type of the road surface ahead of the target vehicle based on the motion state information and the abnormal color block region image includes: determining the trajectory category of the vehicle ahead based on the trajectory information, wherein the trajectory category includes detour category, lane change category, and straight-ahead category; determining the posture category of the vehicle ahead based on the vehicle posture information, wherein the posture category includes bumpy category and smooth category; if the trajectory category is the detour category or the lane change category, and the abnormal color block region image indicates that there is an anomaly on the road surface ahead of the target vehicle, then the anomaly type is determined to be a type to be avoided; if the trajectory category is the straight-ahead category and the posture category is the bumpy category, then the anomaly type of the road surface ahead of the target vehicle is determined based on the abnormal color block region image; if the trajectory category is the straight-ahead category and the posture category is the smooth category, then the anomaly type of the road surface ahead of the target vehicle is determined based on the abnormal color block region image, the vehicle posture information, and the driving scenario of the target vehicle.

[0009] In some implementations, determining the anomaly type of the road surface ahead of the target vehicle based on the abnormal color block region image includes: acquiring the contour information of the abnormal color block in the abnormal color block region image; determining the shape feature score of the abnormal color block based on the contour information; determining the height change value of the road surface anomaly based on the vehicle posture information; determining a first risk probability of the road surface anomaly based on the shape feature score and the height change value; if the first risk probability is greater than or equal to a first probability threshold, then determining the anomaly type as a type to be avoided; if the first risk probability is less than the first probability threshold, then determining the anomaly type as a type to be decelerated and passed.

[0010] In some implementations, determining the shape feature score of the abnormal color block based on the contour information includes: horizontally flipping the abnormal color block based on the center point of the contour information to obtain a horizontally folded color block; determining a horizontal flip overlap rate based on the first overlap area of ​​the abnormal color block and the horizontally folded color block, and the area of ​​the abnormal color block; vertically flipping the abnormal color block based on the center point of the contour information to obtain a vertically folded color block; determining a vertical flip overlap rate based on the second overlap area of ​​the abnormal color block and the vertically folded color block, and the area of ​​the abnormal color block; determining a circular matching degree based on the horizontal flip overlap rate and the vertical flip overlap rate; acquiring multiple boundary points on the contour at preset angular intervals, starting from the center point of the contour information, and determining the boundary point dispersion in each preset angular direction; determining a square matching degree based on the boundary point dispersion in multiple preset angular directions; and determining the shape feature score based on the circular matching degree and the square matching degree.

[0011] In some implementations, determining the anomaly type of the road surface ahead of the target vehicle based on the abnormal color block region image, the vehicle posture information, and the driving scenario of the target vehicle includes: determining a second risk probability based on the abnormal color block region image; determining a third risk probability based on the vehicle posture change information; determining a first weight corresponding to the first risk probability and a second weight corresponding to the second risk probability based on the driving scenario; performing a weighted fusion of the first risk probability and the second risk probability based on the first weight and the second weight to obtain a third risk probability; and determining the anomaly type as a deceleration passage type if the third risk probability is greater than the second probability threshold.

[0012] In some implementations, controlling the driving process of the target vehicle based on the anomaly type includes: if the anomaly type is the type to be avoided, generating a target control strategy to control the target vehicle to detour around the abnormal road surface; if the anomaly type is the type to decelerate through, determining whether the target vehicle has lane-changing or detour conditions; if the lane-changing or detour conditions exist, generating a target control strategy to control the target vehicle to detour around the abnormal road surface; if the lane-changing or detour conditions do not exist, generating a target control strategy to control the target vehicle to decelerate through the abnormal road surface; and controlling the driving process of the target vehicle based on the target control strategy.

[0013] Secondly, this application also provides a vehicle control device applied to a target vehicle, comprising: an information acquisition unit for acquiring motion state information of a vehicle ahead in the same lane as the target vehicle; an image recognition unit for identifying abnormal color block regions in the driving lane based on a road surface image of the target vehicle's driving lane; an anomaly determination unit for determining the anomaly type of the road surface ahead of the target vehicle based on the motion state information and the abnormal color block region image; and a vehicle control unit for controlling the driving process of the target vehicle based on the anomaly type.

[0014] Thirdly, this application also provides a vehicle, including: a memory and a processor, the processor being configured to implement the steps of the vehicle control method described in the first aspect when executing a computer program stored in the memory.

[0015] Fourthly, this application also provides a computer-readable storage medium storing computer-executable instructions or a computer program, which, when executed by a processor, implement the steps of the vehicle control method described in the first aspect.

[0016] Fifthly, this application also provides a computer program product, including a computer program or computer-executable instructions, which, when executed by a processor, implement the steps of the vehicle control method provided in the embodiments of this application.

[0017] In summary, this application obtains the motion state information of a vehicle ahead in the same lane as the target vehicle. Since the target vehicle and the preceding vehicle will pass through the same road surface area sequentially, the motion state of the preceding vehicle can reflect the actual impact of the road surface on vehicle travel. This allows the target vehicle to obtain prior information about the road surface conditions before reaching that location, enabling early perception of road conditions. Based on the road surface image of the target vehicle's lane, abnormal color block regions are identified. Potential abnormal regions are extracted by changes in color or texture in the image, providing spatial location information of road surface anomalies. This allows the target vehicle to visually locate abnormal regions. Combining the motion state information of the preceding vehicle with the abnormal color block region image, the preceding vehicle's motion reflects whether the road surface affects driving, while the color block information represents the specific location of the anomaly. Through the mutual supplementation of these two types of information, the accuracy of determining the type of road surface anomaly can be improved. After determining the type of road surface anomaly, the target vehicle's driving process is controlled based on this anomaly type, enabling the vehicle to adopt corresponding driving strategies for different road surface conditions, thereby achieving effective regulation of the driving process. In summary, the vehicle control method provided in this application achieves early perception and accurate location of road surface anomalies by fusing the motion state information of the preceding vehicle with the abnormal color block information in the road surface image, and accordingly performs targeted control on the vehicle's driving process, thereby improving the vehicle's ability to identify and respond to road surface anomalies. Attached Figure Description

[0018] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit this specification. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A schematic flowchart of a vehicle control method provided in an embodiment of this application; Figure 2 A lane area image provided in an embodiment of this application; Figure 3 A binarized image provided in an embodiment of this application; Figure 4 An image of an abnormal color patch region provided in an embodiment of this application; Figure 5 This is a schematic diagram of the composition structure of a vehicle control device provided in an embodiment of this application; Figure 6 This is a schematic diagram of the composition structure of a vehicle provided in an embodiment of this application. Detailed Implementation

[0019] The terms used in the specification, claims, and drawings of this application, such as "first," "second," "third," "fourth," etc. (if any), are used to distinguish similar objects and not to describe a specific order or sequence. Therefore, it is to be understood that these terms can be used interchangeably where appropriate, allowing the described embodiments to be used in different orders, unless specifically required by the illustrations or description. Furthermore, the terms "is" and "has," and any variations thereof, are intended to cover, non-exclusively, all possible constituent elements. For example, a process, method, system, product, or apparatus comprising several steps or units is not necessarily limited to the steps or units explicitly listed, but may also include other steps or units not explicitly listed, or steps or units inherent to the process, method, product, or apparatus.

[0020] In this application, a "module" or "unit" refers to a computer program or part of a computer program that has a specific function and works in conjunction with other related parts to achieve a predetermined goal. These modules or units can be implemented by software, hardware (e.g., processing circuitry or memory), or a combination of both. One or more processors or memories can implement one or more modules or units. Furthermore, each module or unit can also be part of a larger module or unit.

[0021] The technical solutions of this application will be described in detail below with reference to the accompanying drawings of the embodiments. It should be noted that the described embodiments are only a part of this application, and not all embodiments. In the following description, the "some embodiments" mentioned are only a subset of all possible embodiments, which may be the same or different subsets, and different embodiments can be combined with each other without conflict.

[0022] Figure 1 This is a schematic flowchart illustrating a vehicle control method provided in an embodiment of this application. For example, see [link to example]. Figure 1 The vehicle control method provided in this application embodiment is applied to a target vehicle, which is an intelligent driving vehicle equipped with and executing this vehicle control method, and is the subject of execution and the object of driving control of this method; the vehicle control method may include the following steps 101 to 104: Step 101: Obtain the motion status information of the vehicle ahead that is in the same lane as the target vehicle.

[0023] In some examples, the vehicle ahead is another motor vehicle in the same lane as the target vehicle and located ahead of the target vehicle in its direction of travel. This vehicle provides a reference for road surface anomalies to the target vehicle. Motion state information is information characterizing the dynamic changes of the vehicle ahead during its travel. This information is one of the bases for determining whether there are anomalies on the road surface ahead. Motion state information can be acquired through sensing devices mounted on the target vehicle. An image capture device (ICD) is used to acquire this information; the target vehicle acquires visual images of the vehicle ahead through the image capture device and obtains the motion state information based on the visual image analysis. A radar detection device (RDD) is another acquisition device; the target vehicle emits detection signals through the radar detection device and receives echo signals, obtaining the motion state information of the vehicle ahead based on the echo signals.

[0024] By implementing step 101, the motion state information of the vehicle ahead in the same lane as the target vehicle is obtained. Since the vehicle ahead and the target vehicle will pass through the same road surface area in turn, the motion state of the vehicle ahead can reflect the actual impact of the road surface on the vehicle's movement. This allows the target vehicle to obtain prior information about the road surface conditions before it reaches the road section, thus enabling early perception of the road surface conditions.

[0025] Step 102: Based on the road surface image of the target vehicle's driving lane, identify the abnormal color block area image of the driving lane.

[0026] In some examples, the driving lane is the exclusive road area defined by lane lines where the target vehicle is currently located. This area is the execution range for road surface image acquisition and abnormal color block recognition. The road surface image can be acquired by an image acquisition device mounted on the target vehicle, and this image is the basic data source for identifying abnormal color block regions. For example, the image acquisition device acquires the road surface ahead at a preset frame rate, generating a color digital road surface image that includes road surface texture and lane lines. The abnormal color block region image is extracted from the road surface image and represents a local area image of the road surface within the driving lane that has abnormal color or texture. This image is the key visual basis for determining road surface anomalies. It can be formed by comparing and analyzing the visual features of the road surface image to filter out color block regions that are significantly different from the visual features of normal road surfaces and extract them independently.

[0027] By implementing step 102, abnormal color block regions are identified based on the road surface image of the target vehicle's driving lane. Potential abnormal regions are extracted by utilizing changes in color or texture in the image, which can obtain spatial distribution information of road surface abnormalities, enabling the target vehicle to locate abnormal regions at the visual level.

[0028] Step 103: Based on motion state information and abnormal color block area images, determine the type of abnormality on the road surface in front of the target vehicle.

[0029] In some examples, the anomaly type of the road surface in front of the target vehicle is the classification result of the abnormal traffic conditions existing on the road surface in front of the target vehicle's direction of travel and in the same driving lane as the target vehicle. This classification result is the basis for executing the target vehicle's driving control. The motion state feature data of the vehicle in front can be jointly analyzed with the image features of the abnormal color block area of ​​the driving lane. By using the method of cross-validation of dual information, the road surface anomaly classification result can be determined and output, which can avoid the error caused by single information judgment.

[0030] By implementing step 103, combining the motion status information of the preceding vehicle with the image of the abnormal color block area, on the one hand, the motion status of the preceding vehicle reflects whether the road surface affects the vehicle's driving, and on the other hand, the abnormal color block represents the specific location of the abnormality. This allows the two types of information to complement each other and improves the accuracy of determining the type of abnormality on the road surface ahead.

[0031] Step 104: Control the driving process of the target vehicle based on the anomaly type.

[0032] In some examples, based on the determined type of road surface anomaly ahead of the target vehicle, the vehicle control unit can generate matching driving control commands to adaptively adjust the driving state of the target vehicle; for example, avoidance control can be performed for anomalies with obstacles on the road surface, and deceleration control can be performed for anomalies with risks to traffic on the road surface.

[0033] For example, the vehicle control unit receives real-time data on the type of road anomaly, maintains the ability to respond to control commands instantly, accurately matches the control strategy according to the type of anomaly, ensures that the control operation is consistent with the actual road conditions, and sends the generated driving control commands to the execution components of the target vehicle to achieve stable and precise control of the driving process.

[0034] By implementing step 104, after determining the type of road surface anomaly, the driving process of the target vehicle is controlled based on the type of anomaly, so that the vehicle can take corresponding driving strategies for different road surface conditions, thereby achieving effective regulation of the driving process.

[0035] In summary, this embodiment of the application acquires the motion state information of a vehicle ahead in the same lane as the target vehicle. Since the preceding vehicle and the target vehicle will pass through the same road surface area sequentially, the motion state of the preceding vehicle can reflect the actual impact of the road surface on vehicle travel. This allows the target vehicle to obtain prior information about the road surface conditions before reaching that location, enabling early perception of road conditions. Based on the road surface image of the target vehicle's lane, abnormal color block regions are identified. Potential abnormal regions are extracted by changes in color or texture in the image, which can obtain the spatial location information of road surface anomalies, allowing the target vehicle to locate the abnormal regions visually. Combining the motion state information of the preceding vehicle with the abnormal color block region image, on the one hand, the motion of the preceding vehicle reflects whether the road surface affects driving, and on the other hand, the color block information represents the specific location of the anomaly. Through the mutual supplementation of these two types of information, the accuracy of determining the type of road surface anomaly ahead can be improved. After determining the type of road surface anomaly, the driving process of the target vehicle is controlled based on the type of anomaly, enabling the vehicle to adopt corresponding driving strategies for different road surface conditions, thereby achieving effective adjustment of the driving process. In summary, the vehicle control method provided in this application integrates the motion state information of the vehicle in front with the abnormal color block information in the road image to achieve early perception and accurate location of abnormalities on the road ahead, and accordingly performs targeted control on the vehicle's driving process, thereby improving the vehicle's ability to identify and respond to abnormalities on the road.

[0036] In some embodiments, step 102 may include: performing lane line detection on the road surface image to obtain a lane region image of the driving lane; performing color histogram statistical analysis on the lane region image to obtain a color segmentation threshold; performing binarization processing on the lane region image based on the color segmentation threshold to obtain a binarized image; and performing morphological processing on the binarized image to obtain an abnormal color block region image.

[0037] In some examples, lane detection is a visual processing operation that extracts features and spatially locates lane boundary markers in a road image. This operation can be performed by the onboard image processing unit (IPU) of the target vehicle, detecting lanes by recognizing their color and contour features in the road image; for example, locating lane lines in the form of solid white lines and dashed yellow lines from the road image. The lane region image is the image data extracted from the original road image after lane detection, containing only the area of ​​the target vehicle's driving lane. This image is obtained by cropping the road image region using the boundaries defined by the lane lines. An example is shown below. Figure 2 The image shows the lane area.

[0038] Color histogram statistical analysis is an image processing method that quantifies and statistically analyzes the color distribution of all pixels within a lane area image. It generates color distribution results by counting the number of pixels corresponding to each color value. For example, counting the percentage of pixels representing the basic road surface color within a lane area image is a specific implementation of color histogram statistical analysis. The color segmentation threshold, generated after color histogram statistical analysis, is a critical value used to distinguish normal road surface colors from abnormal colors. This value is adaptively determined based on the color distribution statistical results of the lane area image. A binarized image is an image containing only two pixel values ​​obtained by performing binarization (BP) on the lane area image based on the color segmentation threshold. During processing, image pixels are divided into two categories of feature values ​​according to the threshold. For example, converting a lane area image into an image containing only white and black pixels is a binarized image. See [link to documentation]. Figure 3 The image content shown.

[0039] Morphological processing (MP) is an image processing operation that optimizes pixel regions in a binarized image. This operation can include image dilation and image erosion to eliminate minor disturbances and smooth the edges of abnormal regions, ultimately extracting color-block regions representing road surface anomalies. For example, dilation connects scattered pixels, and erosion removes isolated noise points. (See also...) Figure 4 The image content shown.

[0040] Through the implementation of the above embodiments, lane line detection is first performed on the road surface image to limit the analysis area. Then, an adaptive segmentation threshold is determined based on the color histogram. Combined with binarization and morphological processing, abnormal color block regions are extracted. This makes the extraction of abnormal regions more focused on the current driving lane and has stronger robustness. It can reduce interference from irrelevant areas, improve the accuracy and stability of abnormal color block recognition, and further provide high-quality input data for subsequent road surface anomaly determination.

[0041] In some embodiments, the aforementioned motion state information may include driving trajectory information and vehicle posture information; the aforementioned step 103 may include: determining the trajectory category of the vehicle ahead based on the trajectory information, wherein the trajectory category may include detour category, lane change category, and straight-going category; determining the posture category of the vehicle ahead based on the vehicle posture information, wherein the posture category may include bumpy category and smooth category; if the trajectory category is detour category or lane change category, and the abnormal color block area image indicates that there is an anomaly on the road surface ahead of the target vehicle, then the anomaly type is determined to be a type to be avoided; if the trajectory category is straight-going category and the posture category is bumpy category, then the anomaly type of the road surface ahead of the target vehicle is determined based on the abnormal color block area image; if the trajectory category is straight-going category and the posture category is smooth category, then the anomaly type of the road surface ahead of the target vehicle is determined based on the abnormal color block area image, vehicle posture information, and the driving scenario of the target vehicle.

[0042] In some examples, driving trajectory information is path data representing the continuous changes in the position of the vehicle ahead during its journey. This can be acquired jointly by the image acquisition device and radar detection device on the target vehicle, obtained by continuously tracking the spatial position of the vehicle ahead and recording its temporal position changes. For example, lateral offset data of the vehicle ahead within the lane and longitudinal driving path data along the lane are both driving trajectory information. Vehicle posture information is characteristic data reflecting changes in the vertical and horizontal posture of the vehicle ahead. This can be obtained by acquiring the visual features of the vehicle ahead through the target vehicle's image acquisition device and analyzing changes in vehicle height and tilt. For example, data on the vehicle ahead's body tilting down on one side or rising as a whole are both vehicle posture information.

[0043] The process of determining the trajectory category of a vehicle ahead based on trajectory information involves the information processing unit parsing and calculating the driving trajectory information. Based on the lateral displacement amplitude, displacement duration, and relative position of the vehicle ahead with the lane lines, a preset judgment rule is matched to determine the trajectory category. For example, core displacement features are extracted from the driving trajectory information, and the feature values ​​are compared with preset trajectory classification thresholds to ultimately complete the trajectory category division. The detour category is when the vehicle ahead makes a small lateral displacement within the current driving lane without crossing any lane lines, in order to avoid a road surface area. For example, a vehicle ahead slightly shifts left or right within its lane, always remaining within the area enclosed by the lane lines, which falls under the detour category. The lane change category is when the vehicle ahead actively crosses lane lines to enter an adjacent lane. For example, a vehicle ahead crosses a lane line and completely enters the adjacent lane on the left or right, which falls under the lane change category. The straight-ahead category is when the vehicle ahead travels steadily along the center line of the driving lane without significant lateral displacement, maintaining a straight line. For example, a vehicle ahead does not shift left or right and always travels in a straight line along the lane, which falls under the straight-ahead category.

[0044] The process of determining the attitude category of a vehicle ahead based on its vehicle posture information involves the information processing unit quantitatively analyzing the vehicle posture information. Based on parameters such as the amplitude of vehicle body undulations and tilt angles, the unit compares these parameters with preset attitude judgment thresholds to classify the attitude category. This process can analyze the fluctuation values ​​in the vehicle posture information to determine whether the attitude change exceeds a preset stable range, thus completing the attitude category determination. The bumpy category is characterized by significant vertical undulations, unilateral or bilateral sinking and lifting, and attitude changes exceeding the stable threshold. For example, a vehicle ahead experiencing unilateral sinking after passing a pothole or bilateral lifting after passing a bump qualifies as a bumpy vehicle. The stable category is characterized by no significant vertical undulations or tilt changes in the vehicle ahead, with attitude fluctuations within the preset stable threshold range. For example, a vehicle ahead maintaining a level and stable state without fluctuations while driving on a normal road surface qualifies as a stable vehicle.

[0045] If the trajectory category is detour or lane change, and the abnormal color block area image indicates that there is an anomaly on the road ahead of the target vehicle, then the anomaly type is determined to be the type to be avoided. This is a dual-condition cross-judgment logic executed by the information processing unit. In this logic, the vehicle ahead detouring or changing lanes represents active obstacle avoidance behavior, and the abnormal color block represents a visual anomaly on the road. When both types of information are satisfied at the same time, the information processing unit directly determines that the road anomaly is of a high-risk level and an avoidance operation needs to be performed.

[0046] If the trajectory category is straight and the attitude category is bumpy, the abnormality type of the road surface in front of the target vehicle is determined based on the abnormal color block area image. This is a special judgment logic set by the information processing unit for the straight bumpy scenario. In this logic, if the vehicle in front maintains straight driving without obstacle avoidance behavior but the vehicle body is bumpy, it indicates that there is a hidden abnormality on the road surface. The information processing unit uses the abnormal color block area image as a basis to analyze the color block features to complete the abnormality type determination.

[0047] If the trajectory category is straight-going and the attitude category is stable, the abnormal type of the road surface in front of the target vehicle is determined based on the abnormal color block area image, vehicle attitude information and the driving scenario of the target vehicle. This is a comprehensive judgment logic set by the information processing unit for the straight-going stable scenario. In this logic, the vehicle in front has no obstacle avoidance behavior and no attitude fluctuation. The information processing unit integrates the road surface visual features, vehicle attitude data and the current driving environment to comprehensively determine whether there is an abnormality on the road surface.

[0048] Through the implementation of the above embodiments, the motion state of the vehicle in front is refined into trajectory category and vehicle posture category, and multi-condition judgment logic is constructed by combining abnormal color block information to realize the identification of different road surface anomalies in different scenarios. In particular, by distinguishing between detour, lane change and straight driving behaviors, as well as bumpy and smooth states, it is possible to more accurately distinguish vehicle behavior caused by road surface anomalies from other driving behaviors, reduce the probability of misjudgment, improve the accuracy of anomaly type identification, and enhance the pertinence of control decisions.

[0049] In some embodiments, the aforementioned determination of the anomaly type of the road surface in front of the target vehicle based on the abnormal color block region image may include: acquiring the contour information of the abnormal color block in the abnormal color block region image; determining the shape feature score of the abnormal color block based on the contour information; determining the height change value of the road surface anomaly based on the vehicle posture information; determining the first risk probability of the road surface anomaly based on the shape feature score and the height change value; if the first risk probability is greater than or equal to a first probability threshold, then determining the anomaly type as a type to be avoided; if the first risk probability is less than the first probability threshold, then determining the anomaly type as a type to be decelerated and passed.

[0050] In some examples, anomalous color blocks are areas within the driving lane that exhibit significantly different color and texture features from the regular road surface. These areas are a direct visual indication of road anomalies, and anomalous color blocks can be directly extracted from the image of these areas. For example, dark, irregular color blocks corresponding to road potholes and light-colored, blocky areas corresponding to road bumps are both anomalous color blocks. Contour information is a set of coordinate data describing the shape and spatial extent of the outer boundary of an anomalous color block. This information is the fundamental data for analyzing the shape of anomalous color blocks. Contour information can be obtained by performing edge extraction operations on the anomalous color blocks using an Image Analysis Unit (IAU). For example, the coordinate sequence of the closed boundary of an anomalous color block and the boundary enclosure range data are both contour information. Shape feature scores are numerical indicators used to quantify the regularity of anomalous color blocks, calculated based on the contour information through geometric analysis. They can be generated based on the geometric features of the contour information; for example, regular circular color blocks correspond to higher scores, while irregular pothole color blocks correspond to lower scores.

[0051] The height variation value is a characteristic numerical value representing the vertical undulation amplitude of road surface anomalies. This value reflects the depth of potholes or the height of protrusions and can be calculated based on the vehicle's posture information. For example, the negative undulation value corresponding to a pothole and the positive undulation value corresponding to a protrusion are both height variation values. The first risk probability is generated by fusing the shape feature score and the height variation value. It is used to quantify the probability level of road surface anomalies and can be calculated by fusing the shape feature score and the height variation value. For example, a high-risk road surface anomaly corresponds to a first risk probability close to 1, and a low-risk road surface anomaly corresponds to a smaller first risk probability. The first probability threshold is a pre-set critical probability value used to classify the risk level of road surface anomalies. This value is the criterion for distinguishing anomaly types. For example, the preset first probability threshold is 0.7. A first risk probability of 0.8 or higher than 0.7 is determined to be a type to be avoided, and a first risk probability of 0.5 or lower than 0.7 is determined to be a type to be slowed down before proceeding.

[0052] Through the implementation of the above embodiments, the contour information of abnormal color blocks is extracted and the shape feature score is calculated. At the same time, the road height change value is obtained by combining the vehicle posture information. The risk probability is further determined based on the two, realizing the quantitative assessment of road anomalies from two dimensions: geometric features and dynamic response. It can closely follow the visual shape and physical undulation features of road potholes and bumps, and can effectively distinguish regular interference objects such as manhole covers from real road anomalies. The height change value obtained by relying on the vehicle posture information can accurately reflect the actual risk magnitude of road anomalies. The fusion of the two to obtain the first risk probability realizes the quantitative assessment of road anomaly risk. Then, the first probability threshold is used to clearly divide the types of road anomalies to be avoided and those to be decelerated through. This not only fits the physical characteristics of road anomalies and the actual needs of intelligent driving hierarchical control, but also has a rigorous judgment logic without redundant links. It can also avoid the misjudgment problem of single feature judgment, greatly improve the accuracy of road anomaly type identification, and provide clear execution basis for vehicle control output. It can effectively solve the technical problems of inaccurate road anomaly identification and inaccurate control response in intelligent driving.

[0053] In some embodiments, the aforementioned determination of the shape feature score of the abnormal color block based on contour information may include: horizontally flipping the abnormal color block based on the center point of the contour information to obtain a horizontally folded color block; determining the horizontal flip overlap rate based on the first overlap area of ​​the abnormal color block and the horizontally folded color block, and the area of ​​the abnormal color block; vertically flipping the abnormal color block based on the center point of the contour information to obtain a vertically folded color block; determining the vertical flip overlap rate based on the second overlap area of ​​the abnormal color block and the vertically folded color block, and the area of ​​the abnormal color block; determining the circular matching degree based on the horizontal flip overlap rate and the vertical flip overlap rate; acquiring multiple boundary points on the contour at preset angular intervals, starting from the center point of the contour information, and determining the boundary point dispersion in each preset angular direction; determining the square matching degree based on the boundary point dispersion in multiple preset angular directions; and determining the shape feature score based on the circular matching degree and the square matching degree.

[0054] In some examples, the center point of the contour information is the geometric center coordinate of the abnormal color block contour. This coordinate serves as the reference position for the flip operation and boundary point extraction. The center point can be determined by extracting the extreme coordinates of the contour and calculating their average. For example, the intersection coordinates of the midpoints of the upper and lower boundaries and the midpoints of the left and right boundaries of the contour are the center point of the contour information. A horizontal half-flipped color block is a color block image generated by mirroring the abnormal color block along the horizontal direction with the center point of the contour information as the reference. This can be obtained by performing a horizontal mirror transformation using an image analysis unit. The first overlap area between the abnormal color block and the horizontal half-flipped color block is the pixel area of ​​the overlapping region of the original abnormal color block and the horizontal half-flipped color block in the same coordinate system. This can be obtained by performing a region superposition operation on the two types of color blocks. The horizontal flip overlap rate is the ratio of the first overlap area to the area of ​​the abnormal color block itself, used to characterize the degree of symmetry of the abnormal color block in the horizontal direction. It can be calculated by division; for example, the ratio of the first overlap area to the area of ​​the abnormal color block is 0.85.

[0055] A vertical half-flip color block is a color block image generated by mirroring an abnormal color block along the vertical direction with its center point as the reference. This can be obtained by performing a vertical mirror transformation using an image analysis unit. The second overlap area between the abnormal color block and the vertical half-flip color block is the pixel area of ​​the overlapping region of the original abnormal color block and the vertical half-flip color block in the same coordinate system. This can be obtained by performing region superposition operations on the two types of color blocks. The vertical flip overlap rate is the ratio of the second overlap area to the area of ​​the abnormal color block itself, used to characterize the degree of symmetry of the abnormal color block in the vertical direction. It can be calculated by division; for example, the ratio of the second overlap area to the area of ​​the abnormal color block is 0.88.

[0056] The circular matching degree is a characteristic value that quantifies the degree to which abnormal color blocks fit the standard circular shape. The higher the value, the closer the shape of the color block is to a circle. The circular matching degree can be obtained by averaging or weighted summing the two types of overlap rates. For example, the circular matching degree corresponding to the circular color blocks of road manhole covers can reach more than 0.9.

[0057] Starting from the center point of the contour information, multiple boundary points on the contour are obtained according to a preset angle interval, and the dispersion of the boundary points in each preset angle direction is determined by starting from the center point and connecting radial lines at a fixed angle, obtaining the intersection of the connecting lines and the contour as the boundary points, calculating the dispersion of the boundary point coordinates, extracting the boundary points at ten-degree intervals and calculating the standard deviation.

[0058] Square matching degree is a characteristic value that quantifies the degree to which abnormal color blocks fit the standard square shape. The higher the value, the closer the shape of the color block is to a square. It can be obtained based on boundary point dispersion analysis. Square color blocks with regular edges have smaller boundary point dispersion and correspondingly higher square matching degree. Multi-directional dispersion features can be statistically analyzed and converted into square matching degree. Alternatively, after removing the extreme values ​​of dispersion, the mean can be calculated and then converted into a square matching degree value. The smaller the mean of dispersion, the higher the square matching degree.

[0059] The process of determining the square matching degree based on the boundary point dispersion in multiple preset angular directions can be achieved by first starting from the center point of the contour information and traversing all directions from 0° to 340° at preset angular intervals of 20 degrees, resulting in a total of eighteen preset angular directions. For each preset angular direction, lines are taken from the center point at intervals of 0.2 degrees to both sides, and the intersection points of the lines and the contour are obtained as boundary points. After obtaining twenty boundary points, a straight line is fitted using the least squares method. The ratio of the standard deviation to the mean of the distance from all boundary points to the fitted straight line is calculated to obtain the boundary point dispersion in that direction. This process is repeated for all eighteen directions. Then, the four highest and four lowest dispersion values ​​are removed, and the average of the remaining ten dispersion values ​​is calculated. Finally, through a preset numerical mapping rule, the average dispersion value is converted into a square matching degree within the range of 0 to 1. The smaller the average dispersion value, the higher the square matching degree. For example, for abnormal color blocks in regular square repair areas of the road surface, the average dispersion value in the eighteen directions after removing extreme values ​​is 0.04, and a square matching degree of 0.96 is obtained after mapping transformation; for abnormal color blocks in irregular potholes of the road surface, the average dispersion value in the eighteen directions after removing extreme values ​​is 0.62, and a square matching degree of 0.38 is obtained after mapping transformation.

[0060] The process of determining shape feature scores based on circular and square matching degrees can be achieved by first obtaining pre-calculated circular and square matching degrees, while simultaneously retrieving a pre-stored circular judgment threshold. The circular matching degree is then compared to the circular judgment threshold. If the circular matching degree is greater than or equal to the threshold, it indicates a high degree of fit between the color block shape and the circle, and the shape feature score is obtained by subtracting the circular matching degree from 1. If the circular matching degree is less than the threshold, it indicates a low degree of fit between the color block shape and the circle, and the shape feature score is obtained by subtracting the square matching degree from 1. The final output shape feature score falls within the range of 0 to 1. For example, for an abnormal color block resembling a circular manhole cover on a road surface, the circular matching degree is 0.94, which is greater than the preset circular judgment threshold of 0.7, resulting in a final shape feature score of 0.06. For an abnormal color block resembling an irregular pothole on a road surface, the circular matching degree is 0.32, which is less than the preset circular judgment threshold of 0.7, while the corresponding square matching degree is 0.35, resulting in a final shape feature score of 0.65.

[0061] By implementing the above embodiments, the abnormal color blocks are horizontally and vertically flipped to calculate symmetry, and combined with multi-angle boundary point dispersion analysis, a quantitative description of the shape characteristics of the abnormal color blocks is achieved, thereby distinguishing abnormal areas with regular and irregular shapes. This can effectively reduce misjudgments caused by regular objects (such as manhole covers), improve the accuracy of anomaly identification, and provide a more reliable basis for risk assessment based on shape characteristics.

[0062] In some embodiments, the aforementioned determination of the anomaly type of the road surface in front of the target vehicle based on the abnormal color block region image, vehicle posture information, and the driving scenario of the target vehicle may include: determining a second risk probability based on the abnormal color block region image; determining a third risk probability based on the vehicle posture change information; determining a first weight corresponding to the first risk probability and a second weight corresponding to the second risk probability based on the driving scenario; weighting and fusing the first risk probability and the second risk probability based on the first weight and the second weight to obtain the third risk probability; and determining the anomaly type as a deceleration passage type if the third risk probability is greater than the second probability threshold.

[0063] In some examples, the second risk probability is calculated solely based on the visual features of the anomalous color patch region image. It is a quantified probability value used to individually characterize the visual risk level of road surface anomalies. This value ranges from zero to one and can be obtained by analyzing the region extent, feature contrast, and shape feature score of the anomalous color patch. The larger the coverage area of ​​the anomalous color patch and the more significant the feature differences, the higher the corresponding second risk probability value. The process of determining the second risk probability based on the anomalous color patch region image involves first extracting the pixel proportion and edge sharpness features of the anomalous color patch, and then combining this with the shape feature score and substituting it into a preset calculation formula to obtain the second risk probability.

[0064] The third risk probability is calculated solely based on the vehicle's posture changes. It's a quantitative probability value used to characterize the degree of risk of vehicle bumps caused by road anomalies. This value also ranges from zero to one and can be obtained by analyzing the vertical undulations and unilateral tilt of the vehicle's body. The more severe the body posture fluctuations, the higher the corresponding third risk probability value. The process of determining the third risk probability based on vehicle posture change information involves first extracting the amplitude and frequency parameters of the body posture fluctuations, then performing numerical conversion according to preset risk assessment standards, and finally calculating the third risk probability based on the posture characteristics of the slightly bumpy vehicle ahead.

[0065] The first weight is the scenario adaptation weighting coefficient assigned to the first risk probability in the weighted fusion calculation. It is used to adjust the influence ratio of the first risk probability in the comprehensive risk assessment. It can be dynamically determined based on the current driving scenario of the target vehicle, and the value range is between 0 and 1. The second weight is the scenario adaptation weighting coefficient assigned to the second risk probability in the weighted fusion calculation. It is used to adjust the influence ratio of the second risk probability in the comprehensive risk assessment. It can also be dynamically determined based on the current driving scenario of the target vehicle, and the value range is also between 0 and 1.

[0066] The process of determining the first weight corresponding to the first risk probability and the second weight corresponding to the second risk probability based on the driving scenario can be achieved by first collecting information on ambient light, road surface condition, driving speed, and road type through sensing devices to identify the current driving scenario. Next, pre-stored weight configuration tables for various scenarios, such as strong light / backlight, rain / water accumulation, low-speed urban driving, high-speed driving, normal sunny weather, and road construction / debris, are invoked. Finally, the scenario-specific weight values ​​are matched between the first and second risk probabilities. For example, in a strong light / backlight scenario, the accuracy of color block recognition decreases, so a first weight of 0.7 and a second weight of 0.3 are matched. In a normal sunny weather scenario, both types of perception are accurate, so a first weight of 0.5 and a second weight of 0.5 are matched.

[0067] The third risk probability is a comprehensive road anomaly risk probability generated by integrating the first risk probability, the second risk probability, the first weight, and the second weight. This probability comprehensively combines four types of information: shape features, posture changes, visual color blocks, and driving scenarios. It can truly reflect the actual traffic risk of road anomalies in the current environment, and its value ranges from 0 to 1. It can be calculated using a weighted summation formula, which is: the third risk probability equals the first risk probability multiplied by the first weight plus the second risk probability multiplied by the second weight. For example, when the first risk probability is 0.6, the second risk probability is 0.6, and both weights are 0.5, the calculated third risk probability is 0.6.

[0068] The second probability threshold is a value used to classify the overall risk probability level. It is the only criterion for determining whether a road surface anomaly requires deceleration. The value ranges from 0 to 1. The third risk probability, which is calculated by weighted fusion, can be precisely compared with the second probability threshold. When the third risk probability is greater than the second probability threshold, it is determined that there is a certain risk to passing through the road, but no detour is required. Only a reduction in vehicle speed is needed to pass safely. Thus, the anomaly type is determined to be the deceleration passage type.

[0069] By implementing the above embodiments, different risk probabilities are calculated based on abnormal color block information and vehicle posture information, and each risk is weighted and fused in combination with the driving scenario. This allows for dynamic adjustment of the importance of each information source according to different environmental conditions, thereby maintaining high recognition accuracy in complex scenarios, improving the stability and reliability of the vehicle under various working conditions, and further enhancing the accuracy of control response.

[0070] In some embodiments, step 104 may include: if the anomaly type is an avoidance type, generating a target control strategy to control the target vehicle to bypass the abnormal road surface; if the anomaly type is a deceleration passage type, determining whether the target vehicle has lane-changing or detour conditions; if lane-changing or detour conditions exist, generating a target control strategy to control the target vehicle to bypass the abnormal road surface; if lane-changing or detour conditions do not exist, generating a target control strategy to control the target vehicle to decelerate through the abnormal road surface; and controlling the driving process of the target vehicle based on the target control strategy.

[0071] In some examples, the type to be avoided represents an abnormal situation on the road ahead, such as a deep pothole or a high bump, which makes it unsafe to pass. The vehicle control unit receives the judgment result of the road abnormality type in real time. After confirming that the abnormality type is the type to be avoided, it can immediately call the system's preset detour control rule library, and combine the target vehicle's current speed and distance parameters from the abnormal road surface to generate a complete target control strategy that includes steering angle adjustment, driving path planning, and speed assistance control. For example, when there is a large deep pothole on the road ahead of the target vehicle and the abnormality type is determined to be the type to be avoided, the vehicle control unit generates a detour target control strategy that deviates slightly laterally within the current lane.

[0072] The "Decelerate Through" type indicates a minor road anomaly ahead, allowing for safe passage by slowing down. The sensing unit continuously collects information on obstacles in adjacent lanes, lane markings, and available lateral space within the lane, transmitting this environmental data to the vehicle control unit. Lane change conditions include no obstacles in adjacent lanes, sufficient longitudinal distance for a safe lane change, and no signs prohibiting lane changes. Detour conditions indicate sufficient lateral space within the current lane, allowing for slight steering to avoid the abnormal road surface. The target vehicle can determine the conditions based on the sensing data. If the lane change or detour conditions are met, a detour control rule is used to generate a target control strategy. If the conditions are not met, the system's preset deceleration control rules are invoked to generate a deceleration through target control strategy that includes braking intensity and target vehicle speed. When the anomaly type is "Decelerate Through" and the sensing unit determines that adjacent lanes are safe, the target vehicle can generate a lane change / detour strategy. If the target vehicle has no available lane change or detour space, a deceleration through strategy can be generated.

[0073] Based on the target control strategy, the process of controlling the driving process of the target vehicle involves the vehicle control unit parsing the generated target control strategy into standardized electronic control commands, which are then transmitted to the execution unit via the vehicle communication bus. The execution unit includes steering actuators, power control components, and braking control components, and can directly respond to electronic control commands to complete physical operations. The execution unit can drive the steering actuators to adjust the wheel angles according to the detour strategy to achieve a deviation from the driving path, and drive the power control components to reduce the output power and the braking control components to apply appropriate braking force according to the deceleration strategy to achieve a reduction in vehicle speed. For example, when the target control strategy is to detour within the lane, the execution unit controls the steering components to complete a small turn of less than 30 degrees; when the target control strategy is to decelerate through, the execution unit controls the vehicle to reduce its speed from 60 kilometers per hour to 30 kilometers per hour.

[0074] Through the implementation of the above embodiments, based on the determined anomaly type and current driving conditions (such as whether lane changing or detour conditions are available), a corresponding target control strategy is generated and the vehicle driving process is controlled, enabling the vehicle to take the optimal response under different constraints. This not only improves the flexibility of control decisions, but also takes into account driving efficiency and comfort while ensuring safety, thereby achieving more precise and reasonable vehicle control.

[0075] Furthermore, as an implementation of the foregoing method embodiments, this application also provides a vehicle control device for implementing the foregoing method embodiments. This device embodiment corresponds to the foregoing method embodiments. For ease of reading, this vehicle control device embodiment will not repeat the details of the foregoing method embodiments one by one, but it should be understood that the device in this application embodiment can correspondingly implement all the contents of the foregoing method embodiments. For example... Figure 5As shown, the vehicle control device 20 includes: an information acquisition unit 201, an image recognition unit 202, an anomaly determination unit 203, and a vehicle control unit 204. The information acquisition unit 201 is used to acquire motion state information of vehicles ahead in the same lane as the target vehicle; the image recognition unit 202 is used to identify abnormal color block areas in the driving lane based on a road surface image of the target vehicle's driving lane; the anomaly determination unit 203 is used to determine the anomaly type of the road surface ahead of the target vehicle based on the motion state information and the abnormal color block area image; and the vehicle control unit 204 is used to control the driving process of the target vehicle based on the anomaly type.

[0076] In some embodiments, the image recognition unit 202 is further configured to perform lane line detection on the road surface image to obtain a lane area image of the driving lane; perform color histogram statistical analysis on the lane area image to obtain a color segmentation threshold; perform binarization processing on the lane area image based on the color segmentation threshold to obtain a binarized image; and perform morphological processing on the binarized image to obtain an abnormal color block area image.

[0077] In some embodiments, the motion state information includes driving trajectory information and vehicle posture information; the anomaly determination unit 203 is further configured to determine the trajectory category of the vehicle ahead based on the trajectory information, wherein the trajectory category includes detour category, lane change category, and straight-going category; determine the posture category of the vehicle ahead based on the vehicle posture information, wherein the posture category includes bumpy category and smooth category; if the trajectory category is detour category or lane change category, and the abnormal color block area image indicates that there is an anomaly on the road surface ahead of the target vehicle, then the anomaly type is determined to be a type to be avoided; if the trajectory category is straight-going category and the posture category is bumpy category, then the anomaly type of the road surface ahead of the target vehicle is determined based on the abnormal color block area image; if the trajectory category is straight-going category and the posture category is smooth category, then the anomaly type of the road surface ahead of the target vehicle is determined based on the abnormal color block area image, vehicle posture information, and the driving scenario of the target vehicle.

[0078] In some embodiments, the anomaly determination unit 203 is further configured to acquire the contour information of the abnormal color block in the abnormal color block region image; determine the shape feature score of the abnormal color block based on the contour information; determine the height change value of the road surface anomaly based on the vehicle body posture information; determine the first risk probability of the road surface anomaly based on the shape feature score and the height change value; if the first risk probability is greater than or equal to the first probability threshold, then determine the anomaly type as a type to be avoided; if the first risk probability is less than the first probability threshold, then determine the anomaly type as a type to be decelerated and passed.

[0079] In some embodiments, the anomaly determination unit 203 is further configured to: horizontally flip the abnormal color block based on the center point of the contour information to obtain a horizontally half-folded color block; determine the horizontal flip overlap rate based on the first overlap area of ​​the abnormal color block and the horizontally half-folded color block, and the area of ​​the abnormal color block; vertically flip the abnormal color block based on the center point of the contour information to obtain a vertically half-folded color block; determine the vertical flip overlap rate based on the second overlap area of ​​the abnormal color block and the vertically half-folded color block, and the area of ​​the abnormal color block; determine the circular matching degree based on the horizontal flip overlap rate and the vertical flip overlap rate; obtain multiple boundary points on the contour at preset angular intervals, starting from the center point of the contour information, and determine the boundary point dispersion in each preset angular direction; determine the square matching degree based on the boundary point dispersion in multiple preset angular directions; and determine the shape feature score based on the circular matching degree and the square matching degree.

[0080] In some embodiments, the anomaly determination unit 203 is further configured to determine a second risk probability based on the image of the abnormal color block region; determine a third risk probability based on the vehicle body posture change information; determine a first weight corresponding to the first risk probability and a second weight corresponding to the second risk probability based on the driving scenario; perform weighted fusion of the first risk probability and the second risk probability based on the first weight and the second weight to obtain a third risk probability; and determine the anomaly type as a deceleration passage type if the third risk probability is greater than the second probability threshold.

[0081] In some embodiments, the vehicle control unit 204 is further configured to: generate a target control strategy to control the target vehicle to bypass the abnormal road surface if the abnormality type is an avoidance type; determine whether the target vehicle has lane-changing or detour conditions if the abnormality type is a deceleration passage type; generate a target control strategy to control the target vehicle to bypass the abnormal road surface if lane-changing or detour conditions exist; generate a target control strategy to control the target vehicle to decelerate through the abnormal road surface if lane-changing or detour conditions do not exist; and control the driving process of the target vehicle based on the target control strategy.

[0082] This application also provides a computer-readable storage medium storing computer-executable instructions or a computer program that, when executed by a processor, will cause the processor to perform any step of the vehicle control method provided in this application.

[0083] In some embodiments, the computer-readable storage medium may be a random access memory (RAM), a read-only memory (ROM), flash memory, a magnetic surface memory, an optical disc, or a compact disc read-only memory (CD-ROM); or it may be a variety of devices that include one or any combination of the above-mentioned memories.

[0084] In some embodiments, computer-executable instructions may take the form of programs, software, software modules, scripts, or code, written in any form of programming language (including compiled or interpreted languages, or declarative or procedural languages), and may be deployed in any form, including as stand-alone programs or as modules, components, subroutines, or other units suitable for use in a computing environment.

[0085] In some embodiments, computer-executable instructions may, but do not necessarily, correspond to files in a file system, and may be stored as part of a file that holds other programs or data, for example, in one or more scripts in a HyperText Markup Language (HTML) document, in a single file dedicated to the program in question, or in multiple co-located files (e.g., files that store one or more modules, subroutines, or code sections).

[0086] like Figure 6 As shown, this application also provides a vehicle 30, including a memory 310, a processor 320, and a computer program 311 stored in the memory 310 and executable on the processor. When the processor 320 executes the computer program 311, it implements any step of the above-described vehicle control method.

[0087] This application also provides a computer program product comprising a computer program or computer-executable instructions stored in a computer-readable storage medium. A vehicle's processor reads the computer program or computer-executable instructions from the computer-readable storage medium and executes the computer program or computer-executable instructions, causing the vehicle to perform any step of the vehicle control method described above.

[0088] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A vehicle control method characterized by, Applied to the target vehicle, the vehicle control method includes: Obtain motion status information of vehicles ahead that are in the same lane as the target vehicle; Based on the road surface image of the target vehicle's driving lane, identify the abnormal color block area image of the driving lane; Based on the motion state information and the abnormal color block region image, the abnormal type of the road surface in front of the target vehicle is determined; Based on the aforementioned anomaly type, the driving process of the target vehicle is controlled.

2. The vehicle control method according to claim 1, characterized in that, The process of identifying abnormal color block regions in the driving lane based on the road surface image of the target vehicle's driving lane includes: Lane line detection is performed on the road surface image to obtain the lane area image of the driving lane; Color histogram statistical analysis is performed on the lane area image to obtain the color segmentation threshold; The lane area image is binarized based on the color segmentation threshold to obtain a binarized image; Morphological processing is performed on the binarized image to obtain the image of the abnormal color block region.

3. The vehicle control method according to claim 1, characterized in that, The motion state information includes driving trajectory information and vehicle posture information; determining the anomaly type of the road surface in front of the target vehicle based on the motion state information and the abnormal color block region image includes: Based on the trajectory information, the trajectory category of the vehicle ahead is determined, wherein the trajectory category includes detour category, lane change category, and straight-going category; Based on the vehicle body posture information, the posture category of the vehicle in front is determined, wherein the posture category includes a bumpy category and a smooth category; If the trajectory category is the detour category or the lane change category, and the abnormal color block area image indicates that there is an anomaly on the road surface in front of the target vehicle, then the anomaly type is determined to be a type to be avoided; If the trajectory category is the straight-going category and the posture category is the bumpy category, then the abnormal type of the road surface in front of the target vehicle is determined based on the abnormal color block area image; If the trajectory category is the straight-ahead category and the posture category is the stable category, then based on the abnormal color block area image, the vehicle posture information, and the driving scenario of the target vehicle, the abnormal type of the road surface in front of the target vehicle is determined.

4. The vehicle control method according to claim 3, characterized in that, The step of determining the anomaly type of the road surface in front of the target vehicle based on the image of the abnormal color patch region includes: Obtain the contour information of the abnormal color blocks in the image of the abnormal color block region; Based on the contour information, the shape feature score of the abnormal color block is determined; The height change value of the road surface anomaly is determined based on the vehicle posture information; Based on the shape feature score and the height change value, a first risk probability of the road surface anomaly is determined; If the first risk probability is greater than or equal to the first probability threshold, then the anomaly type is determined to be a type to be avoided; If the first risk probability is less than the first probability threshold, then the anomaly type is determined to be a deceleration passage type.

5. The vehicle control method according to claim 4, characterized in that, The step of determining the shape feature score of the abnormal color block based on the contour information includes: Based on the center point of the contour information, the abnormal color block is horizontally flipped to obtain a horizontal half-fold color block; The horizontal flip overlap rate is determined based on the first overlap area between the abnormal color block and the horizontal half-fold color block, and the area of ​​the abnormal color block. Based on the center point of the contour information, the abnormal color block is vertically flipped to obtain a vertical half-fold color block; Based on the second overlap area between the abnormal color block and the vertical half-fold color block, and the area of ​​the abnormal color block, the vertical flip overlap rate is determined; The circular matching degree is determined based on the horizontal flip overlap rate and the vertical flip overlap rate; Starting from the center point of the contour information, multiple boundary points on the contour are obtained at preset angle intervals, and the dispersion of the boundary points in each preset angle direction is determined. The square matching degree is determined based on the dispersion of the boundary points in multiple preset angular directions; The shape feature score is determined based on the circle matching degree and the square matching degree.

6. The vehicle control method according to claim 3, characterized in that, The step of determining the anomaly type of the road surface in front of the target vehicle based on the abnormal color block region image, the vehicle posture information, and the driving scenario of the target vehicle includes: Based on the image of the abnormal color patch region, a second risk probability is determined; Based on the vehicle body posture change information, the probability of the third risk is determined; Based on the driving scenario, a first weight corresponding to the first risk probability and a second weight corresponding to the second risk probability are determined; Based on the first weight and the second weight, the first risk probability and the second risk probability are weighted and fused to obtain the third risk probability; If the third risk probability is greater than the second probability threshold, then the anomaly type is determined to be a deceleration passage type.

7. The vehicle control method according to claim 4 or 6, characterized in that, The control of the target vehicle's driving process based on the anomaly type includes: If the anomaly type is the type to be avoided, then a target control strategy is generated to control the target vehicle to detour around the abnormal road surface. If the abnormality type is the deceleration passage type, then determine whether the target vehicle has lane change conditions or detour conditions; If the lane-changing condition or the detour condition exists, a target control strategy is generated to control the target vehicle to detour around the abnormal road surface. If the lane-changing condition or the detour condition does not exist, a target control strategy is generated to control the target vehicle to decelerate through the abnormal road surface. Based on the target control strategy, the driving process of the target vehicle is controlled.

8. A vehicle control device, characterized in that, Applied to the target vehicle, the vehicle control device includes: An information acquisition unit is used to acquire motion status information of vehicles ahead that are in the same lane as the target vehicle; An image recognition unit is used to identify abnormal color block areas in the driving lane based on a road surface image of the target vehicle's driving lane; An anomaly determination unit is used to determine the anomaly type of the road surface in front of the target vehicle based on the motion state information and the image of the anomaly color block region; A vehicle control unit is used to control the driving process of the target vehicle based on the type of anomaly.

9. A vehicle comprising: The memory and processor are characterized in that the processor is used to implement the steps of the vehicle control method as described in any one of claims 1 to 7 when executing a computer program stored in the memory.

10. A computer-readable storage medium having stored thereon computer-executable instructions or a computer program, characterized in that, When the computer-executable instructions or the computer program are executed by a processor, the steps of the vehicle control method as described in any one of claims 1 to 7 are implemented.

Images