Intersection opposite-coming vehicle direction and trajectory determination method and related device

Abstract

The application discloses a kind of intersection opposite car direction and trajectory determination method and related equipment, comprising: obtaining opposite vehicle image information;Opposite vehicle corner point information is obtained based on opposite vehicle image information;The corner point close to the vehicle is taken as reference point;Reference point speed information and heading angle information are obtained based on the position information of reference point corresponding to at least two signal periods;Vehicle predicted position information and predicted heading angle information are obtained according to reference point speed information, heading angle information and heading angular velocity information based on basic motion model;According to vehicle predicted position information and predicted heading angle information, one-step prediction is carried out using first Kalman filter, and next time state value and its covariance are obtained;Based on next time state value and its covariance and the measurement data of sensor, the predicted value is corrected using second Kalman filter, and the state estimation value and covariance of current time are calculated;The orientation and trajectory are determined based on the state estimation value and covariance of current time.

Description

Technical Field

[0001] This specification relates to the field of intelligent driving, and more specifically, this application relates to a method and related equipment for determining the position and trajectory of oncoming vehicles at an intersection. Background Technology

[0002] Current autonomous driving systems primarily focus on detecting vehicles traveling in the same direction, such as adaptive cruise control (ACC) and automatic emergency braking (AEB). However, these systems typically lack the ability to detect oncoming vehicles, limiting their application scenarios. Furthermore, existing systems often rely on multiple millimeter-wave radars or other expensive equipment, making it difficult to maintain a competitive advantage in cost-sensitive markets.

[0003] Millimeter-wave radar has a weak ability to identify static objects, failing to effectively distinguish between objects such as roadside trees, buildings, and traffic signs, which may lead to false alarms or missed detections. The insufficient angular resolution of millimeter-wave radar makes it difficult to accurately identify detailed vehicle features, such as vehicle model and color, thus affecting recognition accuracy. In adverse weather conditions (such as rain and snow) or when affected by road contamination, the detection performance of millimeter-wave radar deteriorates significantly, and it may even malfunction, severely impacting system stability. The large amount of data generated by millimeter-wave radar requires complex data processing and algorithm optimization, significantly increasing system complexity and development difficulty, and reducing implementation efficiency.

[0004] Due to these limitations, existing technologies perform poorly in intersection scenarios and rely on millimeter-wave radar or other expensive sensor configurations, resulting in high system costs and making them difficult to adapt to today's highly competitive market environment. Therefore, there is an urgent need to optimize and simplify existing solutions to reduce system complexity and cost. Summary of the Invention

[0005] The summary section introduces a series of simplified concepts, which will be further explained in detail in the detailed description section. This summary section is not intended to limit the key and essential technical features of the claimed technical solution, nor is it intended to determine the scope of protection of the claimed technical solution.

[0006] Firstly, this application proposes a method for determining the orientation and trajectory of oncoming vehicles at an intersection, including:

[0007] Acquire image information of oncoming vehicles;

[0008] Based on the aforementioned oncoming vehicle image information, obtain the corner point information of the oncoming vehicle;

[0009] Use the corner point closest to the vehicle as a reference point;

[0010] Based on the reference point position information corresponding to at least two signal cycles mentioned above, obtain the reference point velocity information and heading angle information;

[0011] Based on the basic motion model, the vehicle's predicted position information and predicted heading angle information are obtained according to the above-mentioned reference point velocity information, heading angle information and heading angular velocity information.

[0012] Based on the vehicle's predicted position information and the predicted heading angle information, a first Kalman filter is used to perform a one-step prediction to obtain the state value and its covariance at the next moment.

[0013] Based on the state value and its covariance at the next time step and the sensor measurement data, the predicted value is corrected using the second Kalman filter, and the state estimate and covariance at the current time step are calculated.

[0014] Based on the current state estimate and covariance, the orientation and trajectory of the oncoming vehicle are determined.

[0015] In one feasible implementation, when the identification is complete, the corner point information of the oncoming vehicle is the eight corner endpoints of the outer contour cuboid corresponding to the oncoming vehicle; when the identification is incomplete, it is a partial corner endpoint of the outer contour cuboid corresponding to the vehicle.

[0016] In one feasible implementation, the acquisition of reference point velocity information and heading angle information based on the reference point position information corresponding to at least two signal periods includes:

[0017] Differential solution is performed based on the reference point position information corresponding to at least two signal periods to obtain the reference point velocity information;

[0018] Based on the reference point position information corresponding to at least two signal cycles mentioned above, curve fitting is performed to obtain heading angle information.

[0019] In one feasible implementation, when the aforementioned opposing vehicles are traveling on a curve, the aforementioned basic motion model is a nonlinear motion model.

[0020] In one feasible implementation, when the opposing vehicles are traveling on a straight road, the basic motion model is a linear motion model.

[0021] In one feasible implementation, the first Kalman filter described above is constructed based on a constant rotational speed and velocity model.

[0022] In one feasible implementation, the filtering gain K of the second Kalman filter described above... k for:

[0023]

[0024] The Jacobian matrix of the sensor measurement matrix. Let P be the Jacobian transpose of the sensor measurement matrix, and let R be the sensor measurement noise covariance matrix. k Let be the state covariance at the current moment.

[0025] Secondly, this application proposes a device for determining the orientation and trajectory of oncoming vehicles at an intersection, comprising:

[0026] The first acquisition unit is used to acquire image information of oncoming vehicles;

[0027] The second acquisition unit is used to acquire corner information of the oncoming vehicle based on the aforementioned oncoming vehicle image information.

[0028] The first determining unit is used to use the corner point that is closer to the vehicle as a reference point;

[0029] The third acquisition unit is used to acquire reference point velocity information and heading angle information based on the reference point position information corresponding to at least two signal cycles mentioned above.

[0030] The fourth acquisition unit is used to acquire vehicle predicted position information and predicted heading angle information based on the above reference point velocity information, the above heading angle information and heading angular velocity information according to the basic motion model;

[0031] The fifth acquisition unit is used to perform a one-step prediction using the first Kalman filter based on the vehicle predicted position information and the predicted heading angle information to obtain the state value and its covariance at the next moment.

[0032] The calculation unit is used to calculate the state estimate and covariance at the current time based on the state value and its covariance at the next time moment and the sensor measurement data, by correcting the predicted value using a second Kalman filter.

[0033] The second determining unit is used to determine the orientation and trajectory of the oncoming vehicle based on the state estimate and covariance at the current moment.

[0034] Thirdly, an electronic device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program stored in the memory to implement the steps of the method for determining the orientation and trajectory of oncoming vehicles at a crossroads as described in any of the first aspects above.

[0035] Fourthly, this application also proposes a computer-readable storage medium storing a computer program thereon, wherein when the computer program is executed by a processor, it implements the method for determining the orientation and trajectory of oncoming vehicles at a crossroads as described in any of the first aspects.

[0036] In summary, existing millimeter-wave radars, due to their low target resolution, cannot accurately identify detailed vehicle features, such as model and color, in detecting oncoming vehicles. This affects the system's effectiveness in complex intersection scenarios. The method proposed in this application, by using a forward-looking vision camera combined with advanced image processing algorithms, can more accurately extract vehicle corner information (such as upper corners and outer cuboid corners), thereby identifying the location and trajectory of oncoming vehicles. The vision device can acquire clearer images, identify more details, and improve the accuracy of judging vehicle motion status. This invention: By using image information acquired by a camera and combining it with algorithms to identify the corners of static objects and dynamic vehicles, it can effectively avoid the problem caused by the radar system's poor ability to identify static objects. The camera can not only distinguish between static and dynamic objects, but also filter out irrelevant static targets through algorithms, thereby improving the system's accuracy. This invention relies only on a system equipped with surround-view cameras and a forward-looking vision camera, significantly reducing the complexity and cost of the system hardware. Meanwhile, processing image data acquired by cameras avoids the complex data processing requirements of millimeter-wave radar in intersection assist functions, making the system more economical and efficient, and adapting to the current cost-sensitive market demands. The camera system performs more stably under certain weather conditions, and combined with image processing algorithms, it can track and predict oncoming vehicles even in complex weather conditions using clear image information. Furthermore, the image processing algorithm can optimize image quality under adverse environmental conditions through image enhancement techniques, ensuring the system's reliability and continuity. This invention uses a Kalman filter to predict and correct vehicle motion states. A first-stage prediction is performed using a first Kalman filter, and the prediction is corrected based on sensor measurement data using a second Kalman filter, enabling more efficient updates to vehicle speed, position, and heading angle. The application of a two-stage Kalman filter significantly improves the prediction accuracy of oncoming vehicles, reduces errors, and achieves precise tracking of the target vehicle's trajectory. By acquiring image data through surround-view cameras and processing it using a Kalman filter-based algorithm, faster response speed and higher real-time performance can be achieved. In environments with significant noise and environmental interference, the Kalman filter can improve system robustness through multiple iterations, making the calculation of oncoming vehicle location and trajectory more reliable. This invention, by optimizing the system structure and reducing reliance on expensive hardware (such as millimeter-wave radar), relies solely on a forward-facing vision camera and algorithm design. This significantly reduces system complexity and cost while improving the accuracy of oncoming vehicle recognition and tracking. In complex intersection scenarios, this invention effectively overcomes the limitations of millimeter-wave radar systems, enhancing the system's adaptability to various weather and environmental conditions, making vehicle location and trajectory prediction more practical and stable. Attached Figure Description

[0037] 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:

[0038] Figure 1 This is a flowchart illustrating a method for determining the orientation and trajectory of oncoming vehicles at an intersection, as provided in an embodiment of this application.

[0039] Figure 2 This application provides a schematic diagram of a scenario for determining the orientation and trajectory of oncoming vehicles at an intersection, as illustrated in an embodiment of the present application.

[0040] Figure 3 This is a schematic diagram illustrating another scenario for determining the orientation and trajectory of oncoming vehicles at an intersection, provided in an embodiment of this application.

[0041] Figure 4 A schematic diagram illustrating another scenario for determining the orientation and trajectory of oncoming vehicles at an intersection, provided in an embodiment of this application.

[0042] Figure 5 A schematic diagram of a device for determining the orientation and trajectory of oncoming vehicles at an intersection, provided in an embodiment of this application;

[0043] Figure 6 This is a schematic diagram of an electronic device for determining the orientation and trajectory of oncoming vehicles at a crossroads, provided in an embodiment of this application. Detailed Implementation

[0044] The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus. The technical solutions of the embodiments of this application will now be clearly and completely described in conjunction with the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them.

[0045] Please see Figure 1This is a flowchart illustrating a method for determining the orientation and trajectory of oncoming vehicles at an intersection, provided in an embodiment of this application. Specifically, it may include:

[0046] S110. Obtain image information of oncoming vehicles;

[0047] For example, image data of oncoming vehicles is acquired from sensors, and this image data will serve as the basis for extracting vehicle corner information in subsequent steps. Image information can be acquired through cameras or other vision devices and used as input data in the next step.

[0048] S120. Obtain corner information of the oncoming vehicle based on the above-mentioned oncoming vehicle image information;

[0049] For example, based on the image information acquired in S110, the corner points of the target vehicle are identified. Corner points are salient points on the outer contour of the vehicle and are typically used to describe the vehicle's spatial position and orientation. In this step, the algorithm needs to analyze the image and extract corner point information related to the vehicle, such as the top corner of the vehicle and the corner points of the outer cuboid.

[0050] S130. Use the corner point closest to the vehicle as a reference point;

[0051] For example, among multiple corner points, the one closest to the vehicle is selected as the reference point. This reference point represents key positional information of oncoming vehicles, and subsequent motion prediction and heading angle estimation will be calculated based on the positional changes of this reference point.

[0052] S140. Based on the reference point position information corresponding to at least two signal cycles, obtain the reference point velocity information and heading angle information;

[0053] For example, the position information of the reference point is acquired over at least two signal cycles. By comparing the changes in the reference point's position during these cycles, the velocity information of the reference point and the heading angle information of the target vehicle are calculated. The heading angle is the angle between the target vehicle's direction of travel and the Z-axis, which is derived from the change in the reference point's position over time.

[0054] S150. Based on the basic motion model, obtain the vehicle's predicted position information and predicted heading angle information according to the above-mentioned reference point velocity information, heading angle information and heading angular velocity information;

[0055] For example, based on the basic motion model, the future position and heading angle of the target vehicle are predicted using the reference point velocity information, heading angle information, and heading angular velocity information calculated in the previous step. The basic motion model can describe the vehicle's linear or curved driving; by combining these parameters, the future driving state of the vehicle can be estimated.

[0056] S160. Based on the vehicle's predicted position information and the predicted heading angle information, a first Kalman filter is used to perform a one-step prediction to obtain the state value and its covariance at the next moment.

[0057] For example, a first Kalman filter is used to perform a one-step state prediction based on the vehicle's predicted position and predicted heading angle. This step outputs the state value and its covariance matrix for the next time step. The state value includes dynamic information such as the vehicle's position and speed, and the covariance matrix reflects the uncertainty of these predicted values.

[0058] S170. Based on the state value and its covariance at the next time step and the sensor measurement data, the predicted value is corrected using the second Kalman filter, and the state estimate and covariance at the current time step are calculated.

[0059] For example, by combining real-time measurement data from the sensors, a second Kalman filter is used to correct the prediction result from the previous step. By comparing the measurement data with the prediction result, the current state estimate and covariance matrix are updated. This step improves the system's accuracy in estimating the current state and reduces prediction errors.

[0060] S180. Determine the orientation and trajectory of the oncoming vehicle based on the current state estimate and covariance.

[0061] For example, the specific location and trajectory of an oncoming vehicle can be determined based on the current state estimate and covariance. By integrating the vehicle's position information, speed, and heading angle, the movement of oncoming vehicles can be accurately tracked, and the vehicle's trajectory information can be updated in real time.

[0062] In summary, existing millimeter-wave radars, due to their low target resolution, cannot accurately identify detailed vehicle features, such as model and color, in detecting oncoming vehicles. This affects the system's effectiveness in complex intersection scenarios. The method proposed in this application, by using a forward-looking vision camera combined with advanced image processing algorithms, can more accurately extract vehicle corner information (such as upper corners and outer cuboid corners), thereby identifying the location and trajectory of oncoming vehicles. The vision device can acquire clearer images, identify more details, and improve the accuracy of judging vehicle motion status. This invention: By using image information acquired by a camera and combining it with algorithms to identify the corners of static objects and dynamic vehicles, it can effectively avoid the problem caused by the radar system's poor ability to identify static objects. The camera can not only distinguish between static and dynamic objects, but also filter out irrelevant static targets through algorithms, thereby improving the system's accuracy. This invention relies only on a system equipped with surround-view cameras and a forward-looking vision camera, significantly reducing the complexity and cost of the system hardware. Meanwhile, processing image data acquired by cameras avoids the complex data processing requirements of millimeter-wave radar in intersection assist functions, making the system more economical and efficient, and adapting to the current cost-sensitive market demands. The camera system performs more stably under certain weather conditions, and combined with image processing algorithms, it can track and predict oncoming vehicles even in complex weather conditions using clear image information. Furthermore, the image processing algorithm can optimize image quality under adverse environmental conditions through image enhancement techniques, ensuring the system's reliability and continuity. This invention uses a Kalman filter to predict and correct vehicle motion states. A first-stage prediction is performed using a first Kalman filter, and the prediction is corrected based on sensor measurement data using a second Kalman filter, enabling more efficient updates to vehicle speed, position, and heading angle. The application of a two-stage Kalman filter significantly improves the prediction accuracy of oncoming vehicles, reduces errors, and achieves precise tracking of the target vehicle's trajectory. By acquiring image data through surround-view cameras and processing it using a Kalman filter-based algorithm, faster response speed and higher real-time performance can be achieved. In environments with significant noise and environmental interference, the Kalman filter can improve system robustness through multiple iterations, making the calculation of oncoming vehicle location and trajectory more reliable. This invention, by optimizing the system structure and reducing reliance on expensive hardware (such as millimeter-wave radar), relies solely on a forward-facing vision camera and algorithm design. This significantly reduces system complexity and cost while improving the accuracy of oncoming vehicle recognition and tracking. In complex intersection scenarios, this invention effectively overcomes the limitations of millimeter-wave radar systems, enhancing the system's adaptability to various weather and environmental conditions, making vehicle location and trajectory prediction more practical and stable.

[0063] In some examples, when the identification is complete, the corner point information of the oncoming vehicle is the eight corner endpoints of the outer contour cuboid corresponding to the oncoming vehicle; when the identification is incomplete, it is some of the corner endpoints of the outer contour cuboid corresponding to the vehicle.

[0064] For example, the outlines identified by the target vehicle are numbered. Note: Generally, each sensor output has its own output rules. This invention is explained in the following order for ease of explanation. (Lower left is number 1, numbered counter-clockwise); other targets are numbered sequentially (if identified), such as... Figure 2 As shown. Based on the reference point, when the target vehicle proceeds straight through the intersection, the upper corner of the vehicle's outer cuboid is selected as the reference point. That is, corner point 3 in the figure; when the target vehicle turns left or right through the intersection at time t, the nearest upper corner point of the outer cuboid of the target vehicle and the vehicle itself is selected as the reference point. When the target vehicle proceeds straight through the intersection without changing direction, the reference point is... It remains the corner point of the rectangular prism surrounding the vehicle from the previous moment.

[0065] like Figure 3 As shown, when the target vehicle first goes straight and then turns left into the left blind spot and passes through the intersection, the reference point first switches from the 3-corner point to the 9-corner point, and then switches from the 9-corner point to the 11-corner point after entering the left blind spot.

[0066] like Figure 4 As shown, when the target vehicle first goes straight and then turns right into the right blind spot and passes through the intersection, the reference point first switches from the 3-corner point to the 10-corner point, and then switches from the 10-corner point to the 12-corner point after entering the right blind spot.

[0067] In some examples, the reference point velocity information and heading angle information are obtained based on the reference point position information corresponding to at least two signal periods, including:

[0068] Differential solution is performed based on the reference point position information corresponding to at least two signal periods to obtain the reference point velocity information;

[0069] Based on the reference point position information corresponding to at least two signal cycles mentioned above, curve fitting is performed to obtain heading angle information.

[0070] For example, reference point location information over at least two signal cycles.

[0071] Operation: Based on the position of the same reference point in two different signal periods, perform differential calculation to calculate the velocity of the target vehicle in each direction. The specific formula is as follows:

[0072]

[0073]

[0074] Where, x reft and z reft These are the lateral and longitudinal coordinates of the reference point in the vehicle coordinate system at time t, where T is the signal period. Using this difference formula, the velocity information v of the reference point in the x and z directions can be obtained. xt and v zt That is, the speed of the vehicle.

[0075] In complex road traffic environments, vehicles frequently change lanes and turn, requiring intelligent vehicles to accurately and quickly identify their heading angle. Based on the vehicle's axisymmetry, the heading angle of the target vehicle... This can be represented as the angle between its driving direction and the Z-axis of the stereo vision system coordinate system, as shown in the figure.

[0076] Within a short period (several sampling cycles or dozens of sampling cycles), the sensor measurement data is fitted piece by piece, and the motion of the target vehicle is represented by the following quadratic curve:

[0077]

[0078] The parameters a, b, and c of the conic section can be obtained using the least squares estimation method, and the slope of the conic section can be obtained by differentiation.

[0079]

[0080] The slope of the curve represents its tangent direction, which is the instantaneous direction of motion of the target vehicle or the opposite direction. Taking the arctangent of the curve's slope yields the angle between the tangent direction and the Z-axis, i.e., the heading angle of the target vehicle relative to the observing vehicle.

[0081]

[0082] In some examples, when the opposing vehicles are traveling on a curve, the basic motion model described above is a nonlinear motion model.

[0083] For example, in vehicle tracking based on the motion characteristics of vehicles at intersections, this invention selects a basic vehicle motion model, and the state vector of the target vehicle is represented as follows:

[0084]

[0085] In the formula, the position of the target vehicle in the vehicle coordinate system is represented by a data set. express, The target vehicle's speed, yaw angle, and yaw rate are respectively .

[0086] When the vehicle is driving on a curve, When the function is:

[0087]

[0088] Where Z represents the longitudinal position of the target vehicle in the vehicle coordinate system (usually indicating the vehicle's forward and backward direction). X represents the lateral position of the target vehicle in the vehicle coordinate system (usually indicating the vehicle's left and right direction). V represents the velocity of the target vehicle (i.e., the speed at which the vehicle moves). Let ω be the yaw angle of the target vehicle, representing the angle between the vehicle's direction of travel and the reference direction (usually the Z-axis). ω is the yaw rate of the target vehicle, representing the rate of rotation of the vehicle when turning. It describes the speed at which the vehicle's angle changes during travel. V is the vehicle's speed. It represents the speed at which the vehicle travels along the curve and is used in the formula to calculate the change in position.

[0089] and These terms describe the periodic characteristics of a vehicle's position change over time during cornering. When cornering, the vehicle's position change is affected by the yaw angle. The influence of yaw rate ω.

[0090] In some examples, when the opposing vehicles are traveling on a straight road, the basic motion model described above is a linear motion model.

[0091] For example, if the vehicle is traveling in a straight line, the state transition function is:

[0092]

[0093]

[0094] The basic principle of the algorithm in this invention is to calculate the filtered value at the current time using the state estimate from the previous time step and the observed value at the current time step. This filtering process can be divided into two stages: the state prediction stage and the state update stage.

[0095] In some examples, the first Kalman filter described above is constructed based on a constant speed and velocity model.

[0096] For example, it is assumed that the target vehicle's speed v and yaw rate ω are constant. That is, the vehicle's speed and steering rate remain unchanged during prediction. This assumption is suitable for short-term vehicle motion prediction, such as short-term motion prediction of oncoming vehicles in a crossroads scenario.

[0097] The Extended Kalman Filter (EKF) treats nonlinear systems by linearizing their state equations using the first term of a Taylor series expansion. The state prediction equation is:

[0098]

[0099] Where: x k+1 This represents the state vector at time k+1.

[0100] g(x k ,u) represents the state transition function based on a constant tachometer and speed model, which describes the state change of the target vehicle from time k to time k+1.

[0101] X k Let be the state vector at time k, including position, velocity, yaw angle, etc.

[0102] u represents noise processing, which includes the impact of external disturbances on the vehicle's state during operation.

[0103] in, This represents the state transition function for a constant speed and velocity model. This indicates noise handling. The one-step state prediction covariance is:

[0104]

[0105] in, Q is the Jacobian matrix obtained by taking the partial derivative of each element in the state transition function, and Q is the covariance matrix for handling noise.

[0106] In some examples, the filtering gain K of the second Kalman filter described above... k for:

[0107]

[0108] The Jacobian matrix of the sensor measurement matrix. Let R be the Jacobian transpose of the sensor measurement matrix, R be the sensor measurement noise covariance matrix, and Pk be the state covariance at the current time.

[0109] For example, in the formula, R is the Jacobian matrix of the sensor measurement matrix, and R is the sensor measurement noise covariance matrix.

[0110] State estimation equation:

[0111]

[0112] In the formula, The sensor measurement value. It is a non-linear mapping function.

[0113] State estimation covariance

[0114]

[0115] In the formula, I is the identity matrix.

[0116] like Figure 5 As shown, this application proposes a device for determining the orientation and trajectory of oncoming vehicles at an intersection, comprising:

[0117] The first acquisition unit 21 is used to acquire image information of oncoming vehicles;

[0118] The second acquisition unit 22 is used to acquire corner information of the oncoming vehicle based on the above-mentioned oncoming vehicle image information;

[0119] The first determining unit 23 is used to use a corner point that is closer to the vehicle as a reference point;

[0120] The third acquisition unit 24 is used to acquire reference point velocity information and heading angle information based on the reference point position information corresponding to at least two signal periods.

[0121] The fourth acquisition unit 25 is used to acquire vehicle predicted position information and predicted heading angle information based on the reference point velocity information, the heading angle information and the heading angle velocity information according to the basic motion model.

[0122] The fifth acquisition unit 26 is used to perform a one-step prediction using the first Kalman filter based on the vehicle predicted position information and the predicted heading angle information to obtain the state value and its covariance at the next moment.

[0123] The calculation unit 27 is used to calculate the state estimate and covariance at the current moment based on the state value and its covariance at the next moment and the sensor measurement data, and to correct the predicted value using a second Kalman filter.

[0124] The second determining unit 28 is used to determine the orientation and trajectory of the oncoming vehicle based on the state estimate and covariance at the current moment.

[0125] In some examples, when the identification is complete, the corner point information of the oncoming vehicle is the eight corner endpoints of the outer contour cuboid corresponding to the oncoming vehicle; when the identification is incomplete, it is some of the corner endpoints of the outer contour cuboid corresponding to the vehicle.

[0126] In some examples, the reference point velocity information and heading angle information are obtained based on the reference point position information corresponding to at least two signal periods, including:

[0127] Differential solution is performed based on the reference point position information corresponding to at least two signal periods to obtain the reference point velocity information;

[0128] Based on the reference point position information corresponding to at least two signal cycles mentioned above, curve fitting is performed to obtain heading angle information.

[0129] In some examples, when the opposing vehicles are traveling on a curve, the basic motion model described above is a nonlinear motion model.

[0130] In some examples, when the opposing vehicles are traveling on a straight road, the basic motion model described above is a linear motion model.

[0131] In some examples, the first Kalman filter described above is constructed based on a constant speed and velocity model.

[0132] In some examples, the filtering gain K of the second Kalman filter described above... k for:

[0133]

[0134] The Jacobian matrix of the sensor measurement matrix. Let P be the Jacobian transpose of the sensor measurement matrix, and let R be the sensor measurement noise covariance matrix. k Let be the state covariance at the current moment.

[0135] like Figure 6 As shown, this application embodiment also provides an electronic device 300, 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 the steps of any of the above-mentioned methods for determining the orientation and trajectory of oncoming vehicles at a crossroads.

[0136] Since the electronic device described in this embodiment is the device used to implement the crossroads oncoming vehicle position and trajectory determination device in this application embodiment, those skilled in the art can understand the specific implementation method and various variations of the electronic device in this embodiment based on the method described in this application embodiment. Therefore, how the electronic device implements the method in this application embodiment will not be described in detail here. Any device used by those skilled in the art to implement the method in this application embodiment is within the scope of protection of this application.

[0137] In practical implementation, when the computer program 311 is executed by the processor, it can achieve the following: Figure 1 Any of the corresponding implementation methods in the embodiments.

[0138] It should be noted that the descriptions of each embodiment in the above embodiments have different focuses. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0139] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0140] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0141] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0142] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0143] This application also provides a computer program product, which includes computer software instructions. When the computer software instructions are run on a processing device, the processing device executes the process of determining the orientation and trajectory of oncoming vehicles at a crossroads in the corresponding embodiment.

[0144] A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of this application is generated. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a server or data center that integrates one or more available media. The available medium may be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state disk (SSD)).

[0145] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0146] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between apparatuses or units, and may be electrical, mechanical, or other forms.

[0147] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0148] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0149] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0150] 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 method for determining the orientation and trajectory of oncoming vehicles at an intersection, characterized in that, include: Acquire image information of oncoming vehicles; Obtain corner information of the oncoming vehicle based on the oncoming vehicle image information; Use the corner point closest to the vehicle as a reference point; Based on the reference point position information corresponding to at least two signal cycles, obtain the reference point velocity information and heading angle information; Based on the basic motion model, the vehicle's predicted position information and predicted heading angle information are obtained according to the reference point velocity information, the heading angle information, and the heading angular velocity information. Based on the vehicle's predicted position information and the predicted heading angle information, a first Kalman filter is used to perform a one-step prediction to obtain the state value and its covariance at the next moment. Based on the state value and its covariance at the next moment and the sensor's measurement data, the predicted value is corrected using a second Kalman filter, and the state estimate and covariance at the current moment are calculated. The orientation and trajectory of the oncoming vehicle are determined based on the current state estimate and covariance.

2. The method for determining the orientation and trajectory of oncoming vehicles at an intersection according to claim 1, characterized in that, When the identification is complete, the corner point information of the oncoming vehicle is the 8 corner endpoints of the outer contour cuboid corresponding to the oncoming vehicle; when the identification is incomplete, it is some of the corner endpoints of the outer contour cuboid corresponding to the vehicle.

3. The method for determining the orientation and trajectory of oncoming vehicles at an intersection according to claim 1, characterized in that, The acquisition of reference point velocity information and heading angle information based on reference point position information corresponding to at least two signal periods includes: The reference point velocity information is obtained by differential solution based on the reference point position information corresponding to at least two signal periods. Curve fitting is performed based on the reference point position information corresponding to at least two signal cycles of the reference point to obtain the heading angle information.

4. The method for determining the orientation and trajectory of oncoming vehicles at an intersection according to claim 1, characterized in that, When the oncoming vehicle is traveling on a curve, the basic motion model is a nonlinear motion model.

5. The method for determining the orientation and trajectory of oncoming vehicles at an intersection according to claim 1, characterized in that, When the oncoming vehicle is traveling on a straight road, the basic motion model is a linear motion model.

6. The method for determining the orientation and trajectory of oncoming vehicles at an intersection according to claim 1, characterized in that, The first Kalman filter is constructed based on a constant rotation rate and velocity model.

7. The method for determining the orientation and trajectory of oncoming vehicles at an intersection according to claim 1, characterized in that, The filtering gain K of the second Kalman filter k for: The Jacobian matrix of the sensor measurement matrix. Let P be the Jacobian transpose of the sensor measurement matrix, and let R be the sensor measurement noise covariance matrix. k Let be the state covariance at the current moment.

8. A device for determining the orientation and trajectory of oncoming vehicles at a crossroads, characterized in that, include: The first acquisition unit is used to acquire image information of oncoming vehicles; The second acquisition unit is used to acquire corner information of the oncoming vehicle based on the oncoming vehicle image information; The first determining unit is used to use the corner point that is closer to the vehicle as a reference point; The third acquisition unit is used to acquire reference point velocity information and heading angle information based on reference point position information corresponding to at least two signal cycles. The fourth acquisition unit is used to acquire vehicle predicted position information and predicted heading angle information based on the reference point velocity information, the heading angle information and the heading angular velocity information according to the basic motion model; The fifth acquisition unit is used to perform a one-step prediction using the first Kalman filter based on the vehicle's predicted position information and the predicted heading angle information, and to obtain the state value and its covariance at the next moment. The calculation unit is used to calculate the state estimate and covariance at the current moment based on the state value and its covariance at the next moment and the measurement data from the sensor, by correcting the predicted value using a second Kalman filter; The second determining unit is used to determine the orientation and trajectory of the oncoming vehicle based on the state estimate and covariance at the current moment.

9. An electronic device, comprising: The memory and processor are characterized in that the processor is used to execute a computer program stored in the memory to implement the steps of the method for determining the orientation and trajectory of oncoming vehicles at a crossroads as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the method for determining the orientation and trajectory of oncoming vehicles at a crossroads as described in any one of claims 1-7.

Images