Dynamic aggregation duration based on vehicle's radar point cloud

Abstract

Examples described herein provide a method that includes detecting, using a radar device of a vehicle, a reflection point on a target object and determining a dynamic aggregation duration for aggregating radar data for the reflection point, the dynamic aggregation duration being a time period specific to the reflection point and based at least in part on a radial velocity of the reflection point, a reflection point angle of the reflection point, and a distribution of headings of the reflection point. The method further includes aggregating radar data for the target object for the time period specific to the reflection point defined by the dynamic aggregation duration. The method further includes performing a perception task using the aggregated radar data for the target object aggregated for the time period specific to the reflection point defined by the dynamic aggregation duration.

Description

Technical Field

[0001] This topic discloses information about vehicles, and specifically about the duration of dynamic aggregation for vehicle-based radar point clouds. Background Technology

[0002] Modern vehicles (e.g., cars, motorcycles, boats, or any other type of vehicle) may be equipped with one or more cameras that provide reversing assistance, capture images of the vehicle's driver to determine driver drowsiness or attention, provide images of the road while the vehicle is in motion for collision avoidance purposes, provide structure recognition (e.g., road signs, etc.), and include combinations and / or multiple cameras. For example, a vehicle may be equipped with multiple cameras, and images from multiple cameras (referred to as "surround view cameras") can be used to create a "surround" or "bird's-eye view" view of the vehicle. Some of the cameras (referred to as "remote cameras") may be used to capture remote images (e.g., object detection for collision avoidance, structure recognition, etc.).

[0003] Such vehicles can also be equipped with sensors for perception tasks, such as radar units, lidar units, and / or the like. Radar (radio detection and ranging) is a technology that uses radio waves to detect and determine the distance, speed, and angle of objects. Radar works by emitting radio signals that bounce off an object and return to the radar system, where the reflected waves are analyzed based on the amount of time between transmission and reception. The measured time can be used to determine the distance between the radar unit and the detected object, which can be used when performing perception tasks.

[0004] Perception tasks can include one or more of object detection, classification, tracking, lane detection, road sign recognition, and obstacle avoidance. Perception tasks are particularly useful for autonomous vehicles to provide them with real-time awareness of their environment to make safe and informed driving decisions. Images from one or more cameras on the vehicle can also be used to detect objects, track targets, etc., including combinations and / or multiple combinations thereof.

[0005] The expectation for accurate object detection remains in applications such as autonomous driving, where real-time environmental perception is crucial for safe navigation. Summary of the Invention

[0006] In one embodiment, a computer-implemented method is provided. The method includes detecting reflection points on a target object using a vehicle's radar apparatus. The method further includes determining a dynamic aggregation duration for aggregating radar data of the reflection points, the dynamic aggregation duration being a reflection point-specific time period, wherein the dynamic aggregation duration is based at least in part on the radial velocity of the reflection point, the reflection point angle of the reflection point, and the heading distribution of the reflection point. The method also includes aggregating radar data of the target object as aggregated radar data within the reflection point-specific time period defined by the dynamic aggregation duration. The method further includes performing a perception task using the aggregated radar data of the target object aggregated within the reflection point-specific time period defined by the dynamic aggregation duration.

[0007] In addition to one or more features described herein, or as an alternative, other implementations of the method may include at least in part based on the results of the perception task derived from the primary driving vehicle.

[0008] In addition to one or more features described herein, or as an alternative, other implementations of the method may include determining the dynamic aggregation duration by determining a position offset, which is an offset between the true position of the reflection point if it is shifted from a past time point to the current time point by the true velocity vector and the predicted position of the reflection point when it is shifted according to the radial velocity of the reflection point, setting a limit on the position offset, deriving a finite aggregation time for each reflection point based on the limit and depending on the radial velocity of the reflection point, the reflection point angle of the reflection point, and the heading distribution of the reflection point, and deriving the aggregation duration for each reflection point based on its limit.

[0009] In addition to one or more of the features described herein, or as an alternative, other implementations of the method may include dynamic aggregation duration based on a Laplace distribution, which is the heading distribution of the reflection points.

[0010] In addition to one or more of the features described herein, or as an alternative, further embodiments of the method may include determining the positional offset using the following formula:

[0011]

[0012] Where v r θ is the velocity of the reflection point, T is the convergence time, θ is the angle of the reflection point relative to the reflection point of the radar device, and α is the heading of the reflection point relative to the reflection point of the radar device.

[0013] In addition to one or more of the features described herein, or as an alternative, further embodiments of the method may include determining the positional offset using the following formula:

[0014]

[0015] Where v r θ is the velocity of the reflection point, T is the convergence time, θ is the angle of the reflection point relative to the reflection point of the radar device, and α is the heading of the reflection point relative to the reflection point of the radar device.

[0016] In addition to one or more of the features described herein, or as an alternative, further embodiments of the method may include determining the limits of the positional offset using the following formula:

[0017]

[0018] Where v r θ is the velocity of the reflection point, T is the convergence time, θ is the angle of the reflection point relative to the reflection point of the radar device, α is the heading of the reflection point relative to the reflection point of the radar device, and D is the limit.

[0019] In addition to one or more of the features described herein, or as an alternative, further embodiments of the method may include determining the limits of the positional offset using the following formula:

[0020] T|v r |tan(θ+α) <D

[0021] Where v r θ is the velocity of the reflection point, T is the convergence time, θ is the angle of the reflection point relative to the reflection point of the radar device, α is the heading of the reflection point relative to the reflection point of the radar device, and D is the limit.

[0022] In addition to one or more features described herein, or as an alternative, other implementations of the method may include deriving the finite aggregation time using the following formula:

[0023]

[0024] Where v r θ is the velocity of the reflection point, T is the convergence time, θ is the angle of the reflection point relative to the reflection point of the radar device, α is the heading of the reflection point relative to the reflection point of the radar device, and D is the limit.

[0025] In addition to one or more features described herein, or as an alternative, other implementations of the method may include deriving the finite aggregation time using the following formula:

[0026]

[0027] Where v r θ is the velocity of the reflection point, T is the convergence time, θ is the angle of the reflection point relative to the reflection point of the radar device, α is the heading of the reflection point relative to the reflection point of the radar device, and D is the limit.

[0028] In addition to one or more features described herein, or alternatively, other embodiments of the method may include: the dynamic aggregation duration being derived using the following formula:

[0029]

[0030] Where T is the aggregation time, and E α It is the expectation of α, D is the limit of the position offset, θ is the angle of the reflection point relative to the reflection point of the radar device, α is the heading of the reflection point relative to the reflection point of the radar device, and v r p is the velocity of the reflection point. r It is the probability distribution function, L1, defined by the following formula:

[0031]

[0032] L2 is defined by the following formula:

[0033]

[0034] In addition to one or more features described herein, or alternatively, other embodiments of the method may include: the dynamic aggregation duration being derived using the following formula:

[0035]

[0036] Where T is the aggregation time, and E α It is the expectation of α, D is the limit of the position offset, θ is the angle of the reflection point relative to the reflection point of the radar device, α is the heading of the reflection point relative to the reflection point of the radar device, and v r It is the velocity of the reflection point, and p r It is a probability distribution function.

[0037] In another embodiment, a vehicle is provided. The vehicle includes a radar device and a processing system, the processing system including a memory having computer-readable instructions and processing means for executing the computer-readable instructions. The computer-readable instructions control the processing system to perform operations including detecting reflection points on a target object using the vehicle's radar device; determining a dynamic aggregation duration for aggregating radar data of the reflection points, the dynamic aggregation duration being a reflection point-specific time period, wherein the dynamic aggregation duration is at least partially based on the distribution of radial velocity, reflection point angle, and heading of the reflection points; aggregating radar data of the target object as aggregated radar data within the reflection point-specific time period defined by the dynamic aggregation duration; and performing a perception task using the aggregated radar data of the target object aggregated within the reflection point-specific time period defined by the dynamic aggregation duration.

[0038] In addition to one or more features described herein, or as an alternative, other implementations of the vehicle may include: operation also includes autonomous driving of the vehicle based at least in part on the results of a perception task.

[0039] In addition to one or more features described herein, or alternatively, other implementations of the vehicle may include determining the duration of dynamic aggregation by determining a position offset, which is the offset between the actual position of the reflection point when it is shifted from a past time point to the current time point with a true velocity vector and the predicted position of the reflection point when it is shifted according to the radial velocity of the reflection point.

[0040] In addition to one or more of the features described herein, or as an alternative, another embodiment of the vehicle may include: determining the duration of the dynamic aggregation includes setting limits on the position offset.

[0041] In addition to one or more features described herein, or as an alternative, other implementations of the vehicle may include: determining the dynamic aggregation duration by deriving a finite aggregation time for each reflection point based on limits and depending on the radial velocity of the reflection point, the reflection point angle of the reflection point, and the heading distribution.

[0042] In addition to one or more features described herein, or as an alternative, other implementations of the vehicle may include determining the dynamic aggregation duration by deriving the aggregation duration for each reflection point based on the limits of each reflection point.

[0043] In another embodiment, a computer program product is provided. The computer program product includes a set of one or more computer-readable storage media and program instructions, which are collectively stored in the set of one or more storage media, for causing a processor set to perform computer operations. The operations include detecting reflection points on a target object using a radar device of a vehicle; determining a dynamic aggregation duration for aggregating radar data of the reflection points, the dynamic aggregation duration being a reflection point-specific time period, wherein the dynamic aggregation duration is at least partially based on the distribution of radial velocity, reflection point angle, and heading of the reflection points; aggregating radar data of the target object as aggregated radar data within the reflection point-specific time period defined by the dynamic aggregation duration; performing a perception task using the aggregated radar data of the target object aggregated within the reflection point-specific time period defined by the dynamic aggregation duration; and autonomously driving the vehicle at least partially based on the results of the perception task.

[0044] In addition to one or more features described herein, or as an alternative, other implementations of the computer program product may include determining the dynamic aggregation duration by determining a position offset, which is the offset between the actual position of the reflection point when it is moved from a past time point to the current time point with a true velocity vector and the predicted position of the reflection point when it is moved according to the radial velocity of the reflection point, setting a limit on the position offset, deriving a finite aggregation time for each reflection point based on the limit and depending on the radial velocity of the reflection point, the reflection point angle of the reflection point, and the heading distribution, and deriving the aggregation duration for each reflection point based on its limit.

[0045] The above-described features and advantages, as well as other features and advantages, of this disclosure will become apparent when taken in conjunction with the accompanying drawings and the following detailed description. Attached Figure Description

[0046] Other features, advantages, and details appear by way of example only in the following detailed description, which is described in detail with reference to the accompanying drawings, wherein:

[0047] Figure 1 A vehicle with a processing system and sensors according to one or more embodiments is shown;

[0048] Figure 2 The illustration is based on one or more embodiments. Figure 1 The processing system;

[0049] Figure 3A Examples of embodiments of the invention are illustrated. Figure 1 and Figure 2 The reflection point detected by the radar device;

[0050] Figure 3B The illustration shows a method according to one or more embodiments. Figure 1 and Figure 2 The reflection point detected by the radar device;

[0051] Figure 4A A block diagram of a system for dynamic aggregation duration of vehicle-based radar point clouds according to one or more embodiments is shown.

[0052] Figure 4B A block diagram of a system for dynamic aggregation duration of vehicle-based radar point clouds according to one or more embodiments is shown.

[0053] Figure 5 A flowchart is depicted illustrating a method for dynamically aggregating the duration of vehicle-based radar point clouds according to one or more embodiments; and

[0054] Figure 6 A block diagram of an impedance balancing processing system for a battery monitoring circuit for a vehicle, according to one or more embodiments, is shown. Detailed Implementation

[0055] The following description is exemplary in nature only and is not intended to limit this disclosure, its application, or use. It should be understood that throughout the drawings, corresponding reference numerals denote the same or corresponding parts and features. As used herein, the term "module" refers to processing circuitry that may include application-specific integrated circuits (ASICs), electronic circuitry, processor (shared, dedicated, or group) and memory executing one or more software or firmware programs, combinational logic circuitry, and / or other suitable components that provide the described functionality.

[0056] One or more embodiments described herein relate to the duration of dynamic aggregation of vehicle-based radar point clouds.

[0057] Modern vehicle systems rely on advanced technologies to perform perception tasks such as object detection, classification, and tracking. These capabilities are useful for systems that achieve accurate and effective navigation (including semi-autonomous or autonomous operation of the vehicle) by understanding the vehicle's environment in real time. Challenges arise when data suffers from sparsity, leading to potential inaccuracies in detecting and estimating the shape of dynamic objects.

[0058] In particular, radar systems in modern vehicles face challenges in accurately detecting and identifying objects due to the sparsity of radar point clouds. This sparsity stems from low angular resolution, which can lead to missed detections of dynamic objects and inaccurate shape estimations. The expectation for accurate object detection remains in applications such as autonomous driving, where real-time environmental perception is crucial for safe navigation.

[0059] Existing methods aim to increase data density through temporal aggregation. For example, radar data about a target object can be collected over time and aggregated to generate more data about the object than would otherwise be collected at a single point in time. While this enhances density, temporal aggregation introduces complexity, such as point dispersion, especially for objects with unknown velocities. This dispersion can affect the perceived location of the object, making accurate detection and tracking more difficult. Addressing the need to increase data density while minimizing point dispersion remains a significant challenge in this field.

[0060] One or more embodiments described herein utilize the dynamic aggregation duration of radar point clouds, which is determined, for example, by the range, Doppler effect, and angle of the reflection point relative to the vehicle. This method minimizes point dispersion while increasing point cloud density, thereby enhancing the accuracy of object detection and other perception tasks. By adjusting the aggregation duration based on the velocity of the reflection point, one or more embodiments effectively balance density and accuracy, thus providing a rich point cloud with reduced dispersion.

[0061] Figure 1 A vehicle 100 with a processing system 102 and a radar device 104 is shown according to one or more embodiments. The vehicle 100 may be a car, truck, van, bus, motorcycle, boat, or any other type of vehicle. According to one embodiment, the vehicle 100 is a hybrid electric vehicle, such as a plug-in hybrid electric vehicle (PHEV) that is partially or fully powered by electricity. According to another embodiment, the vehicle 100 is an electric vehicle powered by electricity. A battery (not shown) is used to provide power to components of the vehicle 100, such as an electric motor (not shown), electrical components (not shown), etc., including combinations thereof and / or multiples thereof. According to one or more embodiments, the vehicle 100 is an autonomous or semi-autonomous vehicle. An autonomous vehicle is a vehicle with autonomous driving capabilities. A semi-autonomous vehicle is a vehicle with some autonomous features (e.g., self-parking, lane keeping, etc.) but lacks fully autonomous control.

[0062] The processing system 102 is located within the vehicle and is responsible for managing and processing the data collected by the radar device 104. The radar device 104 is strategically positioned on the vehicle to collect data from the vehicle's environment. An arrow between the radar device 104 and the processing system 102 indicates the data flow from the radar device 104 to the processing system 102, highlighting the interaction between these components. This setup enables the vehicle 100 to use the data collected by the radar device 104 to perform task perception tasks, which can be used for, for example, autonomous driving.

[0063] Now for reference Figure 2 Other features of the processing system 102 and the radar device 104 are described.

[0064] In particular, Figure 2 The illustration is based on one or more embodiments. Figure 1 The processing system 102 includes a processing device 202, a memory 204, an aggregation duration engine 210, and a perception task engine 212, according to one or more embodiments. It should be understood that the processing system 102 can be any device suitable for performing or supporting infrastructure access control using vehicle-based LiDAR. For example, the processing system 102 can be a device implemented in or otherwise associated with vehicle 100, such as an electronic control unit (also referred to as an electronic control module). As another example, the processing system 102 can be a smartphone, tablet computer, laptop computer, desktop computer, wearable computing device, etc., including combinations thereof and / or multiple thereof. As yet another example, the processing system 102 can be... Figure 6 The processing system 600 and / or may include Figure 6 One or more components of the processing system 600.

[0065] Processing device 202 is responsible for executing instructions and managing the overall operation of processing system 102. Processing device 202 can be any suitable processing circuitry used for executing instructions and processing data. For example, processing device 202 can be a microcontroller, microprocessor, application-specific integrated circuit (ASIC), or any other type of processing unit capable of handling the computational needs of processing system 102. Processing device 202 is Figure 6 Examples of one or more of the processing devices 621 are described in more detail herein.

[0066] Memory 204 stores data useful for the operation of processing system 102 (e.g., radar data 214), computer-readable instructions, and algorithms. This may include real-time data processing, historical data analysis, and storage of firmware or software programs. Memory 204 is any suitable device for storing data (such as radar data 214) and / or instructions. For example, memory 204 may be a combination of volatile memory (e.g., random access memory) and non-volatile memory (e.g., read-only memory, flash memory). Memory 204 is Figure 6 Examples of one or more of system memory 622, random access memory 623, and / or read-only memory 624 are described in more detail herein.

[0067] Processing system 102 receives radar data 214 from radar unit 104 of objects (such as target object 220) in the environment in which vehicle 100 is operating. Radar data 214 is used to generate a point cloud, which is a set of discrete points in space. The point cloud serves as a digital representation of the environment including target object 220. The point cloud can be useful, for example, for performing perception tasks.

[0068] The aggregation duration engine 210 is responsible for dynamically adjusting the aggregation duration of the radar data 214 collected to generate the point cloud. It utilizes parameters such as range, Doppler, and the angle of the reflection point to determine the optimal aggregation time based on the specific situation. By doing so, the aggregation duration engine 210 minimizes point dispersion, a common problem in temporal aggregation, especially for targets with unknown speeds, such as target object 220. The aggregation duration engine 210 ensures that the radar data 214 collected using radar device 104 is sufficiently dense to improve object detection accuracy while maintaining precision by reducing dispersion. The aggregation duration engine 210 allows for more reliable perception of dynamic objects, enhancing the vehicle's overall sensing capabilities by providing the ability to adjust the aggregation duration based on real-time data (e.g., radar data 214).

[0069] The perception task engine 212 processes radar data 214 to perform various perception tasks, such as object detection, classification, and tracking. It integrates radar data 214 collected by radar device 104 and processed by aggregation duration engine 210 to provide real-time perception of the vehicle 100's environment, including target object 220. The perception task engine 212 is useful for applications such as autonomous driving, where accurate and timely perception is used for efficient and effective navigation. By leveraging advanced algorithms and processing techniques, the perception task engine 212 can interpret complex datasets, such as radar data 214, enabling the vehicle 100 (or its operator) to make informed decisions. According to one or more embodiments, the perception task engine 212 enables the vehicle 100 to navigate autonomously or semi-autonomously through its environment, reducing the need for manual intervention.

[0070] According to one or more embodiments, the perception task engine 212 can be used in conjunction with an autonomous driving system (not shown) to control the autonomous navigation capabilities of vehicle 100, thereby allowing vehicle 100 to navigate relative to detected objects. According to one or more embodiments, the autonomous driving system processes information received from the perception task engine 212 and / or radar data 214 received from radar devices 104 (e.g., lidar devices, camera devices, and GPS devices) to determine the precise position and orientation of vehicle 100. The autonomous driving system then generates control signals as needed to steer, accelerate, or brake the vehicle for safe and efficient navigation. The autonomous driving system ensures that vehicle 100 can autonomously perform complex maneuvers, reducing the need for manual intervention.

[0071] Figure 3A A schematic representation 300 of the reflection point 301 detected by radar device 104 is shown, such as... Figure 2 As shown. Now, referring to system 400 as shown in Figure 4, a more detailed description is provided. Figure 3A In particular, Figure 4AA block diagram of system 400 for dynamic aggregation duration is depicted. System 400 is Figure 2 It is part of the processing system 102.

[0072] Reflection point 301 is detected by radar device 104. Reflection point 301 can be, for example, a point on target object 220, or a point on another object within the environment of vehicle 100. The reflection point has a motion vector 302 with velocity v, which indicates the direction and magnitude of motion of reflection point 301 relative to radar device 104 (e.g., relative to vehicle 100 including radar device 104). More specifically, motion vector 302 with velocity v represents the positional offset over aggregation duration T. Angle θ and heading α of reflection point 301 are used to derive velocity v and radial velocity v of reflection point 301. r (Doppler) (not shown). The aggregation time T is also defined, which is the duration for collecting radar data 214. Radial velocity v r Represented as:

[0073] v r =ucos(θ+α),

[0074] And the velocity v can be the radial velocity v r The export is shown below:

[0075]

[0076] These parameters are determined by Figure 2 The aggregation duration engine 210 in the database is used to dynamically adjust the aggregation duration, thereby minimizing point scattering and enhancing object detection accuracy. See now for reference. Figure 4A System 400 provides a more detailed description of determining the aggregation duration.

[0077] refer to Figure 4A Radar reflection point 402 is collected by radar device 104 and fed into frame 404. At frame 404, the maximum aggregation time T... max 406 defines the time window and maximum allowable offset D. 408 is used to identify the maximum aggregation time T. max The reflection point within the time window defined by 406 and the maximum allowable offset D408. That is, for aggregation up to time t, it is determined by [tT]. max Reflection points within the time window defined by :t (e.g., from radar reflection point 402) are aggregated to produce output 410. Output 410 defines the angle θ and velocity v of the detection point. r (Doppler) and the time offset T from t are as follows:

[0078]

[0079] Where i represents [tT] max The discrete time within the range [0…N] defined by the time window :t].

[0080] Aggregation duration engine 210 determines the dynamic aggregation duration. According to one or more embodiments, the dynamic aggregation duration is determined by determining a position offset, setting limits on the position offset (e.g., a maximum distance, such as 1 meter, 2 meters, etc.), deriving a finite aggregation time based on the limits, and deriving the dynamic aggregation duration based on the finite aggregation time. According to one or more embodiments, the dynamic aggregation duration is determined by using the average value of the reflection point heading α. For example, aggregation duration engine 210 applies a probability distribution, such as a Laplace distribution or other suitable distribution, to the reflection point heading α.

[0081] According to one or more embodiments, the positional offset of the aggregation is calculated using the following formula:

[0082]

[0083] Where v r θ is the velocity of reflection point 301, T is the convergence time, θ is the angle of reflection point 301 relative to the reflection point of radar device 104, and α is the heading of reflection point 301 relative to the reflection point of radar device 104.

[0084] According to one or more embodiments, the limits for positional offset are determined using the following formula:

[0085]

[0086] Where v r θ is the velocity of the reflection point, T is the convergence time, θ is the angle of the reflection point relative to the reflection point of the radar device, α is the heading of the reflection point relative to the reflection point of the radar device, and D is the limit.

[0087] According to one or more embodiments, the finite aggregation time is derived using the following formula:

[0088]

[0089] Where v r θ is the velocity of the reflection point, T is the convergence time, θ is the angle of the reflection point relative to the reflection point of the radar device, α is the heading of the reflection point relative to the reflection point of the radar device, and D is the limit.

[0090] According to one or more embodiments, the dynamic aggregation duration is derived using the following formula:

[0091]

[0092] Where T is the aggregation time, and E α It is the expectation of α, D is the limit of the position offset, θ is the angle of the reflection point relative to the reflection point of the radar device, α is the heading of the reflection point relative to the reflection point of the radar device, and v r p is the velocity of the reflection point. r It is the probability distribution function, L1, defined by the following formula:

[0093]

[0094] L2 is defined by the following formula:

[0095]

[0096] At block 412, radar device 104 collects points that satisfy a dynamic aggregation duration (e.g., points falling within a period defined by the dynamic aggregation duration). At block 414, processing system 102 can perform processing on the collected points that satisfy the dynamic aggregation duration to compensate for motion, including ego velocity (e.g., the velocity vector of a vehicle (e.g., vehicle 100) on which radar device 104 is mounted) and / or radial velocity, to generate an aggregated point cloud at block 416. This process is... Figure 2 The aggregation duration engine 210 and perception task engine 212 manage and ensure that accurate perception tasks are performed using radar data 214.

[0097] According to one or more embodiments, the dynamic aggregation duration can be used in conjunction with distance compensation to compensate for distance offset using Doppler. Now refer to Figure 3B and 4B Such an embodiment is described. Specifically, Figure 3B A schematic representation 320 of the reflection point 321 detected by radar device 104 is shown, such as... Figure 2 As shown. Now, referring to system 420 as shown in Figure 4, a more detailed description is provided. Figure 3B In particular, Figure 4B A block diagram is depicted for a system 420 with range-compensated dynamic aggregation duration, which is... Figure 2 It is part of the processing system 102.

[0098] Reflection point 321 is detected by radar device 104. Reflection point 321 can be, for example, a point on target object 220, or a point on another object within the environment of vehicle 100. The reflection point has a motion vector 322 with velocity v, which indicates the direction and magnitude of motion of reflection point 321 relative to radar device 104 (e.g., relative to vehicle 100 including radar device 104). More specifically, the motion vector 322 with velocity v represents the positional offset over a convergence duration T. The angle θ and heading α of reflection point 301 are used to derive the velocity v and radial velocity v of reflection point 301. r (Doppler). The aggregation time T is also defined, which is the duration for collecting radar data 214. Radial velocity v r Represented as:

[0099] v r =vcos(θ+α),

[0100] And the velocity v can be the radial velocity v r The export is shown below:

[0101]

[0102] Marker 320 also shows the representation of v r The radial offset compensation vector 323 and arc 324 performed by T, where arc 324 represents the motion vector 322 with velocity v and the vector passing through v r The radial offset compensation performed by T is the difference between 323 and T.

[0103] These parameters are determined by Figure 2 The aggregation duration engine 210 in the database is used to dynamically adjust the aggregation duration, thereby minimizing point scattering and enhancing object detection accuracy. See now for reference. Figure 4B System 420 describes the determination of the aggregation duration in more detail.

[0104] At box 422, the aggregation duration engine 210 determines the dynamic aggregation duration. According to one or more embodiments, the dynamic aggregation duration is determined by determining a position offset, setting limits on the position offset (e.g., a maximum distance, such as 1 meter, 2 meters, etc.), deriving a finite aggregation time based on the limits, and deriving the dynamic aggregation duration based on the finite aggregation time. According to one or more embodiments, the dynamic aggregation duration is determined by using the average value of the reflection point heading α. For example, the aggregation duration engine 210 applies a probability distribution to the reflection point heading α, such as a Laplace distribution or other suitable distribution.

[0105] According to one or more embodiments, the positional offset of the aggregation is calculated using the following formula:

[0106]

[0107] Where v r θ is the velocity of the reflection point, T is the convergence time, θ is the angle of the reflection point relative to the reflection point of the radar device, and α is the heading of the reflection point relative to the reflection point of the radar device.

[0108] According to one or more embodiments, the limits for positional offset are determined using the following formula:

[0109] T|v r |tan(θ+α) <D

[0110] Where v r θ is the velocity of the reflection point, T is the convergence time, θ is the angle of the reflection point relative to the reflection point of the radar device, α is the heading of the reflection point relative to the reflection point of the radar device, and D is the limit.

[0111] According to one or more embodiments, the finite aggregation time is derived using the following formula:

[0112]

[0113] Where v r θ is the velocity of the reflection point, T is the convergence time, θ is the angle of the reflection point relative to the reflection point of the radar device, α is the heading of the reflection point relative to the reflection point of the radar device, and D is the limit.

[0114] According to one or more embodiments, the dynamic aggregation duration is derived using the following formula:

[0115]

[0116] Where T is the aggregation time, and E α It is the expectation of α, D is the limit of the position offset, θ is the angle of the reflection point relative to the reflection point of the radar device, α is the heading of the reflection point relative to the reflection point of the radar device, and v r It is the velocity of the reflection point, and p r It is a probability distribution function.

[0117] At block 422, radar device 104 collects points that satisfy the dynamic aggregation duration (e.g., points falling within a period defined by the dynamic aggregation duration). At block 424, processing system 102 can perform processing on the collected points that satisfy the dynamic aggregation duration to compensate for motion including ego velocity and / or radial velocity, thereby at block 426 in [tT max Aggregation points are generated within the time window defined by :t]. This process is performed by Figure 2The aggregation duration engine 210 and perception task engine 212 manage and ensure that accurate perception tasks are performed using radar data 214.

[0118] Figure 5 A flowchart of a method 500 for dynamically aggregating the duration of a vehicle-based radar point cloud, according to one or more embodiments, is shown. Method 500 can be implemented using any suitable system or device. For example, method 500 and its steps can be implemented using... Figure 1 and Figure 2 The processing system 102 is composed of Figure 6 This is implemented using processing systems such as 600 (including combinations thereof and / or multiple thereof). Now refer to... Figure 1-4B The method is described in at least part of the 500, but is not limited thereto.

[0119] At box 502, the method begins by using the radar device 104 of vehicle 100 to detect a reflection point (e.g., reflection point 301) on a target object (e.g., target object 220), as... Figure 2 As shown. Radar device 104 collects radar data 214, which is processed by processing system 102.

[0120] At box 504, the dynamic aggregation duration is determined. The dynamic aggregation duration is used to aggregate radar data from reflectors. It is a time period specific to a reflector and is based at least in part on the distribution of the reflector's radial velocity, reflector angle, and heading. The dynamic aggregation duration optimizes data collection, thereby minimizing point dispersion.

[0121] At box 506, radar data for the target object is aggregated within a time period specific to the reflection point, defined by the dynamic aggregation duration. Aggregated radar data is generated as a result of the aggregation at box 506.

[0122] Finally, at box 508, aggregated radar data (e.g., radar data 214 aggregated within a time period defined by the dynamic aggregation duration) is used to perform perception tasks. More specifically, aggregated radar data of target objects aggregated within a time period specific to a reflection point as defined by the dynamic aggregation duration is used to perform perception tasks. For example, perception tasks performed by perception task engine 212 involve processing radar data collected during the dynamic aggregation duration to detect, classify, and track objects in the environment of vehicle 100. These tasks are useful for providing real-time perception, enabling vehicle 100 to make informed decisions. For example, in autonomous driving, perception tasks help identify obstacles, road signs, and other vehicles, thereby allowing for safe navigation. Perception task engine 212 integrates data collected by radar device 104 and processed by aggregation duration engine 210 to enhance the accuracy and reliability of these tasks.

[0123] According to one or more embodiments, method 500 includes an autonomous vehicle, at least in part based on the results of a perception task. An autonomous vehicle 100 means that the vehicle 100 operates without human intervention (or with limited human interaction), using its systems to navigate and make driving decisions (e.g., turning, merging, accelerating, decelerating, etc., including combinations thereof and / or multiples thereof). This involves using data from sensors (such as radar device 104) to detect and respond to the environment of the vehicle 100, thereby ensuring safe and efficient driving. The vehicle 100 may perform tasks such as steering, acceleration, and braking based on real-time perception and analysis.

[0124] Additional processes may also be included, and it should be understood that... Figure 5 The processes described herein are illustrative, and other processes may be added, or existing processes may be removed, modified, or rearranged without departing from the scope of this disclosure. It should also be understood that... Figure 5 The process described herein can be implemented as programming instructions stored on a non-transitory computer-readable storage medium, when executed by a computing system (e.g., Figure 1 and Figure 2 Processing system 102 Figure 6 Processing systems such as 600, including combinations and / or multiples thereof, include processors (e.g., Figure 2 Processing device 202 Figure 6 When a processor 621, etc., including combinations and / or multiples thereof, is executed, the processor performs the process described herein.

[0125] One or more embodiments offer significant technical benefits, including enhanced object detection accuracy by increasing radar point cloud density while minimizing point dispersion. This improvement is achieved through dynamic aggregation duration, which adjusts the aggregation duration based on parameters such as distance, Doppler, and reflector angle. By optimizing radar data collection, one or more embodiments provide rich point clouds, resulting in more reliable perception tasks. This is particularly beneficial for autonomous or semi-autonomous driving, where accurate real-time perception of the environment is crucial for safe navigation. These and other benefits are possible in the various embodiments described herein.

[0126] It should be understood that one or more embodiments described herein can be implemented in conjunction with any other type of computing environment now known or developed in the future. For example, Figure 6 A block diagram of a processing system 600 for implementing the techniques described herein is depicted. According to one or more embodiments described herein, the processing system 600 is an example of a cloud computing node in a cloud computing environment. In the example, the processing system 600 has one or more central processing units (also referred to as “processors”, “processing resources”, or “processing devices”) 621a, 621b, 621c, etc. (collectively or generally referred to as processor 621 and / or processing device 621). In aspects of this disclosure, each processor 621 may include a Reduced Instruction Set Computer (RISC) microprocessor. The processor 621 is coupled to system memory 622 and / or various other components via a system bus 633. System memory 622 may include one or more temporary and / or permanent memory devices, such as random access memory (RAM) 623, read-only memory (ROM) 624, etc., including combinations and / or multiples thereof. System bus 633 may include a Basic Input / Output System (BIOS) that controls certain basic functions of the processing system 600.

[0127] Further depictions include an input / output (I / O) adapter 627 and a network adapter 626 coupled to the system bus 633. The I / O adapter 627 may be a Small Computer System Interface (SCSI) adapter that communicates with a hard disk 635 and / or storage device 636 or any other similar component. The I / O adapter 627, hard disk 635, and storage device 636 are collectively referred to herein as mass storage 634. An operating system 640 for execution on the processing system 600 may be stored in the mass storage 634. The network adapter 626 interconnects the system bus 633 with an external network 638, enabling the processing system 600 to communicate with other such systems.

[0128] A display (e.g., a display monitor) 639 is connected to the system bus 633 via a display adapter 632, which may include a graphics adapter to improve the performance of graphics-intensive applications and video controllers. In one aspect of this disclosure, adapters 626, 627, and / or 632 may be connected to one or more I / O buses connected to the system bus 633 via an intermediate bus bridge (not shown). Suitable I / O buses for connecting peripheral devices such as hard disk controllers, network adapters, and graphics adapters typically include common protocols such as Peripheral Component Interconnect (PCI). Additional input / output devices are shown connected to the system bus 633 via a user interface adapter 628 and the display adapter 632. A keyboard 629, a mouse 630, and a speaker 631 may be interconnected to the system bus 633 via a user interface adapter 628, which may include, for example, a super I / O chip integrating multiple device adapters into a single integrated circuit.

[0129] In some aspects of this disclosure, the processing system 600 includes a graphics processing unit (GPU) 637. The GPU 637 is a dedicated electronic circuit designed to manipulate and modify memory to accelerate the creation of images in a frame buffer intended for output to a display. Typically, the GPU 637 is highly efficient in manipulating computer graphics and image processing, and has a highly parallel architecture that makes it more efficient than general-purpose CPUs for parallel processing of algorithms handling large blocks of data.

[0130] Therefore, as configured herein, the processing system 600 includes processing power in the form of a processor 621, storage capacity including system memory 622 and mass storage 634, input devices such as a keyboard 625 and a mouse 630, and output capacity including a speaker 631 and a display 639. In some aspects of this disclosure, a portion of the system memory 622 and the mass storage 634 jointly store the operating system 640 to coordinate the functions of the various components shown in the processing system 600.

[0131] The terms “a” and “an” do not indicate a limitation of quantity, but rather that at least one of the referenced items is present. Unless the context clearly indicates otherwise, the term “or” means “and / or”. Throughout the specification, the reference to “aspect” means that a particular element described in connection with that aspect (e.g., a feature, structure, step, or characteristic) is included in at least one aspect described herein and may or may not be present in other aspects. Furthermore, it should be understood that the described elements may be combined in any suitable manner in the aspects.

[0132] When an element, such as a layer, film, region, or substrate, is referred to as being “on” another element, it can be directly on the other element, or there may be intermediate elements present. Conversely, when an element is referred to as being “directly” on another element, there are no intermediate elements present.

[0133] Unless otherwise stated herein, all test standards are the most recent standards in force as of the filing date of this application, or, if priority is claimed, the filing date of the earliest priority application in which a test standard appears.

[0134] Unless otherwise defined, the technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure pertains.

[0135] While the foregoing disclosure has been described with reference to exemplary embodiments, those skilled in the art will understand that various changes can be made and elements can be substituted with equivalents without departing from its scope. Furthermore, many modifications can be made to adapt particular situations or materials to the teachings of this disclosure without departing from the basic scope of this disclosure. Therefore, it is intended that this disclosure be limited to the specific embodiments disclosed, but will include all embodiments falling within its scope.

Claims

1. A computer-implemented method, comprising: Use the vehicle's radar device to detect reflection points on the target object; A dynamic aggregation duration is determined for aggregating radar data of the reflection point, the dynamic aggregation duration being a time period specific to the reflection point, wherein the dynamic aggregation duration is based at least in part on the radial velocity of the reflection point, the reflection point angle of the reflection point, and the heading distribution of the reflection point; Within the time period specific to the reflection point, defined by the dynamic aggregation duration, the radar data of the target object is aggregated as aggregated radar data; and The perception task is performed using the aggregated radar data of the target object aggregated within the time period specific to the reflection point, as defined by the dynamic aggregation duration.

2. The computer-implemented method of claim 1, further comprising at least in part based on the results of the perception task derived from the driver of the vehicle.

3. The computer-implemented method according to claim 1, wherein determining the duration of the dynamic aggregation includes: Determine the position offset, which is the offset between the actual position of the reflection point when it is moved from a past time point to the current time point with the actual velocity vector and the predicted position of the reflection point when it is moved according to the radial velocity of the reflection point; Set limits for the position offset; Based on the limit and depending on the radial velocity of the reflection point, the reflection point angle, and the heading distribution, a finite aggregation time for each reflection point is derived. and The aggregate duration of each reflection point is derived based on its limit.

4. The computer-implemented method of claim 3, wherein the dynamic aggregation duration is based on a Laplace distribution, the Laplace distribution being the heading distribution of the reflection point.

5. The computer-implemented method according to claim 3, wherein, The position offset is determined using the following formula: Where v r θ is the velocity of the reflection point, T is the convergence time, θ is the angle of the reflection point relative to the reflection point of the radar device, and α is the heading of the reflection point relative to the reflection point of the radar device.

6. The computer-implemented method according to claim 3, wherein, The position offset is determined using the following formula: Where v r θ is the velocity of the reflection point, T is the convergence time, θ is the angle of the reflection point relative to the reflection point of the radar device, and α is the heading of the reflection point relative to the reflection point of the radar device.

7. The computer-implemented method according to claim 3, wherein, The limits of the position offset are determined using the following formula: Where v r θ is the velocity of the reflection point, T is the convergence time, θ is the angle of the reflection point relative to the reflection point of the radar device, α is the heading of the reflection point relative to the reflection point of the radar device, and D is the limit.

8. The computer-implemented method according to claim 3, wherein, The limits of the position offset are determined using the following formula: T|v r |tan(θ+α)<D Where v r θ is the velocity of the reflection point, T is the convergence time, θ is the angle of the reflection point relative to the reflection point of the radar device, α is the heading of the reflection point relative to the reflection point of the radar device, and D is the limit.

9. The computer-implemented method of claim 3, wherein the finite aggregation time is derived using the following formula: Where v r θ is the velocity of the reflection point, T is the convergence time, θ is the angle of the reflection point relative to the reflection point of the radar device, α is the heading of the reflection point relative to the reflection point of the radar device, and D is the limit.

10. The computer-implemented method of claim 3, wherein the finite aggregation time is derived using the following formula: Where v r Where is the velocity of the reflection point, T is the convergence time, θ is the angle of the reflection point relative to the reflection point of the radar device, α is the heading of the reflection point relative to the reflection point of the radar device, and D is the limit. The dynamic aggregation duration is derived using the following formula: Where T is the aggregation time, and E α It is the expectation of α, D is the limit of the position offset, θ is the angle of the reflection point relative to the reflection point of the radar device, α is the heading of the reflection point relative to the reflection point of the radar device, and v r p is the velocity of the reflection point. r It is the probability distribution function, L1, defined by the following formula: L2 is defined by the following formula:

Images