Lane information display methods, devices and electronic equipment

Abstract

This application relates to the field of intelligent transportation and provides a lane information prompting method, device, and electronic device, including: acquiring the location information of first targets on each lane and the centerline information of each lane; determining the probability that each first target belongs to each lane based on the location information of each first target and the centerline information of each lane; determining the basic risk of each first target; for each lane, selecting second targets from the first targets based on the probability that each first target belongs to the lane and the corresponding basic risk of the first target; and determining the elements of the probabilistic hazard set of the lane based on the identifier of each second target, the probability that the corresponding second target belongs to the lane, and the corresponding basic risk of the second target. This method helps reduce the probability of missed risk warnings due to ignoring the risk of a certain lane.

Description

Technical Field

[0001] This application belongs to the field of intelligent transportation technology, and in particular relates to lane information prompting methods, devices, electronic devices, computer-readable storage media and computer program products. Background Technology

[0002] Intelligent transportation, also known as smart transportation, is a data-driven system that integrates technologies such as the Internet of Things, big data, artificial intelligence (AI), 5G, and vehicle-road cooperation to digitize and network all elements of "people, vehicles, roads, cloud, and network," ultimately achieving a modern integrated transportation system with real-time perception, intelligent decision-making, dynamic collaboration, and safe and efficient operation.

[0003] Lane assignment is a fundamental aspect of intelligent transportation systems. Existing technical solutions generally employ a "hard assignment" strategy, the core logic of which is to uniquely assign a target to a specific lane. Specifically, lane boundaries are first determined through lane line detection, and then the target is rigidly assigned to a single lane based on whether its geometric center is within the boundary.

[0004] However, this deterministic allocation method has obvious flaws in real-world scenarios and directly raises the following key issues: Risk underestimation is a problem. Large vehicles crossing lanes or vehicles cutting into lanes pose a safety threat to two adjacent lanes simultaneously. However, a hard-assignment strategy forcibly assigns them to a specific lane, inevitably leading to the complete neglect of the risk in the other lane, resulting in missed risk warnings and failing to accurately and comprehensively reflect the overall safety situation of multi-lane coordination. Summary of the Invention

[0005] This application provides a lane information prompting method, device, and electronic device, which can solve the problem of missed risk warnings caused by neglecting the risk of a certain lane in existing methods.

[0006] In a first aspect, embodiments of this application provide a lane information prompting method, including: Obtain the location information of the first target in each lane and the centerline information of each lane; Based on the location information of each first target and the centerline information of each lane, the probability of each first target belonging to each lane is determined respectively. Determine the basic risks for each of the first objectives, wherein the basic risks are the inherent risks of the first objectives; For each lane, based on the probability of each first target belonging to the lane and the basic risk of the corresponding first target, a second target is selected from each first target, wherein the second target is a first target whose risk is greater than a preset first risk threshold; based on the identifier of each second target, the probability of the corresponding second target belonging to the lane and the basic risk of the corresponding second target, the elements of the probability hazard set of the lane are determined.

[0007] The beneficial effects of the embodiments of this application compared with the prior art are: In this embodiment, based on the acquired location information of each first target and the centerline information of each lane, the probability of each first target belonging to each lane is determined. After determining the basic risk of each first target, for each lane, based on the probability of each first target belonging to the lane and the corresponding basic risk of the first target, second targets are selected from the first targets. Then, based on the identifier of the second target, the probability of the second target belonging to the lane, and the basic risk of the second target, the elements of the probabilistic risk set of the lane are determined. Since the second target is a first target with a risk greater than a preset first risk threshold, and the basic risk is the inherent risk of the second target, when the elements of the probabilistic risk set of a lane are determined based on the probability of the second target belonging to that lane, the identifier of the second target, and the basic risk of the second target, the elements of the probabilistic risk set of that lane can not only reflect the probability of the second target (i.e., the first target with a risk greater than the preset first risk threshold) belonging to that lane, but also reflect the inherent risk of the second target. Correspondingly, the elements of the probabilistic risk set of each lane can reflect the probability of each second target belonging to each lane and the inherent risk of each second target. That is, since it is possible to determine the probability of each secondary target belonging to each lane and the corresponding risk, it is possible to reduce the probability of missed risk warnings caused by ignoring the risk of a certain lane, thereby enabling a true and comprehensive reflection of the overall safety situation of multi-lane cooperation.

[0008] Optionally, the lane information prompting method provided in this application embodiment further includes: Assess the confidence level of the existence of each lane; The step of determining the probability that each first target belongs to each lane based on the location information of each first target and the centerline information of each lane includes: Based on the location information of each first target, the centerline information of each lane, and the confidence level of the existence of each lane, the probability of each first target belonging to each lane is determined.

[0009] Optionally, determining the probability that each first target belongs to each lane based on the location information of each first target and the centerline information of each lane includes: Based on the position information of each first target in the current frame and the center position information of each lane in the current frame, the update probability of each first target belonging to each lane in the current frame is determined respectively. For each of the first targets: based on the update probability of the first target belonging to each of the lanes in the current frame, and based on the probability of the first target belonging to each of the lanes in the previous frame of the current frame, determine the probability of the first target belonging to each of the lanes in the current frame.

[0010] Optionally, the step of selecting a second target from each of the first targets based on the probability that each of the first targets belongs to the lane and the basic risk of the first target includes: For each lane, a corresponding weighted risk value is determined based on the probability that each first target belongs to the lane and the basic risk of the corresponding first target. Each weighted risk value is compared with a preset first risk threshold, and the first target corresponding to the weighted risk value that is greater than the first risk threshold is determined as the second target.

[0011] Optionally, after determining the elements of the probabilistic hazard set of the lane, the method further includes: Remove the elements corresponding to the weighted risk values ​​in the probability hazard set that are less than a preset second risk threshold, where the preset second risk threshold is less than the first risk threshold.

[0012] Optionally, the lane information prompting method provided in this application embodiment further includes: For each lane, the expected risk of the lane is determined based on the probability that each first target belongs to the lane and the basic risk of each first target; Output the expected risk of the lane.

[0013] Optionally, after determining the expected risk of the lane, the method further includes: For each lane, a risk upper bound for the lane is determined based on the lane's expected risk, the probability that each of the first targets belongs to the lane, and the basic risk of each of the first targets. Output the upper risk bound of the lane.

[0014] Secondly, embodiments of this application provide a lane information prompting device, comprising: The target information and lane information acquisition module is used to acquire the position information of the first target on each lane and the centerline information of each lane; The lane affiliation probability determination module is used to determine the probability that each first target belongs to each lane based on the location information of each first target and the centerline information of each lane. The inherent risk determination module is used to determine the basic risk of each of the first targets, wherein the basic risk is the inherent risk of the first target; The lane probability hazard set determination module is used to, for each lane, filter out a second target from each of the first targets based on the probability of each first target belonging to the lane and the basic risk of the corresponding first target, wherein the second target is a first target whose risk is greater than a preset first risk threshold; and determine the elements of the lane probability hazard set based on the identifier of each second target, the probability of the corresponding second target belonging to the lane and the basic risk of the corresponding second target.

[0015] Thirdly, embodiments of this application provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method described in the first aspect.

[0016] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method described in the first aspect.

[0017] Fifthly, embodiments of this application provide a computer program product that, when run on an electronic device, causes the electronic device to perform the method described in the first aspect.

[0018] It is understood that the beneficial effects of the second to fifth aspects mentioned above can be found in the relevant descriptions in the first aspect mentioned above, and will not be repeated here. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below.

[0020] Figure 1 This is a flowchart illustrating a lane information prompting method provided in an embodiment of this application; Figure 2 This is a schematic diagram of the structure of a lane information prompting device provided in an embodiment of this application; Figure 3This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0021] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0022] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0023] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0024] Furthermore, in the description of this application and the appended claims, the terms "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0025] References to "one embodiment" or "some embodiments" in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized.

[0026] In intelligent transportation systems (such as autonomous driving and intelligent traffic signal control), it is usually necessary to determine the lane to which a moving target (such as a vehicle) belongs.

[0027] If a "hard attribution" strategy is used to determine the lane to which a moving target belongs, it will lead to the problem of missed risk assessment.

[0028] To reduce the probability of missed risk detection, this application provides a lane information prompting method.

[0029] The lane information prompting method determines the elements of the lane's probability risk set by calculating the probability that a first target belongs to the lane and combining the identifier of the first target with the basic risk of the first target.

[0030] The lane information prompting method provided in the embodiments of this application is described below with reference to the accompanying drawings.

[0031] Figure 1 This illustration shows a flowchart of a lane information prompting method according to an embodiment of this application. This method can be applied to electronic devices, including in-vehicle devices, desktop computers, laptops, etc. The lane information prompting method provided in this embodiment is described in detail below: S11, obtain the position information of the first target in each lane and the centerline information of each of the aforementioned lanes.

[0032] The first target mentioned above is an object moving at a speed in the lane. For example, the first target includes vehicles, running animals, and so on.

[0033] In this embodiment, the location information of the first target can be determined in the following way: if a positioning device is provided on the first target, the electronic device can obtain the location information of the first target by interacting with the positioning device. Of course, the electronic device can also determine the relative position of the first target and a target with known location information by one or more sensors such as visual sensors, millimeter-wave radar, and lidar, and then combine the location information of the target with the known location information to obtain the location information of the first target; no limitation is made here.

[0034] In this embodiment of the application, the centerline information of the lane can be obtained in the following ways: from a high-precision map containing the coordinates of the lane centerline, from data collected by vehicle-mounted sensors (such as visual sensors, lidar, etc.), from roadside perception devices, etc., and so on. No limitation is made here.

[0035] S12, based on the location information of each of the aforementioned first targets and the centerline information of each of the aforementioned lanes, determine the probability that each of the aforementioned first targets belongs to each of the aforementioned lanes.

[0036] In this embodiment of the application, considering that the lateral offset of the first target relative to the center line of the lane implies the movement intention of the first target, an unnormalized belonging likelihood probability can be constructed based on the lateral offset of the first target relative to the center line of the lane, and then the unnormalized belonging likelihood probability can be normalized to obtain the probability that the first target belongs to the lane.

[0037] In some embodiments, considering that the lane identification results obtained when lanes are identified by a perception algorithm may be erroneous. For example, when detecting lanes using an online lane line detection algorithm, online lane line detection is prone to misidentifying non-lane lines (such as guardrails, shadows, shoulders, and debris) as lane lines in complex scenarios (such as backlighting, rain, obstructions, and severely worn markings). Therefore, directly using these misdetected lane lines will incorrectly assign the first target to a non-existent lane, resulting in a serious deviation in the first target attribution determination. Therefore, when determining the probability that the first target belongs to a lane, the confidence level of the lane's existence can also be considered. That is, in some embodiments, the lane information prompting method provided in this application further includes: Assess the confidence level of the existence of each of the aforementioned lanes.

[0038] Correspondingly, the above-mentioned determination of the probability that each of the first targets belongs to each of the lanes based on the location information of each of the first targets and the centerline information of each of the lanes includes: Based on the location information of each of the aforementioned first targets, the centerline information of each of the aforementioned lanes, and the confidence level of the existence of each of the aforementioned lanes, the probability of each of the aforementioned first targets belonging to each of the aforementioned lanes is determined.

[0039] In this embodiment, the confidence level of lane existence can be assessed based on the source of the lane; that is, the higher the reliability of the source, the higher the confidence level of lane existence, and vice versa. For example, if the lane source is a high-precision map, the confidence level of lane existence from the high-precision map can be set to a number close to 1 (i.e., the difference between the lane existence confidence level and 1 is less than a preset value, which is a number less than 1 and close to 0, such as 0.05). As another example, if the lane source is online lane line detection, the confidence level of lane existence can be determined based on the lane line detection confidence level or continuity score to suppress false detections and flickering.

[0040] After determining the confidence level of lane existence, an unnormalized assignment likelihood probability is constructed based on the confidence level of lane existence and the lateral offset of the first target relative to the lane centerline. Then, the unnormalized assignment likelihood probability is normalized to obtain the probability that the first target belongs to the lane.

[0041] In this embodiment of the application, considering that the obtained position information of the first target usually has a certain error, the lateral offset of the first target relative to the lane centerline calculated based on the position information of the first target usually also has a certain error. Therefore, when constructing the unnormalized belonging likelihood probability, the error of the lateral offset can also be considered.

[0042] Optionally, assuming there are multiple first targets and multiple lanes, then for the first target and lane Calculate the first target Relative to this lane Lateral offset of the centerline (To avoid conflict with the vertical axis) Obfuscation, used here (representing the horizontal direction), then the constructed unnormalized attribution likelihood probability can be as follows: .

[0043] in: This represents the unnormalized likelihood of belonging (dimensionless).

[0044] lane The existence confidence level (dimensionless, value range [0,1], used to suppress interference from false detection lanes).

[0045] Indicates the first objective Center to Lane The horizontal distance of the centerline can be set in meters.

[0046] Indicates the first objective The lateral offset error, in meters, can be set as follows. Optionally, It can be set to half the vehicle width, such as 1.0 meter.

[0047] right After normalization, the resulting value can be used as the primary objective. Belonging to the lane The probability of.

[0048] In some embodiments, when using unnormalized assignment likelihoods for normalization, if the assignment likelihoods for all lanes are extremely low (e.g., due to extreme weather, lane line completion loss, false detections, etc.), both the denominator and numerator will approach 0. Since dividing by 0 is mathematically meaningless, when both the denominator and numerator approach 0, the calculation result will be infinite or non-numerical, potentially causing system crashes. Therefore, a smoothing constant can be set during normalization. That is: .

[0049] in: Indicates the first objective Belonging to the lane The probability is the normalized value. This represents the smoothing constant (dimensionless, typical value can be set to...). ).

[0050] Because a smoothing constant is added during normalization calculations, the probability distribution can be made more accurate even when the likelihood of all lanes is low. It does not degenerate, but rather becomes nearly uniformly distributed. This treatment enhances numerical stability in extreme cases, such as when lane lines are completely lost.

[0051] In some embodiments, considering that the first target is unlikely to change from one lane to another in a very short time, and the first target observed in a single frame Belonging to the lane The probability of a target being assigned to a lane is easily affected by sensor noise, occlusion, false detection, etc., and may fluctuate drastically (e.g., a first target is suddenly assigned to the wrong lane in a single frame). Therefore, to suppress random noise in single-frame observations, the posterior probability of the previous frame can be used as the prior of the current frame. That is, the probability of each first target belonging to its respective lane is determined based on the position information of each first target and the centerline information of each lane, including: Based on the position information of each of the aforementioned first targets in the current frame and the center position information of each of the aforementioned lanes in the current frame, the update probability of each of the aforementioned first targets belonging to each of the aforementioned lanes in the current frame is determined respectively. For each of the aforementioned first targets: based on the pending update probability of the aforementioned first target belonging to each of the aforementioned lanes in the current frame, and based on the probability of the aforementioned first target belonging to each of the aforementioned lanes in the previous frame of the current frame, the probability of the aforementioned first target belonging to each of the aforementioned lanes in the current frame is determined.

[0052] In this embodiment of the application, when considering historical results, the normalized assignment likelihood calculated based on the position information of each of the aforementioned first targets in the current frame and the center position information of each of the aforementioned lanes in the current frame can be used as the update probability for each of the first targets to be assigned to each lane in the current frame. Alternatively, the normalized assignment likelihood calculated by adding the confidence level of lane existence and / or the smoothing constant can be used as the update probability for each of the first targets to be assigned to each lane in the current frame. The calculation method of the normalized assignment likelihood is detailed in the description above and will not be repeated here.

[0053] After obtaining the updateable probabilities of the first target belonging to each lane in the current frame, the final probability of the first target belonging to each lane in the current frame is determined by using the posterior probability of the first target belonging to a lane, which is proportional to the weighted product of the observed probability in the current frame and the posterior probability in the previous frame. Optionally, assume the current frame is frame [frame number missing]. The first goal In frame Belonging to the lane The probability of needing to be updated is (i.e., the observation probability of the current frame), the first target In frame ( Belonging to the lane The probability (i.e., the posterior probability of the previous frame) is The first goal In frame Belonging to the lane The probability is normalized ,and .

[0054] in, This is the inertia factor (dimensionless, typically set to 0.7). Wherein, The larger the value, the stronger the influence of historical probability, and the smoother the output. Conversely, The smaller the value, the faster the system responds to the current observation. This update is equivalent to exponential smoothing in the logarithmic domain. After calculation, it is necessary to... Normalize to make Thus, the first objective was achieved. In frame Belonging to the lane The final probability.

[0055] In this embodiment of the application, since the calculation of the final probability of the first target belonging to each lane in the current frame takes into account not only the current observation but also the historical results, the random noise in the single-frame observation is suppressed, thereby improving the accuracy of the final probability.

[0056] Optionally, considering direct use As a frame When the prior probability is used, it is equivalent to defaulting to the first target. Stay in the original lane Assuming the first objective is... If a lane change occurs rapidly in the current frame, the current frame observation will just be pointing to the new lane, but the previous frame observation will still be pointing to the original lane. And the probability is very high, so, according to Determine the primary objective Belonging to the lane When the probability is high, historical relationships can cause the probability of belonging to a new lane to increase slowly, resulting in a probability lag. Therefore, this application introduces a lane transition matrix to correct for state transitions: .

[0057] Among them, the above-mentioned This is the lane transfer matrix (a dimensionless probability matrix, typically diagonally dominant, but allowing for a small probability of transfer to an adjacent lane), used to model the lateral movement tendency of vehicles. replace In This can significantly improve the response speed when changing lanes.

[0058] S13, determine the basic risks of each of the aforementioned first objectives, wherein the aforementioned basic risks are the inherent risks of the aforementioned first objectives.

[0059] The fundamental risk of the first objective mentioned above refers to the inherent and stable danger of the first objective itself. Taking a vehicle as an example, the fundamental risk of a vehicle is determined by its own motion state, behavioral intentions, vehicle condition, and vehicle type. It is unrelated to the current lane, temporary noise, or observation errors; it is an inherent and stable danger of the vehicle itself.

[0060] In this embodiment, the basic risk of the first target can be calculated in the following way: selecting features that reflect the stability of the first target itself; normalizing each feature; setting weights for different features and summing them by weight; and limiting the result of the weighted summation.

[0061] When the primary target is a vehicle, features reflecting the stability of the primary target itself are selected, which may include: vehicle speed, lateral acceleration (reflecting whether the vehicle makes sharp turns or swerves), rate of change of heading angle (reflecting whether the vehicle's direction of travel is stable), degree of lateral position fluctuation (reflecting whether the vehicle frequently deviates from its lane), and vehicle type (such as large vehicles and non-motorized vehicles, which inherently pose a higher risk). Each feature is normalized, which may include mapping each feature value to a range of 0 to 1. A larger value indicates a higher risk associated with that feature; for example, a faster speed results in a higher risk after normalization. The larger the value, the better. Weights are assigned to different features and a weighted sum is calculated. This can include: assigning weights based on the degree of influence of each feature on risk; these weights reflect the importance of different factors to vehicle stability. For example, lateral acceleration and heading fluctuations have a greater impact on risk and can be assigned higher weights. All normalized features are then weighted and summed to obtain a comprehensive basic risk value. The weighted sum is then limited, for example, by restricting the weighted sum to the range of 0 to 1, to ensure that the final output basic risk value is within a reasonable range, facilitating subsequent integration with collision risk, lane risk, etc.

[0062] S14, for each of the above lanes, based on the probability that each of the above first targets belongs to the above lane and the basic risk of the corresponding first target, a second target is selected from each of the above first targets, the second target being the above first target whose risk is greater than a preset first risk threshold; based on the identifier of each of the above second targets, the probability that the corresponding second target belongs to the above lane and the basic risk of the corresponding second target, the elements of the probability hazard set of the above lane are determined.

[0063] In this embodiment of the application, the identification of the second target can be represented by the location information of the second target. Of course, it can also be represented by a combination of road segment, lane and linear offset. There is no limitation here.

[0064] In this embodiment of the application, a probabilistic hazard set is maintained for each lane. The elements of the probabilistic hazard set are determined based on the identifier of the second target, the probability that the second target belongs to the lane, and the basic risk of the second target. The second target is a first target that carries a risk.

[0065] In this embodiment, the second objective is a first objective whose risk exceeds a preset first risk threshold. This second objective can be obtained by filtering from various first objectives in the following manner: For each of the above lanes, a corresponding weighted risk value is determined based on the probability that each of the above first targets belongs to the above lane and the basic risk of the corresponding first target. Each of the aforementioned weighted risk values ​​is compared with a preset first risk threshold, and the first target corresponding to the weighted risk value that is greater than the first risk threshold is determined as the second target.

[0066] In this embodiment of the application, the weighted risk value can be obtained by multiplying the probability of the first target belonging to the lane and the basic risk of the corresponding first target, or by summing the probability of the first target belonging to the lane and the basic risk of the corresponding first target. No limitation is made here.

[0067] After identifying the second target, the identifier of the second target, the probability of the second target belonging to the lane, and the basic risk of the second target can be used as elements of the probabilistic hazard set of the lane. Alternatively, the identifier of the second target, the probability of the second target belonging to the lane, and the weighted risk value of the second target can also be used as elements of the probabilistic hazard set of the lane. No limitation is made here.

[0068] Optionally, after generating the probabilistic hazard set, the probabilistic hazard set can be output, or the probabilistic hazard set can be output after certain conditions are met.

[0069] Optionally, after generating the elements of the probabilistic hazard set, to further prevent the second target from jumping back and forth between the probabilistic hazard set and the non-probabilistic hazard set, a dual-threshold mechanism can be adopted. That is, a second risk threshold is set, which is less than the first risk threshold. If it is determined that the weighted risk value corresponding to an element in the probabilistic hazard set is less than the second risk threshold, then the element is removed from the probabilistic hazard set. In this case, after determining the elements of the probabilistic hazard set of the aforementioned lane, the following steps are also included: Remove the elements corresponding to the weighted risk values ​​of the above-mentioned probability risk set that are less than the preset second risk threshold.

[0070] In this embodiment, the weighted risk value of an element in the probabilistic hazard set is compared with a preset second risk threshold. If the weighted risk value is determined to be less than the second risk threshold, the element is removed. The element is then added to the non-probabilistic hazard set. Since the second risk threshold is less than the first risk threshold, when the weighted risk value is less than the first risk threshold but still greater than the second risk threshold, the element corresponding to the weighted risk value remains in the probabilistic hazard set, thereby preventing the second target from jumping back and forth between the probabilistic and non-probabilistic hazard sets.

[0071] The following example illustrates how to select a second target from a first target, generate elements of the probabilistic hazard set for that lane based on the second target's identifier, the probability that the second target belongs to the lane, and the weighted risk value of the second target, and how to remove elements from the probabilistic hazard set.

[0072] Assuming the first goal Belonging to the lane The probability is adopted This indicates that the primary objective is... Basic risk adopts This indicates that the primary objective is... The weighted risk value adopts This indicates that the preset first risk threshold is adopted. This indicates that the preset second risk threshold is adopted. This indicates that the probability hazard set adopts... Indicate, then .like Then the first target As the second objective The second objective use It is stated that, according to , , Determine the probability hazard set An element If, at a later time, it is determined that the element... Then the element from Remove.

[0073] That is, the probability hazard set The elements in the data satisfy the dual threshold conditions corresponding to the first risk threshold and the second risk threshold: .

[0074] in, .

[0075] In this embodiment, based on the acquired location information of each first target and the centerline information of each lane, the probability of each first target belonging to each lane is determined. After determining the basic risk of each first target, for each lane, based on the probability of each first target belonging to the lane and the corresponding basic risk of the first target, second targets are selected from the first targets. Then, based on the identifier of the second target, the probability of the second target belonging to the lane, and the basic risk of the second target, the elements of the probabilistic risk set of the lane are determined. Since the second target is a first target with risk, and the basic risk is the inherent risk of the second target, when the elements of the probabilistic risk set of a lane are determined based on the probability of the second target belonging to that lane, the identifier of the second target, and the basic risk of the second target, the elements of the probabilistic risk set of that lane can reflect not only the probability of the second target (i.e., the first target with risk) belonging to that lane, but also the inherent risk of the second target. Correspondingly, the elements of the probabilistic risk set of each lane can reflect the probability of each second target belonging to each lane and the inherent risk of each second target. That is, since it is possible to determine the probability of each secondary target belonging to each lane and the corresponding risk, it is possible to reduce the probability of missed risk warnings caused by ignoring the risk of a certain lane, thereby enabling a true and comprehensive reflection of the overall safety situation of multi-lane cooperation.

[0076] The above description primarily introduced the risk posed to a lane by a single secondary (or primary) objective. The following section will discuss the overall risk posed to a lane by multiple objectives. Optionally, this overall risk includes the expected risk and the upper bound of the risk.

[0077] Assuming lane random risk The definition is as follows: .

[0078] in, Indicates "first goal" In this frame, it is considered to belong to the lane. This is a random event. It should be noted that since a primary target cannot belong to two lanes simultaneously, different lanes... In fact, they are negatively correlated (i.e., mutually exclusive). This applies when calculating single-lane risk. At that time, because only the aggregation within that lane is considered, i.e., different primary targets The attribution between them can usually be approximated as independent (unless they move in groups), therefore, the independence assumption is an acceptable engineering approximation in single-lane risk calculation.

[0079] It is a lane The calculation process for the random risk (a random variable, not a deterministic value) is as follows: For each first objective in the scenario Introduce a Bernoulli random variable The random variable has a probability The value is 1, with a probability ( The value is 0. The first objective obtained in the previous text is calculated using Gaussian likelihood and then smoothed using Bayesian smoothing. Belonging to the lane The posterior probability of the lane. Total risk This is the sum of the basic risk scores of all the first targets drawn and assigned to that lane. Because... It is random, therefore, It is also a random variable, and its expected value is the expected risk. .

[0080] In this embodiment of the application, the lane information prompting method provided in this embodiment of the application further includes: For each of the aforementioned lanes, the expected risk of the lane is determined based on the probability that each of the aforementioned first targets belongs to the aforementioned lane and the basic risk of each of the aforementioned first targets. Output the expected risk of the aforementioned lanes.

[0081] Among them, lane The expected risk level (i.e., average risk) is as follows: .

[0082] In this embodiment of the application, after calculating the expected risk for each lane, the expected risk is output.

[0083] After calculating the expected risk, an upper bound on the risk can also be calculated to address potential risks arising from ownership uncertainty.

[0084] In some embodiments, after determining the expected risk of the lane as described above, the lane information prompting method provided in this application further includes: For each of the aforementioned lanes, the risk upper bound of the lane is determined based on the expected risk of the lane, the probability that each of the aforementioned first targets belongs to the lane, and the basic risk of each of the aforementioned first targets. Output the upper risk limit for the aforementioned lanes.

[0085] In this embodiment of the application, the following two methods can be used to calculate the upper bound of risk.

[0086] Method 1 is an approximation based on variance: .

[0087] in: For lane The upper bound of risk (dimensionless). This is a safety factor (dimensionless, such as 2.0).

[0088] The radical term in the above formula is the fluctuation term in the engineering approximation. Based on the independence assumption, the risk contribution of each first objective is treated as a Bernoulli random variable, and the variance formula is used. Calculate the standard deviation of total risk. According to the formula above, The larger the value, the higher the uncertainty of attribution (i.e., the flatter the distribution).

[0089] It should be noted that, and It is not necessary to strictly limit it to If a normalized value is required downstream, a saturation function (e.g., ...) can be used. Alternatively, you can use cropping and scaling to map it to a fixed range.

[0090] Method 2, Conditional Value at Risk: This method measures the worst-case scenario. The average risk under the following circumstances, where For confidence level, a typical value can be set to 0.95 or 0.99. Let: .

[0091] in, Represents random variables of quantiles.

[0092] To address the potential security risks arising from the uncertainty of target attribution, it is necessary to accurately characterize the extreme tail features of the risk, which requires first constructing risk variables. The complete probability distribution is typically achieved through Monte Carlo sampling in engineering: the system performs a large number of repeated random samplings based on the probability of each target's attribution, statistically obtains the empirical probability distribution of the risk variable, and then calculates the quantiles and conditional expectations of the risk based on this distribution, ultimately obtaining the conditional risk value.

[0093] In this embodiment of the application, the corresponding method can be selected to calculate the risk upper bound according to the actual situation. For example, in scenarios with high security requirements, the second method can be selected, while in scenarios with high real-time requirements, the first method can be selected.

[0094] The aforementioned risk expectation and risk upper bound belong to lane-level risk indicators (the probabilistic hazard set belongs to object-level risk indicators). Optionally, to prevent lane-level prompts from frequently switching within the critical risk range, a third risk threshold and a fourth risk threshold can be preset, with the fourth risk threshold being smaller than the third risk range. If the risk upper bound of a lane is greater than the aforementioned third risk threshold, the risk expectation and risk upper bound of that lane are output to indicate that the lane has a risk. After determining that a lane has a risk, if it is determined that the risk upper bound of that lane is less than the aforementioned fourth risk threshold, the output of the risk expectation and risk upper bound of that lane is stopped.

[0095] In some embodiments, after outputting the risk expectation and risk upper bound of the output lane, the aforementioned probabilistic hazard set can be output to illustrate which targets constitute the lane risk warning.

[0096] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0097] Corresponding to the lane information prompting method described in the above embodiments, Figure 2 This diagram illustrates a structural block diagram of a lane information prompting device according to an embodiment of this application. For ease of explanation, only the parts related to the embodiment of this application are shown.

[0098] Reference Figure 2 The lane information display device 2 is applied to electronic equipment and includes a target information and lane information acquisition module 21, a lane ownership probability determination module 22, a target inherent risk determination module 23, and a lane probability hazard set determination module 24. Among them: The target information and lane information acquisition module 21 is used to acquire the position information of the first target in each lane and the centerline information of each of the aforementioned lanes.

[0099] The lane affiliation probability determination module 22 is used to determine the probability that each of the first targets belongs to each of the lanes based on the location information of each of the first targets and the centerline information of each of the lanes.

[0100] The target inherent risk determination module 23 is used to determine the basic risk of each of the aforementioned first targets, wherein the basic risk is the inherent risk of the aforementioned first targets.

[0101] The lane probability hazard set determination module 24 is used to, for each of the above lanes, select a second target from each of the above first targets based on the probability of each of the above first targets belonging to the above lanes and the basic risk of the corresponding first targets, wherein the second target is the above first target whose risk is greater than a preset first risk threshold; and determine the elements of the probability hazard set of the above lanes based on the identifier of each of the above second targets, the probability of the corresponding second target belonging to the above lanes and the basic risk of the corresponding second target.

[0102] In this embodiment, based on the acquired location information of each first target and the centerline information of each lane, the probability of each first target belonging to each lane is determined. After determining the basic risk of each first target, for each lane, based on the probability of each first target belonging to the lane and the corresponding basic risk of the first target, second targets are selected from the first targets. Then, based on the identifier of the second target, the probability of the second target belonging to the lane, and the basic risk of the second target, the elements of the probabilistic risk set of the lane are determined. Since the second target is a first target with a risk greater than a preset first risk threshold, and the basic risk is the inherent risk of the second target, when the elements of the probabilistic risk set of a lane are determined based on the probability of the second target belonging to that lane, the identifier of the second target, and the basic risk of the second target, the elements of the probabilistic risk set of that lane can not only reflect the probability of the second target (i.e., the first target with a risk greater than the preset first risk threshold) belonging to that lane, but also reflect the inherent risk of the second target. Correspondingly, the elements of the probabilistic risk set of each lane can reflect the probability of each second target belonging to each lane and the inherent risk of each second target. That is, since it is possible to determine the probability of each secondary target belonging to each lane and the corresponding risk, it is possible to reduce the probability of missed risk warnings caused by ignoring the risk of a certain lane, thereby enabling a true and comprehensive reflection of the overall safety situation of multi-lane cooperation.

[0103] Optionally, the lane information prompting method provided in this application embodiment is characterized by further comprising: The lane presence confidence assessment module is used to assess the confidence level of the presence of each of the aforementioned lanes.

[0104] Correspondingly, the lane ownership probability determination module 22 mentioned above is specifically used for: Based on the location information of each of the aforementioned first targets, the centerline information of each of the aforementioned lanes, and the confidence level of the existence of each of the aforementioned lanes, the probability of each of the aforementioned first targets belonging to each of the aforementioned lanes is determined.

[0105] In this embodiment of the application, the confidence level of the existence of the lane can be assessed based on the source of the lane; that is, the higher the reliability of the source, the higher the confidence level of the existence of the lane.

[0106] Optionally, the lane ownership probability determination module 22 mentioned above includes: The lane attribution probability determination unit is used to determine the probability of each of the first targets belonging to each of the lanes in the current frame based on the position information of each of the first targets in the current frame and the center position information of each of the lanes in the current frame. The lane attribution probability update unit is used to, for each of the first targets, determine the probability that the first target belongs to each of the lanes in the current frame based on the probability to be updated that the first target belongs to each of the lanes in the current frame, and based on the probability that the first target belonged to each of the lanes in the previous frame of the current frame.

[0107] Optionally, when the lane probability hazard set determination module 24 selects a second target from each of the first targets based on the probability that each of the first targets belongs to the lane and the basic risk of the first target, it is specifically used for: For each of the above lanes, a corresponding weighted risk value is determined based on the probability that each of the above first targets belongs to the above lane and the basic risk of the corresponding first target. Each of the aforementioned weighted risk values ​​is compared with a preset first risk threshold, and the first target corresponding to the weighted risk value that is greater than the first risk threshold is determined as the second target.

[0108] Optionally, the lane information prompting device 2 provided in this application embodiment further includes: The element removal module is used to remove the elements corresponding to the weighted risk values ​​in the probabilistic hazard set of the lane that are less than a preset second risk threshold after the elements of the probabilistic hazard set of the lane are determined. The preset second risk threshold is less than the first risk threshold.

[0109] Optionally, the lane information prompting device 2 provided in this application embodiment further includes: The risk expectation determination module is used to determine the risk expectation of each of the above lanes based on the probability that each of the above first targets belongs to the above lanes and the basic risk of each of the above first targets. The risk expectation output module is used to output the risk expectation of the above lanes.

[0110] Optionally, the lane information prompting device 2 provided in this application embodiment further includes: The risk upper bound determination module is used to determine the risk upper bound of each lane after determining the risk expectation of the lane, based on the risk expectation of the lane, the probability that each of the first targets belongs to the lane, and the basic risk of each of the first targets. The risk upper bound output module is used to output the risk upper bound of the above lanes.

[0111] It should be noted that the information interaction and execution process between the above-mentioned devices / units are based on the same concept as the method embodiments of this application. For details on their specific functions and technical effects, please refer to the method embodiments section, and they will not be repeated here.

[0112] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 3 As shown, the electronic device 3 of this embodiment includes: at least one processor 30 ( Figure 3 The diagram shows only one processor, memory 31, and computer program 32 stored in the memory 31 and executable on at least one processor 30. When the processor 30 executes the computer program 32, it implements the steps in any of the above method embodiments.

[0113] The aforementioned electronic device 3 can be a desktop computer, laptop, handheld computer, or cloud server, etc. This electronic device may include, but is not limited to, a processor 30 and a memory 31. Those skilled in the art will understand that... Figure 3 This is merely an example of electronic device 3 and does not constitute a limitation on electronic device 3. It may include more or fewer components than shown in the figure, or combine certain components, or different components. For example, it may also include input / output devices, network access devices, etc.

[0114] The processor 30 may be a Central Processing Unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0115] In some embodiments, the aforementioned memory 31 may be an internal storage unit of the electronic device 3, such as a hard disk or memory of the electronic device 3. In other embodiments, the aforementioned memory 31 may be an external storage device of the electronic device 3, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the electronic device 3. Furthermore, the aforementioned memory 31 may include both internal storage units and external storage devices of the electronic device 3. The aforementioned memory 31 is used to store operating systems, applications, bootloaders, data, and other programs, such as the program code of the aforementioned computer programs. The aforementioned memory 31 may also be used to temporarily store data that has been output or will be output.

[0116] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the above device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments 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. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0117] This application also provides a network device, which includes: at least one processor, a memory, and a computer program stored in the memory and executable on the at least one processor, wherein the processor executes the computer program to implement the steps in any of the above method embodiments.

[0118] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, can implement the steps in the above-described method embodiments.

[0119] This application provides a computer program product that, when run on an electronic device, enables the electronic device to implement the steps described in the various method embodiments above.

[0120] 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, all or part of the processes in the methods of the above embodiments of this application can be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include at least: any entity or device capable of carrying computer program code to a photographic device / electronic device, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium. Examples include USB flash drives, portable hard drives, magnetic disks, or optical disks. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electrical carrier signals or telecommunication signals.

[0121] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0122] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0123] In the embodiments provided in this application, it should be understood that the disclosed apparatus / network devices and methods can be implemented in other ways. For example, the apparatus / network device embodiments described above are merely illustrative. For instance, the division of modules or 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; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0124] 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.

Claims

1. A lane information prompting method, characterized in that, include: Obtain the location information of the first target in each lane and the centerline information of each lane; Based on the location information of each first target and the centerline information of each lane, the probability of each first target belonging to each lane is determined respectively. Determine the basic risks for each of the first objectives, wherein the basic risks are the inherent risks of the first objectives; For each lane, based on the probability of each first target belonging to the lane and the basic risk of the corresponding first target, a second target is selected from each first target, wherein the second target is a first target whose risk is greater than a preset first risk threshold; based on the identifier of each second target, the probability of the corresponding second target belonging to the lane and the basic risk of the corresponding second target, the elements of the probability hazard set of the lane are determined.

2. The lane information prompting method as described in claim 1, characterized in that, Also includes: Assess the confidence level of the existence of each of the lanes; The step of determining the probability that each first target belongs to each lane based on the location information of each first target and the centerline information of each lane includes: Based on the location information of each first target, the centerline information of each lane, and the confidence level of the existence of each lane, the probability of each first target belonging to each lane is determined.

3. The lane information prompting method as described in claim 1, characterized in that, The step of determining the probability that each first target belongs to each lane based on the location information of each first target and the centerline information of each lane includes: Based on the position information of each first target in the current frame and the center position information of each lane in the current frame, the update probability of each first target belonging to each lane in the current frame is determined respectively. For each of the first targets: based on the update probability of the first target belonging to each of the lanes in the current frame, and based on the probability of the first target belonging to each of the lanes in the previous frame of the current frame, determine the probability of the first target belonging to each of the lanes in the current frame.

4. The lane information prompting method as described in any one of claims 1 to 3, characterized in that, The step of selecting a second target from each of the first targets based on the probability that each first target belongs to the lane and the basic risk of the corresponding first target includes: For each lane, a corresponding weighted risk value is determined based on the probability that each first target belongs to the lane and the basic risk of the corresponding first target. Each weighted risk value is compared with a preset first risk threshold, and the first target corresponding to the weighted risk value that is greater than the first risk threshold is determined as the second target.

5. The lane information prompting method as described in claim 4, characterized in that, After determining the elements of the probabilistic hazard set of the lane, the method further includes: Remove the elements corresponding to the weighted risk values ​​in the probability hazard set that are less than a preset second risk threshold, where the preset second risk threshold is less than the first risk threshold.

6. The lane information prompting method according to any one of claims 1 to 3, characterized in that, Also includes: For each lane, the expected risk of the lane is determined based on the probability that each first target belongs to the lane and the basic risk of each first target; Output the expected risk of the lane.

7. The lane information prompting method as described in claim 6, characterized in that, Following the determination of the expected risk of the lane, the method further includes: For each lane, a risk upper bound for the lane is determined based on the lane's expected risk, the probability that each of the first targets belongs to the lane, and the basic risk of each of the first targets. Output the upper risk bound of the lane.

8. A lane information display device, characterized in that, include: The target information and lane information acquisition module is used to acquire the position information of the first target on each lane and the centerline information of each lane; The lane affiliation probability determination module is used to determine the probability that each first target belongs to each lane based on the location information of each first target and the centerline information of each lane. The inherent risk determination module is used to determine the basic risk of each of the first targets, wherein the basic risk is the inherent risk of the first target; The lane probability hazard set determination module is used to, for each lane, filter out a second target from each of the first targets based on the probability of each first target belonging to the lane and the basic risk of the corresponding first target, wherein the second target is a first target whose risk is greater than a preset first risk threshold; and determine the elements of the lane probability hazard set based on the identifier of each second target, the probability of the corresponding second target belonging to the lane and the basic risk of the corresponding second target.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method as described in any one of claims 1 to 8.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 8.

Images