Dynamic event classification method and apparatus

Abstract

The embodiment of the application discloses a dynamic event classification method and device, the method comprises the following steps: obtaining at least one correlation degree between m kinds of dynamic events according to n dimensions, grouping the m kinds of dynamic events according to the at least one correlation degree to obtain a plurality of groups; the m kinds of dynamic events are events causing changes in a dynamic information layer of the high-precision map, and n and m are integers greater than 1; wherein a first kind of dynamic event in the m kinds of dynamic events belongs to a first group in the plurality of groups; according to the obtained first dynamic event being the first kind of dynamic event, it is determined that the first dynamic event belongs to the first group. The application can save resource overhead when storing and managing dynamic events of a high-precision map, and improve storage and management efficiency.

Description

Technical Field

[0001] This application relates to the field of data processing, specifically to a method and apparatus for dynamic event classification in high-precision maps. Background Technology

[0002] High-precision maps, as a key capability for achieving autonomous driving, will effectively complement existing sensors in autonomous driving, providing vehicles with more reliable perception capabilities. Compared to traditional navigation maps, high-precision maps serving autonomous driving have higher requirements in all aspects and can work in conjunction with sensors and algorithms to support decision-making.

[0003] Because the external environment dynamically changes and affects vehicle operation during autonomous driving, high-precision maps increasingly require more dynamic information in addition to static layers to meet the development needs of the transportation sector. However, the classification of dynamic events in high-precision maps is currently in the conceptual stage, with existing concepts often classifying dynamic events based on event update frequency. However, the classification boundaries based on update frequency are unreasonable, leading to unclear storage structures when storing large amounts of dynamic events and resulting in significant storage and computing resource overhead for management.

[0004] In summary, how to save resource overhead when storing and managing dynamic events of high-precision maps is a technical problem that needs to be solved by those skilled in the art. Summary of the Invention

[0005] This application provides a dynamic event classification method and device, which can save resource consumption and improve storage and management efficiency when storing and managing dynamic events of high-precision maps.

[0006] Firstly, this application provides a dynamic event classification method, which includes:

[0007] At least one correlation degree is obtained between m types of dynamic events based on n dimensions. The m types of dynamic events are then grouped into multiple groups based on the at least one correlation degree. The m types of dynamic events are events that cause changes in the dynamic information layer in the high-precision map. The n and m are integers greater than 1. The first type of dynamic event among the m types of dynamic events belongs to the first group among the multiple groups. Based on the fact that the first dynamic event is the first type of dynamic event, it is determined that the first dynamic event belongs to the first group.

[0008] Compared to existing technologies, this application classifies various dynamic events used in high-precision maps from multiple dimensions, which can achieve a more reasonable classification and make the storage structure of various dynamic events clear and reasonable, which is easy to manage. This can save resource consumption and improve storage and management efficiency when storing and managing dynamic events of high-precision maps.

[0009] In one possible implementation, the above n dimensions include at least two of the following dimensions: the fine-grained dynamic event range covered by the event, the meaning of the event, the geometric representation of the event, the impact of the event, the target of the event, the source of the event, the priority level of the event, or the update frequency of the event status.

[0010] This application provides several dimensions for classifying dynamic events, and classifying dynamic events using at least two of these dimensions can improve the rationality of the classification compared to existing technologies, thereby facilitating storage and management.

[0011] In one possible implementation, obtaining at least one correlation degree between m types of dynamic events based on n dimensions includes: obtaining the correlation degree between each pair of the m types of dynamic events on each of the n dimensions; and combining the correlation degrees on each dimension to obtain the at least one correlation degree.

[0012] In this application, the correlation between dynamic events is first considered from a single dimension, and then the correlation between dynamic events from multiple dimensions is combined as the final classification basis, thereby realizing a more reasonable classification of dynamic events based on multiple dimensions.

[0013] In one possible implementation, the above-mentioned grouping of the m types of dynamic events into multiple groups based on the at least one correlation degree includes: comparing each of the at least one correlation degree with a first numerical range; and classifying the two dynamic events corresponding to the correlation degree within the first numerical range into the same group.

[0014] In this application, by pre-setting classification criteria, only dynamic events that meet the criteria can be classified into the same category, thereby ensuring the rationality of the classification.

[0015] In one possible implementation, the above n dimensions are prioritized, and at least one correlation degree is obtained between m dynamic events based on the n dimensions. The m dynamic events are then grouped according to this at least one correlation degree to obtain multiple groups, including:

[0016] Obtain the first degree of correlation between each pair of the m dynamic events in the first dimension;

[0017] Based on the first correlation, the m types of dynamic events are grouped to obtain a first plurality of groups. The first plurality of groups includes a second group, which includes p dynamic events from the m types of dynamic events, where p is a positive integer greater than 1.

[0018] Obtain the second correlation degree between each pair of the p dynamic events in the second dimension, where the second dimension is the dimension with a lower priority than the first dimension among the n dimensions;

[0019] Based on this second correlation, the p dynamic events are grouped to obtain a second set of groups.

[0020] In this application, dynamic events are classified according to multiple dimensions in descending order of priority. First, dynamic events are classified based on the highest priority dimension. Then, each dimension is further classified based on the classification results of the next higher priority dimension, until the classification is completed based on the lowest priority dimension. The result of this classification is the final classification result. This application comprehensively considers multiple dimensions when classifying dynamic events, thereby improving the rationality of event classification.

[0021] In one possible implementation, the m types of dynamic events are associated with multiple layers of the high-precision map according to the multiple groupings mentioned above.

[0022] For example, the above m types of dynamic events include a second dynamic event and a third dynamic event. The second dynamic event and the first type of dynamic event both belong to the first group among the above multiple groups, and the third dynamic event belongs to the second group among the above multiple groups; the above method also includes:

[0023] The first dynamic event and the second dynamic event are associated and expressed in the first layer, and the third dynamic event is associated and expressed in the second layer. The first layer and the second layer are two different dynamic information layers in the high-precision map.

[0024] In this application, by associating dynamic events in the same group to express them in the same layer, and associating dynamic events in different groups to express them in different layers, various dynamic events can be expressed in a high-precision map in an efficient and orderly manner.

[0025] Secondly, this application provides a dynamic event classification method, which includes:

[0026] At least one correlation degree is obtained between m types of dynamic events based on n dimensions, and the m types of dynamic events are grouped into multiple groups based on the at least one correlation degree; the m types of dynamic events are events that cause changes in the dynamic information layer in the high-precision map, and n and m are integers greater than 1.

[0027] Compared to existing technologies, this application classifies various dynamic events used in high-precision maps from multiple dimensions, which can achieve a more reasonable classification and make the storage structure of various dynamic events clear and reasonable, which is easy to manage. This can save resource consumption and improve storage and management efficiency when storing and managing dynamic events of high-precision maps.

[0028] In one possible implementation, the above n dimensions include at least two of the following dimensions: the fine-grained dynamic event range covered by the event, the meaning of the event, the geometric representation of the event, the impact of the event, the target of the event, the source of the event, the priority level of the event, or the update frequency of the event status.

[0029] This application provides several dimensions for classifying dynamic events, and classifying dynamic events using at least two of these dimensions can improve the rationality of the classification compared to existing technologies, thereby facilitating storage and management.

[0030] In one possible implementation, obtaining at least one correlation degree between m types of dynamic events based on n dimensions includes: obtaining the correlation degree between each pair of the m types of dynamic events on each of the n dimensions; and combining the correlation degrees on each dimension to obtain the at least one correlation degree.

[0031] In this application, the correlation between dynamic events is first considered from a single dimension, and then the correlation between dynamic events from multiple dimensions is combined as the final classification basis, thereby realizing a more reasonable classification of dynamic events based on multiple dimensions.

[0032] In one possible implementation, the above-mentioned grouping of the m dynamic events according to the at least one correlation degree to obtain multiple groups includes: comparing each of the at least one correlation degree with a first numerical range; and classifying the two dynamic events corresponding to the correlation degree in the first numerical range into the same group.

[0033] In this application, by pre-setting classification criteria, only dynamic events that meet the criteria can be classified into the same category, thereby ensuring the rationality of the classification.

[0034] In one possible implementation, the above n dimensions are prioritized, and at least one correlation degree is obtained between m dynamic events based on the n dimensions. The m dynamic events are then grouped according to this at least one correlation degree to obtain multiple groups, including:

[0035] Obtain the first degree of correlation between each pair of the m dynamic events in the first dimension;

[0036] Based on the first correlation, the m types of dynamic events are grouped to obtain a first plurality of groups. The first plurality of groups includes a second group, which includes p dynamic events from the m types of dynamic events, where p is a positive integer greater than 1.

[0037] Obtain the second correlation degree between each pair of the p dynamic events in the second dimension, where the second dimension is the dimension with a lower priority than the first dimension among the n dimensions;

[0038] Based on this second correlation, the p dynamic events are grouped to obtain a second set of groups.

[0039] In this application, dynamic events are classified according to multiple dimensions in descending order of priority. First, dynamic events are classified based on the highest priority dimension. Then, each dimension is further classified based on the classification results of the next higher priority dimension, until the classification is completed based on the lowest priority dimension. The result of this classification is the final classification result. This application comprehensively considers multiple dimensions when classifying dynamic events, thereby improving the rationality of event classification.

[0040] In one possible implementation, the m types of dynamic events are associated with multiple layers of the high-precision map according to the multiple groupings mentioned above.

[0041] For example, the above m types of dynamic events include a second dynamic event and a third dynamic event. The second dynamic event and the first type of dynamic event both belong to the first group among the above multiple groups, and the third dynamic event belongs to the second group among the above multiple groups; the above method also includes:

[0042] The first dynamic event and the second dynamic event are associated and expressed in the first layer, and the third dynamic event is associated and expressed in the second layer. The first layer and the second layer are two different dynamic information layers in the high-precision map.

[0043] In this application, by associating dynamic events in the same group to express them in the same layer, and associating dynamic events in different groups to express them in different layers, various dynamic events can be expressed in a high-precision map in an efficient and orderly manner.

[0044] Thirdly, this application provides an apparatus comprising:

[0045] The acquisition unit is used to obtain at least one degree of correlation between m types of dynamic events based on n dimensions.

[0046] A grouping unit is used to group the m types of dynamic events into multiple groups based on at least one correlation degree; the m types of dynamic events are events that cause changes in the dynamic information layer in the high-precision map, where n and m are integers greater than 1; wherein, the first type of dynamic event among the m types of dynamic events belongs to the first group among the multiple groups;

[0047] The determining unit is configured to determine that the first dynamic event belongs to the first group based on the first dynamic event being the first type of dynamic event.

[0048] In one possible implementation, the above n dimensions include at least two of the following dimensions: the fine-grained dynamic event range covered by the event, the meaning of the event, the geometric representation of the event, the impact of the event, the target of the event, the source of the event, the priority level of the event, or the update frequency of the event status.

[0049] In one possible implementation, the above-mentioned acquisition unit is specifically used for:

[0050] Obtain the pairwise correlation between each of the m types of dynamic events in each of the n dimensions;

[0051] The correlation degree is obtained by combining the correlation degree of each dimension.

[0052] In one possible implementation, the above-mentioned grouping unit is specifically used for:

[0053] Each of the at least one correlation degree is compared with a first numerical range;

[0054] The two dynamic events corresponding to the correlation within the first numerical range are grouped into the same group.

[0055] In one possible implementation, the acquisition unit is specifically used to: obtain the first correlation degree between each pair of the m dynamic events in the first dimension;

[0056] The aforementioned grouping unit is specifically used to: group the m types of dynamic events according to the first correlation degree to obtain a first plurality of groups, the first plurality of groups including a second group, the second group including p dynamic events among the m types of dynamic events, where p is a positive integer greater than 1;

[0057] The acquisition unit is also specifically used to: obtain the second correlation degree between each pair of the p kinds of dynamic events in the second dimension, wherein the second dimension is the dimension with a lower priority than the first dimension among the n dimensions;

[0058] The grouping unit is also specifically used to: group the p dynamic events according to the second correlation to obtain a second set of groups.

[0059] In one possible implementation, the above-described device further includes:

[0060] The association unit is used to associate the above m types of dynamic events with multiple layers of the above high-precision map according to the above multiple groups.

[0061] Fourthly, this application provides an apparatus comprising:

[0062] The acquisition unit is used to obtain at least one degree of correlation between m types of dynamic events based on n dimensions.

[0063] A grouping unit is used to group the m types of dynamic events into multiple groups based on the at least one correlation degree; the m types of dynamic events are events that cause changes in the dynamic information layer in the high-precision map, and n and m are integers greater than 1.

[0064] In one possible implementation, the above n dimensions include at least two of the following dimensions: the fine-grained dynamic event range covered by the event, the meaning of the event, the geometric representation of the event, the impact of the event, the target of the event, the source of the event, the priority level of the event, or the update frequency of the event status.

[0065] In one possible implementation, the above-mentioned acquisition unit is specifically used for:

[0066] Obtain the pairwise correlation between each of the m types of dynamic events in each of the n dimensions;

[0067] The correlation degree is obtained by combining the correlation degree of each dimension.

[0068] In one possible implementation, the above-mentioned grouping unit is specifically used for:

[0069] Each of the at least one correlation degree is compared with a first numerical range;

[0070] The two dynamic events corresponding to the correlation within the first numerical range are grouped into the same group.

[0071] In one possible implementation, the acquisition unit is specifically used to: obtain the first correlation degree between each pair of the m dynamic events in the first dimension;

[0072] The aforementioned grouping unit is specifically used to: group the m types of dynamic events according to the first correlation degree to obtain a first plurality of groups, the first plurality of groups including a second group, the second group including p dynamic events among the m types of dynamic events, where p is a positive integer greater than 1;

[0073] The acquisition unit is also specifically used to: obtain the second correlation degree between each pair of the p kinds of dynamic events in the second dimension, wherein the second dimension is the dimension with a lower priority than the first dimension among the n dimensions;

[0074] The grouping unit is also specifically used to: group the p dynamic events according to the second correlation to obtain a second set of groups.

[0075] In one possible implementation, the above-described device further includes:

[0076] The association unit is used to associate the above m types of dynamic events with multiple layers of the above high-precision map according to the above multiple groups.

[0077] Fifthly, this application provides an apparatus that may include a processor for implementing the dynamic event classification method described in the first aspect. The apparatus may also include a memory coupled to the processor, which, when executing a computer program stored in the memory, can implement the dynamic event classification method described in the first aspect or any possible implementation thereof. The apparatus may also include a communication port for communicating with other devices; exemplaryly, the communication port may be a transceiver, circuit, bus, module, or other type of communication port.

[0078] In one possible implementation, the device may include:

[0079] Memory, used to store computer programs;

[0080] The processor is configured to obtain at least one correlation degree between m types of dynamic events based on n dimensions, group the m types of dynamic events into multiple groups based on the at least one correlation degree, and determine that the first dynamic event belongs to the first group based on the first dynamic event obtained. The m types of dynamic events are events that cause changes in the dynamic information layer in the high-precision map, where n and m are integers greater than 1; and the first type of dynamic event among the m types of dynamic events belongs to the first group among the multiple groups.

[0081] It should be noted that the computer program in the memory of this application can be pre-stored or downloaded from the Internet and stored after use of the device. This application does not specifically limit the source of the computer program in the memory. The coupling in the embodiments of this application is an indirect coupling or connection between devices, units, or modules, which can be electrical, mechanical, or other forms, for information interaction between devices, units, or modules.

[0082] Sixthly, this application provides an apparatus that may include a processor for implementing the dynamic event classification method described in the second aspect above. The apparatus may also include a memory coupled to the processor, which, when executing a computer program stored in the memory, can implement the dynamic event classification method described in the second aspect or any possible implementation thereof. The apparatus may also include a communication port for communicating with other devices; exemplaryly, the communication port may be a transceiver, circuit, bus, module, or other type of communication port.

[0083] In one possible implementation, the device may include:

[0084] Memory, used to store computer programs;

[0085] The processor is used to obtain at least one correlation degree between m types of dynamic events based on n dimensions, and to group the m types of dynamic events into multiple groups based on the at least one correlation degree. The m types of dynamic events are events that cause changes in the dynamic information layer in the high-precision map, where n and m are integers greater than 1.

[0086] It should be noted that the computer program in the memory of this application can be pre-stored or downloaded from the Internet and stored after use of the device. This application does not specifically limit the source of the computer program in the memory. The coupling in the embodiments of this application is an indirect coupling or connection between devices, units, or modules, which can be electrical, mechanical, or other forms, for information interaction between devices, units, or modules.

[0087] In a seventh aspect, this application provides a computer-readable storage medium storing a computer program that is executed by a processor to implement the method described in any one of the first or second aspects.

[0088] Eighthly, this application provides a computer program product that, when read and executed by a computer, will execute the method described in any one of the first or second aspects.

[0089] The solutions provided in the second to fifth aspects above are used to implement or cooperate with the methods provided in the first aspect above, and therefore can achieve the same or corresponding beneficial effects as the first aspect, which will not be elaborated here.

[0090] In summary, compared with existing technologies, this application classifies various dynamic events used in high-precision maps from multiple dimensions, which can achieve a more reasonable classification, making the storage structure of various dynamic events clear and reasonable, and easy to manage. This can save resource consumption and improve storage and management efficiency when storing and managing dynamic events of high-precision maps. Attached Figure Description

[0091] The accompanying drawings used in the embodiments of this application will be described below.

[0092] Figure 1 The diagram shown is a flowchart of a dynamic event classification method provided in an embodiment of this application.

[0093] Figure 2 and Figure 3 The image shown is a schematic diagram of the dynamic event classification results provided in this application;

[0094] Figure 4 This is a schematic diagram of the logical structure of a device provided in an embodiment of this application;

[0095] Figure 5 A schematic diagram of the hardware structure of the device provided in the embodiments of this application. Detailed Implementation

[0096] The technical solutions in the embodiments of this application are described below with reference to the accompanying drawings.

[0097] To better understand the dynamic event classification method provided by the embodiments of the present invention, the applicable scenarios of the embodiments of the present invention will be described below by way of example.

[0098] For example, the dynamic event classification method provided in this application can be applied to high-precision maps. With the continuous development of autonomous driving, semi-autonomous driving, and transportation networking, high-precision maps have gradually become a better tool for serving these fields. For instance, in the field of autonomous driving, because high-precision maps provide a more accurate, clear, and comprehensive description of roads, and can reflect various dynamic events on the road in real time, they provide vehicles with more room for prediction, enabling them to plan their driving in advance and ensuring smooth and economical driving. Furthermore, high-precision maps can help vehicles reduce computational load. When a vehicle needs to pass through an intersection, it needs to perceive the status of traffic lights ahead in advance. High-precision maps can help it locate the specific area where the traffic lights are located, thereby effectively reducing the computational load of full-range scanning and recognition.

[0099] The dynamic event classification method provided in this application can also be applied to other fields, not limited to the high-precision map field described above.

[0100] Because the classification of dynamic events in existing high-precision maps is mostly at the conceptual stage and the classification methods are unclear, the server's storage structure for dynamic events is unclear, resulting in high management resource overhead. To solve this problem, this application provides a dynamic event classification method, which can be applied to, but is not limited to, the application areas described above. Optionally, this method can be implemented in the server, and then the classification results can be applied to the high-precision map. In the embodiments of this application, grouping is equivalent to classification, with one group representing one class.

[0101] See Figure 1 This method may include, but is not limited to, the following steps:

[0102] S101. Obtain at least one correlation degree between m types of dynamic events based on n dimensions, and group the m types of dynamic events according to the at least one correlation degree to obtain multiple groups; where n and m are integers greater than 1, and the m types of dynamic events are events that cause changes in the dynamic information layer in the high-precision map.

[0103] In a specific embodiment, the above m types of dynamic events may include some or all of the following events:

[0104] Weather conditions can include dynamic events such as temperature, air pressure, humidity, sunny, cloudy, windy, foggy, rainy, lightning, snowy, frosty, thunderstorms, and hail.

[0105] Road cover conditions, or road surface environment, can include dynamic events such as water accumulation, snow accumulation, icing, and damage to road cover.

[0106] Road adhesion coefficient: This refers to the adhesion between the road and the vehicle tires. The greater the adhesion, the less likely the vehicle is to slip.

[0107] Road visibility conditions: This can refer to the visible distance caused by factors such as fog, haze, and light.

[0108] Temporary points of interest (POIs) can include non-fixed points of interest, temporary parking spaces, temporary charging stations, and temporary public service points, among other dynamic events.

[0109] Recommendation information can include dynamic events such as user interest recommendations and itinerary suggestions.

[0110] Road traffic conditions: This can include dynamic events such as traffic accidents, traffic control, road construction, and accident-prone areas.

[0111] Traffic flow conditions: This can include the amount of traffic on each road.

[0112] Traffic light status: This can include dynamic events such as traffic light phase status, semantic information, timing data, and operating status.

[0113] Vehicle obstacles: This can include the type of vehicle causing the obstacle, motion information, headlight and door information, control information, and driving behavior prediction information.

[0114] Other obstacles: This can include all obstacles other than vehicle obstacles.

[0115] Risk warning information can include information such as the type of risk (collision risk in all directions, landslide risk, etc.), risk level, and avoidance suggestions.

[0116] Collaborative driving status: This can refer to vehicle collaborative guidance information, including lane changing collaboration, steering collaboration, special vehicle avoidance, and parking collaboration.

[0117] Route planning: This can refer to the planned route based on existing road conditions.

[0118] Changes in road network topology: This can refer to temporary changes in drivable routes caused by road traffic incidents, manifested as changes in the geometric representation and attributes of road lines and lane lines.

[0119] The above examples exemplify some dynamic events applied to high-precision maps. By updating these dynamic events to the high-precision map in real time, various dynamic conditions on the road can be provided to vehicles and other objects using the high-precision map, offering valuable reference for further route planning and behavior prediction. It should be noted that the dynamic events actually used in high-precision maps are not limited to the events described above, and this application does not limit the specific dynamic events or their number.

[0120] The above n dimensions may include at least two of the following dimensions:

[0121] The fine-grained scope of dynamic events covered by the event: This dimension refers to the scope of which sub-dynamic events are included in a certain known dynamic event. Based on this, sub-dynamic events within the known scope can be grouped into the same category.

[0122] For example, a dynamic event like weather conditions can include multiple sub-dynamic events such as temperature, air pressure, humidity, sunny, cloudy, windy, foggy, rainy, lightning, snowy, frosty, thunder, and hail. These are clearly known, so these sub-dynamic events can be categorized into the same group.

[0123] Meaning of events: This dimension refers to classifying events by the meaning of their names.

[0124] For example, the dynamic events of rain and snow can be understood from their names to belong to the same weather state, so rain and snow can be classified as weather states.

[0125] Geometric representation of events: This dimension refers to the geometric form in which each dynamic event is represented in a high-precision map.

[0126] For example, dynamic events in weather conditions such as rain, snow, sunshine, and overcast skies can be represented in high-precision maps using area-level polygon sets. Similarly, road construction and traffic accidents can generally be represented using irregular shapes composed of the vertices and connecting lines of polygons at the road or lane level.

[0127] The aforementioned area-level planar geometric representation can be a polygonal surface or a toroidal surface, etc. The difference between area-level and road-level geometric representation is that area-level events with a large impact range can directly express their geometric location in the map tiles, while road-level events need to express their geometric location on the road. The relative coordinates of the location points in the road geometric representation are determined relative to a certain feature reference point on the road.

[0128] Impact of the event: This dimension refers to the impact of a dynamic event on traffic and roads.

[0129] For example, road construction and traffic control can affect traffic flow or even make roads impassable. Rain and fog can reduce visibility and distance.

[0130] The target of an event: This dimension refers to the objects affected or those who should be informed after a dynamic event occurs. For example, water accumulation or coverings affect the road surface and impede traffic flow, while cooperative driving information affects vehicles or drivers, etc.

[0131] Source of the event: This dimension refers to the source of dynamic events. For example, weather conditions come from the meteorological bureau, and road traffic information comes from the transportation bureau, etc.

[0132] Event status update frequency: This dimension refers to the frequency with which dynamic events themselves are updated. For example, for road construction, a short construction period results in a high update frequency, while a long construction period results in a low update frequency. Another example is the uncertain update frequency of weather conditions.

[0133] Event priority: This dimension refers to the importance of dynamic events or their impact on driving safety. For example, weather conditions have a higher priority than temporary points of interest (POIs) information because weather conditions directly affect driving safety and stability, while POIs are merely supplementary dynamic events; their presence is beneficial, but their absence has little impact on driving safety and stability.

[0134] The above examples illustrate several considerations that can be used for dynamic event classification. It should be noted that the actual considerations are not limited to those described above, and this application does not limit the specific considerations.

[0135] Based on the above introduction, the above method of obtaining at least one correlation degree between m types of dynamic events according to n dimensions and grouping the m types of dynamic events into multiple groups according to the at least one correlation degree can be implemented in various ways. Two possible implementation methods are exemplarily introduced below.

[0136] The first implementation method will be introduced below.

[0137] In this implementation, the correlation degree between each pair of the m dynamic events on each of the n dimensions is first obtained; the correlation degree on each dimension is combined to obtain at least one correlation degree. Then, each of the at least one correlation degree is compared with a first numerical range; the two dynamic events corresponding to the correlation degree within the first numerical range are grouped into the same group.

[0138] In a specific embodiment, the correlation degree between each pair of dynamic events in each dimension can be parameterized to obtain a specific value. Then, the comprehensive correlation degree between each pair of dynamic events is calculated by combining the correlation degrees of each dimension. The comprehensive correlation degree between each pair of m dynamic events is at least one of the aforementioned correlation degrees. In one possible implementation, the comprehensive correlation degree between each pair of dynamic events can be calculated using the following formula:

[0139]

[0140] Alternatively, in one possible implementation, the overall correlation between any two dynamic events can be calculated using the following formula:

[0141]

[0142] In formulas (1) and (2) above, R is the comprehensive correlation degree between the pairwise dynamic events, r_i is the correlation degree of the two dynamic events in the i-th dimension, and k is the weight adjustment parameter. The weights of the above n dimensions can be the same or different, and can be adjusted by k. k_i is the weight adjustment parameter of the i-th dimension.

[0143] After calculating the comprehensive correlation between each pair of dynamic events, these comprehensive correlations can be compared with a first numerical range. The two dynamic events corresponding to the correlations within this first numerical range are grouped into the same group, i.e., classified into the same category. This first numerical range can be set according to actual circumstances, and this application does not impose any restrictions on it.

[0144] To make it easier to understand, the following example is provided.

[0145] Suppose the dynamic events to be classified are five types: weather conditions, road surface conditions, road construction, traffic accidents, and traffic light conditions. And suppose we classify these five dynamic events based on two dimensions: the meaning of the events and their geometric representation. First, we obtain the pairwise correlation between each of the five dynamic events along each of these two dimensions. These correlations can be pre-configured or calculated based on set conditions. Examples of the pairwise correlations between the five dynamic events along each of these two dimensions can be found in Tables 1 and 2.

[0146] Table 1. The correlation between pairs of dynamic events in terms of their meaning.

[0147] Event Name Weather conditions Road surface environment ROAD WORK Traffic accidents Traffic light status Weather conditions 1 0.8 0 0 0 Road surface environment 0.8 1 0.2 0.2 0 ROAD WORK 0 0.2 1 0.8 0.4 Traffic accidents 0 0.2 0.8 1 0.4 Traffic light status 0 0 0.4 0.4 1

[0148] Table 1 exemplarily illustrates the pairwise correlations between five dynamic events—weather conditions, road surface conditions, road construction, traffic accidents, and traffic light status—on the dimension of event meaning. In this application, the semantic correlation between dynamic events is exemplarily represented by coefficients after parameter normalization. It should be noted that the semantic correlation between dynamic events can also be represented in other ways, and this application does not limit this. For example, it can also be represented by a scoring method, with a maximum score of 10 points, and higher scores indicating higher correlation. The representation of correlations on other dimensions can be found in the description here, and will not be repeated hereafter.

[0149] Specifically, weather conditions affect road surface conditions, and their causal relationship is relatively strong, as is their semantic correlation. Therefore, the semantic correlation coefficient between weather conditions and road surface conditions is relatively high at 0.8. However, weather conditions are not semantically related to the other three dynamic events, and their corresponding semantic correlation coefficient is 0.

[0150] Road surface environment, road construction, and traffic accidents are all events that occur on the road surface. Although there is a semantic correlation, it is relatively low. Therefore, the semantic correlation coefficient between road surface environment and road construction and traffic accidents is 0.2 respectively. In addition, there is no semantic correlation between road surface environment and traffic light status, so the corresponding semantic correlation coefficient is 0.

[0151] Road construction and traffic accidents are both road occupancy events, and their meanings are strongly related. Therefore, the meaning correlation coefficient between road construction and traffic accidents is 0.8. In addition, road construction and traffic lights are both traffic-related events, and their meanings are related but relatively weak. Therefore, the meaning correlation coefficient is 0.4.

[0152] Traffic accidents and traffic lights are both traffic-related events, and while they have some semantic correlation, it is relatively low, resulting in a semantic correlation coefficient of 0.4.

[0153] Table 2. The degree of pairwise correlation between dynamic events in the geometric dimension of events.

[0154] Event Name Weather conditions Road surface environment ROAD WORK Traffic accidents Traffic light status Weather conditions 1 0.8 0.4 0.4 0 Road surface environment 0.8 1 0.8 0.8 0 ROAD WORK 0.4 0.8 1 1 0 Traffic accidents 0.4 0.8 1 1 0 Traffic light status 0 0 0 0 1

[0155] Table 2 illustrates, for example, the pairwise correlations between five dynamic events—weather conditions, road surface conditions, road construction, traffic accidents, and traffic light conditions—on the geometric representation dimension of events.

[0156] Specifically, small-scale weather events can be represented using road-level polygon vertices and connections, while large-scale events are represented using region-level polygon vertices and connections, similar to the road surface environment. Therefore, the geometric representation correlation coefficient between weather conditions and the road surface environment is 0.8. It should be noted that, as described above, the location reference methods for road-level and region-level surface representations differ. The road surface environment is generally road-level, while weather conditions can be either road-level or region-level, so the representations are similar but not entirely identical. Furthermore, the geometric representation correlation between weather conditions and road construction and traffic accidents is relatively weak, resulting in a geometric representation correlation coefficient of 0.4 for both.

[0157] Small-scale events in the road surface environment are represented in the same way as road construction and traffic accidents, which can all be represented by polygon vertices and connections to express irregular occupancy areas. However, large-scale events are represented differently, and the geometric representation correlation is also stronger. Therefore, the geometric representation correlation coefficient between the road surface environment and road construction and traffic accidents is 0.8.

[0158] Road construction and traffic accidents can generally be represented by polygon vertices and connecting lines to depict irregular occupancy areas. Therefore, the geometric correlation coefficient between these two dynamic events is 1.

[0159] Traffic light status is generally an extended attribute associated with the traffic light identifier ID on the map, and does not require geometric representation. Therefore, it has no geometric relationship with other events.

[0160] After obtaining the pairwise correlation between the five dynamic events in each of the two dimensions mentioned above, the overall correlation between the five dynamic events can be calculated by combining the correlations in each dimension. The overall correlation between the five dynamic events can be calculated using formula (2) as an example. Assuming that the weight adjustment parameters for the two dimensions are the same and both are set to 1, the calculation formula can be simplified to:

[0161]

[0162] Taking weather conditions and road surface environment as two dynamic events as examples, according to Tables 1 and 2, the correlation coefficients of weather conditions and road surface environment are both 0.8 in terms of the meaning of the event and the geometric expression of the event. Substituting 0.8 into formula (3) yields:

[0163]

[0164] The comprehensive correlation between the five dynamic events calculated according to the above formula (3) can be found in Table 3.

[0165] Table 3. Comprehensive correlation between pairs of dynamic events

[0166]

[0167]

[0168] After obtaining the correlation between each of the above five types of dynamic events, the dynamic events can be classified using a pre-set threshold. Assuming the preset threshold is 0.7, if the overall correlation between two dynamic events is greater than or equal to the threshold of 0.7, or in other words, within the range of 0.7 to 1, then the two dynamic events can be classified into the same category; if the overall correlation between two dynamic events is less than the threshold of 0.7, or in other words, within the range of 0 to 0.7, then the two dynamic events do not belong to the same category.

[0169] Therefore, based on this standard and the comprehensive correlation in Table 3, the above five types of events are classified as follows:

[0170] As shown in Table 3, the overall correlation between weather conditions and road surface environment is 0.8, which is greater than the threshold of 0.7. However, the overall correlation between weather conditions and other dynamic events is much lower than the threshold of 0.7. Furthermore, the overall correlation between road surface environment and other dynamic events is also lower than the threshold of 0.7. Therefore, weather conditions and road surface environment can be classified into the same major category.

[0171] The overall correlation between road construction and traffic accidents is 0.9, which is greater than the threshold of 0.7. However, the overall correlation between road construction and other dynamic events is far below the threshold of 0.7. Furthermore, the overall correlation between traffic accidents and other dynamic events is also below the threshold of 0.7. Therefore, road construction and traffic accidents can be classified into the same major category.

[0172] The overall correlation between traffic light status and any dynamic event is far below the threshold of 0.7, therefore traffic light status is classified as a separate category.

[0173] In summary, the classification of the above five types of dynamic events has been completed.

[0174] In another possible implementation, the correlation between the aforementioned dynamic events can also be calculated using distance. The distance between dynamic events referred to here is an abstract distance, meaning the degree of uncorrelation between the two dynamic events. The closer the distance values ​​between dynamic events, the stronger the correlation between them. For example, for the normalized correlation mentioned above, the distance can be obtained by subtracting the correlation from 1. For easier understanding, please refer to Tables 4 and 5.

[0175] Table 4. Distances between pairs of dynamic events in terms of the meaning dimension of events

[0176] Event Name Weather conditions Road surface environment ROAD WORK Traffic accidents Traffic light status Weather conditions 0 0.2 1 1 1 Road surface environment 0.2 0 0.8 0.8 1 ROAD WORK 1 0.8 0 0.2 0.6 Traffic accidents 1 0.8 0.2 0 0.6 Traffic light status 1 1 0.6 0.6 0

[0177] As can be seen, the distance between each pair of dynamic events in Table 4 is obtained by subtracting the corresponding correlation degree in Table 1 from 1.

[0178] Table 5. Distances between pairs of dynamic events in the geometric dimension of events.

[0179] Event Name Weather conditions Road surface environment ROAD WORK Traffic accidents Traffic light status Weather conditions 0 0.2 0.6 0.6 1 Road surface environment 0.2 0 0.2 0.2 1 ROAD WORK 0.6 0.2 0 0 1 Traffic accidents 0.6 0.2 0 0 1 Traffic light status 1 1 1 1 0

[0180] As can be seen, the distance between each pair of dynamic events in Table 5 is obtained by subtracting the corresponding correlation degree in Table 2 from 1.

[0181] After obtaining the pairwise distances between the five dynamic events in each of the two dimensions mentioned above, the combined distances between the five dynamic events can be calculated by integrating the distances in each dimension. The combined distance between the five dynamic events can be calculated using formula (1) as an example. Assuming that the weight adjustment parameters for the two dimensions are the same and both are set to 1, the calculation formula can be simplified to:

[0182]

[0183] Taking weather conditions and road surface environment as two dynamic events as examples, according to Tables 4 and 5, the distance coefficients for both weather conditions and road surface environment are 0.2 in terms of the meaning and geometric expression of the events. Substituting 0.2 into formula (4) yields:

[0184]

[0185] The comprehensive distances between each pair of the five dynamic events calculated according to the above formula (4) can be found in Table 6.

[0186] Table 6. Overall Distance Between Each Pair of Dynamic Events

[0187] Event Name Weather conditions Road surface environment ROAD WORK Traffic accidents Traffic light status Weather conditions 0 0.283 1.17 1.17 1.41 Road surface environment 0.283 0 0.82 0.82 1.41 ROAD WORK 1.17 0.82 0 0.2 1.17 Traffic accidents 1.17 0.82 0.2 0 1.17 Traffic light status 1.41 1.41 1.17 1.17 0

[0188] After obtaining the pairwise distances between the above five dynamic events, the dynamic events can be classified using a pre-set threshold. Assuming the preset threshold is 0.3, if the combined distance between two dynamic events is less than or equal to the threshold of 0.3, or in other words, within the range of 0 to 0.3, then the two dynamic events can be classified into the same category; if the combined distance between two dynamic events is greater than the threshold of 0.3, or in other words, within the range of 0.3 to 0.1, then the two dynamic events do not belong to the same category.

[0189] Therefore, based on this standard and the comprehensive distance in Table 6, the above five types of events are classified as follows:

[0190] As shown in Table 6, the combined distance between weather conditions and road surface environment is 0.283, which is less than the threshold of 0.3. However, the combined distance between weather conditions and other dynamic events is much greater than the threshold of 0.3. Furthermore, the combined distance between road surface environment and other dynamic events is also greater than the threshold of 0.3. Therefore, weather conditions and road surface environment can be classified into the same major category.

[0191] The combined distance between road construction and traffic accidents is 0.2, which is less than the threshold of 0.3. However, the combined distance between road construction and other dynamic events is much greater than the threshold of 0.3. Furthermore, the combined distance between traffic accidents and other dynamic events is also greater than the threshold of 0.3. Therefore, road construction and traffic accidents can be classified into the same major category.

[0192] The overall distance between the traffic light state and any dynamic event is much greater than the threshold of 0.3, therefore the traffic light state is classified as a separate category.

[0193] In summary, the classification of the above five types of dynamic events has been completed. The above is merely an illustrative description of the process of classifying dynamic events, and this application does not limit the dimensions used in specific embodiments, the number of dynamic events, the specific correlation, the classification threshold, etc.

[0194] It should be noted that the five dynamic events exemplified above are not all dynamic events with the smallest granularity. In one possible implementation, the m dynamic events to be classified can be dynamic events with the smallest granularity, such as rain, water accumulation, and construction. Then, the classification method described above is used to classify these m dynamic events. The classification results can be found in [reference needed]. Figure 2 . Figure 2 The classification result shown is only one example. In actual processing, different classification results will be obtained depending on the dimensions considered.

[0195] Optionally, assuming that classifying the m types of dynamic events results in multiple groups, the dynamic events in these multiple groups can be considered as various types of dynamic events. The above classification method can then be used to further classify these various types of dynamic events, iterating step by step until a tree structure diagram of the dynamic event classification is obtained. For example, see [reference needed]. Figure 3 . Figure 3 The classification result shown is only one example. In actual processing, different classification results will be obtained depending on the dimensions considered.

[0196] In one possible implementation, the m types of dynamic events to be classified may simultaneously include dynamic events with larger granularity and dynamic events with smaller granularity. The smaller granularity dynamic events can be sub-dynamic events of the larger granularity dynamic events, or in other words, the smaller granularity dynamic events are subclasses of the larger granularity dynamic events, while the larger granularity dynamic events are parent classes of the smaller granularity dynamic events, such as weather conditions, rain, and snow. In this case, the parent and child classes have a high degree of correlation or a small distance, and therefore can be classified into the same category.

[0197] The following describes a second possible implementation method, namely, implementation method two, which involves obtaining at least one degree of correlation between m types of dynamic events based on n dimensions and grouping the m types of dynamic events into multiple groups based on this at least one degree of correlation.

[0198] In this implementation, the above n dimensions are ordered by priority. So we can first obtain the correlation between each pair of the m dynamic events on the first dimension; then group the m dynamic events according to these correlations to get multiple groups. Then, we continue to classify the dynamic events divided by the higher-level dimension according to the priority of the dimensions from high to low, until the classification is completed based on the dimension with the lowest priority. The result of this classification is the final classification result.

[0199] For example, assuming n is 3, and the three dimensions are ordered from highest to lowest priority as the first dimension, the second dimension, and the third dimension, then firstly, the correlation between each pair of the m dynamic events in the first dimension is obtained. Based on these correlations, the m dynamic events are classified into c1 dynamic event categories, where c1 is a positive integer greater than 0. Each of the c1 dynamic event categories includes one or more of the m dynamic events. The specific process of classifying dynamic events based on correlation can be found in the description in Table 3 above, and will not be repeated here.

[0200] Then, based on the above c1 dynamic event categories, the correlation degree between each pair of dynamic events in each of the c1 dynamic event categories is obtained in the second dimension. The dynamic events in each category are further classified according to the correlation degree between each pair of dynamic events in each category. After each category is further classified, c2 dynamic event categories are obtained, where c2 is a positive integer greater than 0.

[0201] Then, based on the aforementioned c2 dynamic event categories, the pairwise correlation degree between dynamic events in each of these c2 categories is obtained on the third dimension. The dynamic events in each category are further classified based on the pairwise correlation degree. After all categories are further classified, c3 dynamic event categories are obtained, where c3 is a positive integer greater than 0. These c3 dynamic event categories are the final classification result.

[0202] To make it easier to understand, the following example is provided.

[0203] Let's take the classification of the five dynamic events mentioned above as an example, focusing on both the meaning and geometric representation of the events. Assuming that the meaning of an event has higher priority than its geometric representation, we will first classify the five events—weather conditions, road surface conditions, road construction, traffic accidents, and traffic light conditions—based on their meaning.

[0204] Specifically, we first obtain the pairwise correlation between these five dynamic events in terms of their meaning, as shown in Table 1. Similarly, we can set a threshold as the basis for classification. For example, a threshold of 0.7 would group two dynamic events with a correlation greater than the threshold into the same category. Thus, weather conditions and road conditions can be grouped together as the "Environment" category, and road construction and traffic accidents as the "Traffic" category. Theoretically, traffic light status should be classified separately, but given its high correlation with the meaning of the "Traffic" category, it can also be classified as part of the "Traffic" category. Therefore, based on the dimension of event meaning, the above five dynamic events can be classified into two main categories: Environment and Traffic.

[0205] Based on the above classification into environment and traffic categories, we further categorize these two categories of dynamic events according to their geometric representations. First, we obtain the geometric representation correlation between weather conditions and road surface conditions in the environment category, and the pairwise geometric representation correlation between road construction, traffic accidents, and traffic light status in the traffic category. These correlations are shown in Table 2 and will not be elaborated here. Similarly, we can set a threshold of 0.7. Two dynamic events with a correlation greater than the threshold are grouped into the same category. Therefore, the geometric representation correlation between weather conditions and road surface conditions in the environment category is 0.8, which is greater than the threshold, so weather conditions and road surface conditions are still grouped into the same category. The geometric representation correlation between road construction and traffic accidents in the traffic category is 1, indicating that the geometric representations of these two events are completely identical, so they are still grouped into the same category. However, the geometric representation correlation between traffic light status and both road construction and traffic accidents is 0, so traffic light status is no longer grouped with road construction and traffic accidents, but is treated as a separate category.

[0206] In summary, the final classification results for the above five types of events are as follows: weather conditions and road conditions are grouped into one category, road construction and traffic accidents are grouped into another category, and traffic light conditions are classified as a separate category.

[0207] It should be noted that in this application, different classification dimensions are used to classify the above m types of dynamic events, which can yield different classification results. Therefore, under the premise of satisfying certain usability, a certain degree of grayness and flexibility is allowed in the classification of some events.

[0208] S102. Based on the fact that the first dynamic event is the first type of dynamic event, determine that the first dynamic event belongs to the first group.

[0209] In a specific embodiment, after classifying dynamic events using the above method, the classification results can be saved for later retrieval. The classification results can be stored in the form of a dynamic event classification table or in a hierarchical indexed tree storage structure.

[0210] Optionally, to facilitate storage and retrieval, each dynamic event can be encoded before storage. This allows for efficient searching through encoding. It's important to note that if a hierarchical indexed tree structure is used to store dynamic events, both the individual dynamic events and their respective categories can be encoded for easier querying. Furthermore, there's a mapping between the category codes and the codes of the dynamic events or subcategories they encompass. For example, suppose dynamic events are stored in three layers: the first layer includes multiple coarse-grained dynamic event categories; the second layer includes multiple fine-grained dynamic event categories within each of these coarse-grained categories; and the third layer includes the dynamic events within each of these fine-grained categories. Then, the dynamic event categories in the first and second layers are encoded, and the dynamic events in the third layer are also encoded. The codes of the fine-grained categories in the second layer map to the codes of their respective coarse-grained categories in the first layer, and the codes of the dynamic events in the third layer map to the codes of their respective categories in the second layer. When querying dynamic events, the corresponding events can be quickly found through the mapping relationship between the codes of each layer.

[0211] Furthermore, based on the aforementioned grouping, each dynamic event can be associated with a corresponding layer. For example, similar dynamic events can be associated with the same layer, while different dynamic events can be associated with different layers. This allows for the efficient and orderly representation of each dynamic event on the high-precision map through different layers. For instance, among the m types of dynamic events mentioned above are a second dynamic event and a third dynamic event. The second dynamic event and the first dynamic event belong to the first group among the aforementioned groupings, while the third dynamic event belongs to the second group. Therefore, the first and second dynamic events can be associated with the first layer of the dynamic information layer in the high-precision map, and the third dynamic event can be associated with the second layer of the dynamic information layer in the high-precision map. The first and second layers are two different dynamic information layers in the high-precision map. For example, the first dynamic event could be rain, the second dynamic event could be snow, and the third dynamic event could be a traffic light status. Figure 2 As you can see, rain and snow belong to the sunny / cloudy / rainy / snow group, while traffic light status belongs to its own group, namely the traffic light status group. Therefore, rain and snow can be represented on the same layer, while traffic light status can be represented on another layer.

[0212] Optionally, to facilitate the management of each layer, they can also be identified by codes. Each layer code can uniquely identify a layer, and the association between layers and dynamic events can also be achieved through the association between codes.

[0213] Specifically, in the practical application of high-precision maps, the server can acquire various dynamic events in real time. Upon acquiring these dynamic events, it needs to know their respective categories in order to retrieve the corresponding information. For example, assuming that different dynamic events are associated with their specific representation in the high-precision map, knowing the category of the dynamic event allows for quick retrieval of the corresponding representation, thus enabling the dynamic events to be quickly expressed in the high-precision map and provided to map users such as autonomous vehicles or drivers for reference.

[0214] Specifically, assuming the first dynamic event obtained above belongs to the first type of dynamic event, the specific category to which this first type of dynamic event belongs can be determined by querying the stored dynamic event classification table or hierarchical index structure, thereby allowing further retrieval of corresponding related information. For example, with... Figure 3 Taking the classification results shown as an example, assuming the first dynamic event obtained is snowfall, the server can determine through comparison that the snowfall event belongs to the weather state dynamic event category, and weather state belongs to the environmental information group. Therefore, the server can determine that the snowfall event belongs to the environmental information group. Furthermore, the server can find the information of the high-precision map layer where the snowfall event is located within the environmental information group and update the snowfall status in that layer.

[0215] In one possible implementation, if a new dynamic event occurs that is not included in the aforementioned m types of dynamic events, then this new dynamic event can be used together with the aforementioned m types of dynamic events. Figure 1 The classification method described is used to classify the data, and the new classification results are obtained and stored.

[0216] Alternatively, in another possible implementation, when a new dynamic event occurs that is not included in the aforementioned m types of dynamic events, the category to which the new dynamic event belongs can be determined by calculating the correlation between the new dynamic event and each category in the classification results, based on the results of the classification of the m types of dynamic events. Then, the new dynamic event can be assigned to its corresponding category. Optionally, the new dynamic event can be encoded in the same way, and then the encoded information can be mapped to the encoding of its category.

[0217] In summary, compared to existing technologies, this application classifies various dynamic events used in high-precision maps from multiple dimensions, achieving a more rational classification. This results in a clear and reasonable storage structure for various dynamic events, facilitating management and saving resource costs while storing and managing dynamic events of high-precision maps, thus improving storage and management efficiency. Furthermore, different categories of dynamic events have different priorities. The rational dynamic event categories obtained through this application help to quickly determine the priority of dynamic events. Therefore, when data transmission bandwidth is limited, higher-priority data is prioritized for transmission to users of high-precision maps, enabling faster and more effective operational responses. For vehicle driving, this further improves driving safety and stability.

[0218] The foregoing mainly describes the dynamic event classification method provided in the embodiments of this application. It is understood that each device, in order to achieve the corresponding functions, includes hardware structures and / or software modules for executing each function. Those skilled in the art should readily recognize that, based on the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein, this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed by hardware or by computer software driving hardware 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.

[0219] This application embodiment can divide the device into functional modules according to the above method example. For example, each function can be divided into its own functional module, or two or more functions can be integrated into one module. The integrated module can be implemented in hardware or as a software functional module. It should be noted that the module division in this application embodiment is illustrative and only represents one logical functional division. In actual implementation, there may be other division methods.

[0220] When dividing each function into modules according to its corresponding function. Figure 4 A schematic diagram of a possible logical structure of the device is shown. This device could be the aforementioned server, a chip within the server, or a processing system within the server, etc. The device 400 includes an acquisition unit 401 and a grouping unit 402. Wherein:

[0221] Acquisition unit 401 is used to acquire at least one degree of correlation between m types of dynamic events based on n dimensions. Acquisition unit 401 can perform... Figure 1 The acquisition operation described in step 101 shown.

[0222] Grouping unit 402 is configured to group the m types of dynamic events into multiple groups based on the at least one correlation degree; the m types of dynamic events are events that cause changes in the dynamic information layer in the high-precision map, where n and m are integers greater than 1; wherein the first type of dynamic event among the m types of dynamic events belongs to the first group among the multiple groups; the grouping unit 402 can perform... Figure 1 The grouping operation described in step 101 shown.

[0223] In one possible implementation, the device 400 further includes a determining unit, configured to determine, based on the acquired first dynamic event being the first type of dynamic event, that the first dynamic event belongs to the first group. The determining unit can perform... Figure 1 The operation described in step 102 shown.

[0224] In one possible implementation, the above n dimensions include at least two of the following dimensions: the fine-grained dynamic event range covered by the event, the meaning of the event, the geometric representation of the event, the impact of the event, the target of the event, the source of the event, the priority level of the event, or the update frequency of the event status.

[0225] In one possible implementation, the acquisition unit 401 is specifically used for:

[0226] Obtain the pairwise correlation between each of the m types of dynamic events in each of the n dimensions;

[0227] The correlation degree is obtained by combining the correlation degree of each dimension.

[0228] In one possible implementation, the grouping unit 402 is specifically used for:

[0229] Each of the at least one correlation degree is compared with a first numerical range;

[0230] The two dynamic events corresponding to the correlation within the first numerical range are grouped into the same group.

[0231] In one possible implementation, the acquisition unit 401 is specifically used to: obtain the first correlation degree between each pair of the m dynamic events in the first dimension;

[0232] The aforementioned grouping unit 402 is specifically used to: group the m types of dynamic events according to the first correlation degree to obtain a first plurality of groups, the first plurality of groups including a second group, the second group including p dynamic events among the m types of dynamic events, where p is a positive integer greater than 1;

[0233] The acquisition unit 401 is further used to: obtain the second correlation degree between each pair of the p kinds of dynamic events in the second dimension, wherein the second dimension is the dimension with a lower priority than the first dimension among the n dimensions;

[0234] The grouping unit 402 is also specifically used to: group the p dynamic events according to the second correlation degree to obtain a second plurality of groups.

[0235] In one possible implementation, the device 400 further includes:

[0236] The association unit is used to associate the above m types of dynamic events with multiple layers of the above high-precision map according to the above multiple groups.

[0237] Figure 4 The specific operation and beneficial effects of each unit in the device 400 shown can be found in the above description. Figure 5 The description of the method and its possible implementations is omitted here.

[0238] Figure 5 The diagram illustrates a possible hardware structure of the device provided in this application. This device can be the first network device in the method described in the above embodiments. The device 500 includes a processor 501, a memory 502, and a communication port 503. The processor 501, communication port 503, and memory 502 can be interconnected or connected to each other via a bus 504.

[0239] For example, memory 502 is used to store computer programs and data of device 500. Memory 502 may include, but is not limited to, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), or compact disc read-only memory (CD-ROM).

[0240] In implementation Figure 4 In the case of the illustrated embodiment, execution Figure 4 The software or program code required for the function of all or part of the units is stored in memory 502.

[0241] In implementation Figure 4 In the embodiment, if the software or program code required for the function of some units is stored in memory 502, then in addition to calling the program code in memory 502 to implement some functions, processor 501 can also cooperate with other components (such as communication port 503) to complete the task. Figure 4 Other functions described in the embodiments (such as the function of receiving data).

[0242] There can be multiple communication ports 503, which are used to support device 500 in communication, such as receiving or sending data or signals.

[0243] For example, processor 501 may be a central processing unit, a general-purpose processor, a digital signal processor, an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. The processor may also be a combination that implements computational functions, such as a combination of one or more microprocessors, a combination of a digital signal processor and a microprocessor, etc. Processor 501 can be used to read the program stored in the aforementioned memory 502 and execute the aforementioned... Figure 1 The method described herein, and possible implementations thereof. For example, the processor 501 may perform the following operations:

[0244] At least one correlation degree is obtained between m types of dynamic events based on n dimensions, and the m types of dynamic events are grouped into multiple groups based on the at least one correlation degree; the m types of dynamic events are events that cause changes in the dynamic information layer in the high-precision map, and n and m are integers greater than 1.

[0245] In one possible implementation, the processor 501 described above can perform the following operations:

[0246] At least one correlation degree is obtained between m types of dynamic events based on n dimensions. The m types of dynamic events are then grouped into multiple groups based on the at least one correlation degree. The m types of dynamic events are events that cause changes in the dynamic information layer in the high-precision map. The n and m are integers greater than 1. The first type of dynamic event among the m types of dynamic events belongs to the first group among the multiple groups. Based on the fact that the first dynamic event is the first type of dynamic event, it is determined that the first dynamic event belongs to the first group.

[0247] Figure 5 The specific operations performed by the device 500 and its beneficial effects can be found in the above description. Figure 1 The description of the method and its possible implementations is omitted here.

[0248] This application also provides an apparatus including a processor, a communication port, and a memory, configured to perform the methods described in any of the above embodiments and their possible embodiments.

[0249] In one possible implementation, the device is a chip or a system on a chip (SoC).

[0250] This application also provides a computer-readable storage medium storing a computer program that is executed by a processor to implement the methods described in any of the above embodiments and their possible embodiments.

[0251] This application also provides a computer program product, which, when read and executed by a computer, will execute the methods described in any of the above embodiments and their possible embodiments.

[0252] This application also provides a computer program that, when executed on a computer, will enable the computer to implement the methods described in any of the above embodiments and their possible embodiments.

[0253] In summary, compared with existing technologies, this application classifies various dynamic events used in high-precision maps from multiple dimensions, which can achieve more reasonable classification, thereby saving resource consumption and improving storage and management efficiency when storing and managing dynamic events of high-precision maps.

[0254] In this application, the terms "first," "second," etc., are used to distinguish identical or similar items that have substantially the same function and purpose. It should be understood that there is no logical or temporal dependency between "first," "second," and "nth," nor does it limit the quantity or order of execution. It should also be understood that although the following description uses the terms "first," "second," etc., to describe various elements, these elements should not be limited by the terms. These terms are merely used to distinguish one element from another. For example, without departing from the various examples described, a first image can be referred to as a second image, and similarly, a second image can be referred to as a first image. Both the first image and the second image can be images, and in some cases, they can be separate and distinct images.

[0255] It should also be understood that, in the various embodiments of this application, the sequence number of each process 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.

[0256] It should also be understood that the term “comprising” (also referred to as “includes”, “including”, “comprises” and / or “comprising”) as used in this specification specifies the presence of the stated features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0257] It should also be understood that the phrases "an embodiment," "an embodiment," and "a possible implementation" used throughout the specification mean that a specific feature, structure, or characteristic related to an embodiment or implementation is included in at least one embodiment of this application. Therefore, the phrases "in an embodiment," "an embodiment," or "a possible implementation" appearing throughout the specification do not necessarily refer to the same embodiment. Furthermore, these specific features, structures, or characteristics can be combined in any suitable manner in one or more embodiments.

[0258] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A dynamic event classification method, characterized in that, The method includes: At least one correlation degree is obtained between m types of dynamic events based on n dimensions, and the m types of dynamic events are grouped into multiple groups based on the at least one correlation degree; the at least one correlation degree is obtained based on the pairwise correlation degree between the m types of dynamic events; the m types of dynamic events are events that cause changes in dynamic information layers in high-precision maps, and n and m are integers greater than 1; wherein, the first type of dynamic event among the m types of dynamic events belongs to the first group among the multiple groups; Based on the fact that the first dynamic event is the first type of dynamic event, it is determined that the first dynamic event belongs to the first group; The m types of dynamic events are associated with multiple layers of the high-precision map based on the multiple groupings; The n dimensions include at least two of the following dimensions: the fine-grained dynamic event scope covered by the event, the meaning of the event, the geometric representation of the event, the impact of the event, the target of the event, the source of the event, the priority level of the event, or the update frequency of the event status.

2. The method according to claim 1, characterized in that, The step of obtaining at least one correlation degree among m types of dynamic events based on n dimensions includes: Obtain the pairwise correlation degree between the m types of dynamic events in each of the n dimensions; The at least one correlation degree is obtained by combining the correlation degrees on each dimension.

3. The method according to claim 1 or 2, characterized in that, The process of grouping the m types of dynamic events according to at least one correlation degree to obtain multiple groups includes: Each of the at least one degree of correlation is compared with a first numerical range; The two dynamic events corresponding to the correlation within the first numerical range will be grouped into the same group.

4. The method according to claim 1 or 2, characterized in that, The n dimensions are prioritized. At least one correlation degree is obtained between m types of dynamic events based on the n dimensions. The m types of dynamic events are then grouped according to the at least one correlation degree to obtain multiple groups, including: Obtain the first degree of correlation between each pair of the m dynamic events mentioned in the first dimension; The m types of dynamic events are grouped according to the first correlation degree to obtain a first plurality of groups. The first plurality of groups include a second group, and the second group includes p types of dynamic events among the m types of dynamic events, where p is a positive integer greater than 1. Obtain the second correlation degree between each pair of the p dynamic events in the second group on the second dimension, where the second dimension is the dimension with a lower priority than the first dimension among the n dimensions; The p dynamic events are grouped according to the second correlation degree to obtain a second plurality of groups.

5. A dynamic event classification device, characterized in that, The device includes: The acquisition unit is used to acquire at least one correlation degree between m kinds of dynamic events according to n dimensions, wherein the at least one correlation degree is obtained based on the pairwise correlation degree between the m kinds of dynamic events; A grouping unit is used to group the m types of dynamic events into multiple groups based on the at least one correlation degree; the m types of dynamic events are events that cause changes in dynamic information layers in a high-precision map, and n and m are integers greater than 1; wherein, the first type of dynamic event among the m types of dynamic events belongs to the first group among the multiple groups; The determining unit is configured to determine that the first dynamic event belongs to the first group based on the fact that the acquired first dynamic event is the first type of dynamic event; The m types of dynamic events are associated with multiple layers of the high-precision map based on the multiple groupings; The n dimensions include at least two of the following dimensions: the fine-grained dynamic event scope covered by the event, the meaning of the event, the geometric representation of the event, the impact of the event, the target of the event, the source of the event, the priority level of the event, or the update frequency of the event status.

6. The device according to claim 5, characterized in that, The acquisition unit is specifically used for: Obtain the pairwise correlation degree between the m types of dynamic events in each of the n dimensions; The at least one correlation degree is obtained by combining the correlation degrees on each dimension.

7. The device according to claim 5 or 6, characterized in that, The grouping unit is specifically used for: Each of the at least one degree of correlation is compared with a first numerical range; The two dynamic events corresponding to the correlation within the first numerical range will be grouped into the same group.

8. The device according to claim 5 or 6, characterized in that, The acquisition unit is specifically used to: obtain the first correlation degree between each pair of the m dynamic events in the first dimension; The grouping unit is specifically used to: group the m types of dynamic events according to the first correlation degree to obtain a first plurality of groups, the first plurality of groups including a second group, the second group including p types of dynamic events among the m types of dynamic events, where p is a positive integer greater than 1; The acquisition unit is further configured to: obtain the second correlation degree between each pair of the p kinds of dynamic events in the second group on the second dimension, wherein the second dimension is a dimension with a lower priority than the first dimension among the n dimensions; The grouping unit is further configured to: group the p dynamic events according to the second correlation degree to obtain a second plurality of groups.

9. A dynamic event classification device, characterized in that, The device includes a processor and a memory, wherein the memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory, causing the device to perform the method as described in any one of claims 1 to 4.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that is executed by a processor to implement the method of any one of claims 1 to 4.

Images