Data processing method and device for crowdsourcing map construction, equipment and medium

Abstract

The invention discloses a data processing method and device for map construction, equipment and a medium, and relates to the technical field of crowdsourcing maps, and the method comprises the steps: obtaining a local vectorization semantic map of a vehicle driving on a target road section, and a historical crowdsourcing map corresponding to the target road section; the local vectorization semantic map is constructed based on sensor data collected by a sensor of the vehicle, and the sensor data comprises positioning data, sensing data and sensor parameters of the vehicle; obtaining a first confidence coefficient and a second confidence coefficient based on matching alignment processing of the historical crowdsourcing map and the local vectorization semantic map; based on the first confidence coefficient and the second confidence coefficient, determining whether to collect a processing strategy of target data for constructing a global crowdsourcing map, the target data including a local vector semantic map and positioning data; therefore, the data for crowdsourcing map construction can be accurately, economically and efficiently collected, the data volume is reduced, and the crowdsourcing map construction efficiency is improved.

Description

Technical Field

[0001] This application relates to, but is not limited to, the field of crowdsourcing map technology, and in particular, to a data processing method, apparatus, device, and medium for constructing a crowdsourcing map. Background Art

[0002] A crowdsourcing map is a high-precision map that uses low-cost sensor hardware installed on different vehicles to collect a large amount of road information collected by different vehicles at different times for semantic mapping and data fusion, and then performs data aggregation for production. It should be noted that since the data fusion and aggregation processes need to process a large amount of data, this process is usually executed in a cloud server, and a large amount of collected data needs to be transmitted to the cloud. However, the above method has at least the problem of a large amount of data. Summary of the Invention

[0003] One of the objectives of this application is to provide a data processing method, apparatus, device, and medium for constructing a crowdsourcing map to solve the problem of a large amount of data in the prior art.

[0004] To solve the above problems, the technical solution of this application is implemented as follows:

[0005] In a first aspect, this application provides a data processing method for constructing a crowdsourcing map, and the method includes:

[0006] Obtain a local vectorized semantic map when the vehicle is driving on a target road section, and a historical crowdsourcing map corresponding to the target road section, where the local vectorized semantic map is constructed based on sensor data collected by the vehicle's sensors, and the sensor data includes the vehicle's positioning data, perception data, and sensor parameters;

[0007] Based on the matching and alignment processing of the historical crowdsourcing map and the local vectorized semantic map, obtain a matching confidence; where the matching confidence includes a first confidence corresponding to the local vectorized semantic map and a second confidence corresponding to the historical crowdsourcing map;

[0008] Based on the first confidence and the second confidence, determine a processing strategy for whether to collect target data for constructing a global crowdsourcing map, where the target data includes the local vector semantic map and the positioning data.

[0009] According to the above technical means, in the embodiment of the present application, after obtaining the local vectorized semantic map corresponding to the target road section and the historical crowdsourcing map corresponding to the target road section, based on the matching and alignment processing of the historical crowdsourcing map and the local vectorized semantic map, the first confidence level corresponding to the local vectorized semantic map and the second confidence level corresponding to the historical crowdsourcing map can be obtained; furthermore, based on the first confidence level and the second confidence level, a processing strategy for determining whether to collect target data for constructing a global crowdsourcing map can be determined; in this way, based on the first confidence level and the second confidence level, data for crowdsourcing map construction can be collected accurately, economically and efficiently, the amount of data can be reduced, and the efficiency of crowdsourcing map construction can be improved.

[0010] Further, the local vectorized semantic map is constructed based on the ego-vehicle coordinate system, and the historical crowdsourcing map is constructed based on the world coordinate system. The matching and alignment processing based on the historical crowdsourcing map and the local vectorized semantic map to obtain the matching confidence level includes: determining the mapping relationship between the ego-vehicle coordinate system and the world coordinate system; based on the mapping relationship, converting the first map into a reference map with the same coordinate system as the second map, where the first map is one of the historical crowdsourcing map and the local vectorized semantic map, and the second map is the other of the historical crowdsourcing map and the local vectorized semantic map; obtaining the first object information of the first semantic map object in the reference map and the second object information of the second semantic map object in the second map; based on the first object information and the second object information, performing an association matching process on the first semantic map object in the reference map and the second semantic map object in the second map to obtain a matching result; based on the matching result, determining the first confidence level and the second confidence level.

[0011] According to the above technical means, the embodiment of the present application can align the local vectorized semantic map and the historical crowdsourcing map in the same coordinates. In the same coordinate system, the first object information of the first semantic map object in the reference map and the second object information of the second semantic map object in the second map can be obtained; further, based on the first object information and the second object information, an association matching is performed on the first semantic map object and the second semantic map object to obtain the first confidence level and the second confidence level. In this way, the accuracy of the object information and the confidence level is ensured, and thus it is possible to more accurately determine whether to collect data for crowdsourcing map construction.

[0012] Further, the object information includes the spatial position, geometric attributes, and semantic attributes of the semantic map object. Based on the first object information and the second object information, performing an association matching process on the semantic map objects in the reference map and the semantic map objects in the second map to obtain a matching result, including: determining a determination result of whether there are first and second semantic map objects with the same semantic attributes based on the first object information and the second object information; if the determination result indicates that there are first and second semantic map objects with the same semantic attributes, performing an association matching process on the first and second semantic map objects with the same semantic attributes based on the first geometric attributes and the first spatial position of the first semantic map object with the same semantic attributes, and the second geometric attributes and the second spatial position of the second semantic map object with the same semantic attributes to obtain the matching result; if the determination result indicates that there are no first and second semantic map objects with the same semantic attributes, obtaining a matching result indicating that the association matching of the first semantic map object or the second semantic map object fails.

[0013] According to the above technical means, the embodiments of the present application can determine based on whether the semantic attributes of the semantic map objects in the reference map and the second map are the same. In the case where the semantic attributes are the same, determine the association matching of the first semantic map object and the second semantic map object to improve the accuracy of the matching; in the case where the semantic attributes are different, it indicates that the association matching of the first semantic map object and the second semantic map object fails, and there is no need to perform an association matching determination based on other parameters in the object information. In this way, the processing efficiency is improved.

[0014] Further, the performing an association matching process on the first and second semantic map objects with the same semantic attributes based on the first geometric attributes and the first spatial position of the first semantic map object with the same semantic attributes, and the second geometric attributes and the second spatial position of the second semantic map object with the same semantic attributes to obtain the matching result includes: for the first and second semantic map objects with the same semantic attributes, if the relationship between the first spatial position and the second spatial position and the relationship between the first geometric attributes and the second geometric attributes satisfy the association matching condition, obtaining a matching result indicating that the association matching between the first semantic map object and the second semantic map object is successful; if the relationship between the first spatial position and the second spatial position and the relationship between the first geometric attributes and the second geometric attributes do not satisfy the association matching condition, obtaining a matching result indicating that the association matching between the first semantic map object and the second semantic map object fails.

[0015] According to the above technical means, in the case where the semantic attributes of two semantic map objects are the same, the embodiments of the present application can, based on the spatial positions and geometric attributes in the object information, determine the association between the first semantic map object and the second semantic map object again. When the spatial positions and geometric attributes of the first semantic map object and the second semantic map object meet the association matching conditions, it is determined that the first semantic map object and the second semantic map object are associated and matched. In this way, by setting the association matching conditions based on the object information, the accuracy of the map element attributes and the spatial position accuracy are ensured.

[0016] Further, determining the first confidence level and the second confidence level according to the matching result includes: according to the matching result, determining the first quantity of the first semantic map object in the reference map, the second quantity of the second semantic map object in the second map, and the third quantity of the semantic map objects that are successfully associated and matched among the first semantic map object and the second semantic map object; based on the first quantity, the second quantity, and the third quantity, determining the first confidence level and the second confidence level.

[0017] According to the above technical means, the embodiments of the present application can calculate the first confidence level and the second confidence level based on the quantity of the semantic map objects that are successfully associated and matched, the quantity of the first semantic map objects in the reference map, and the quantity of the second semantic map objects in the second map, and can more accurately determine whether to collect the target data.

[0018] Further, determining the first confidence level and the second confidence level based on the first quantity, the second quantity, and the third quantity includes: determining the ratio of the third quantity to the first quantity as the first confidence level; determining the ratio of the third quantity to the second quantity as the second confidence level.

[0019] According to the above technical means, the embodiments of the present application can more accurately determine whether to collect the target data by calculating the proportions of the semantic map objects that are successfully associated and matched in the reference image and the second image.

[0020] Further, the processing strategy for determining whether to collect the target data for constructing the global crowdsourcing map based on the first confidence level and the second confidence level includes: if the first confidence level meets the first confidence level condition and the second confidence level meets the second confidence level condition, determining the first processing strategy of not collecting the target data; if the first confidence level does not meet the first confidence level condition, and / or, the second confidence level does not meet the second confidence level condition, determining the second processing strategy of collecting the target data, where the second processing strategy includes: encrypting the target data and storing the encrypted target data.

[0021] According to the above technical means, in the embodiments of the present application, when both the first confidence level and the second confidence level meet the confidence level conditions, the target data may not be collected, and when the first confidence level and / or the second confidence level do not meet the confidence level conditions, the target data may be collected. In this way, the data for crowdsourcing map construction can be collected accurately, economically and efficiently, and the crowdsourcing map construction efficiency can be improved.

[0022] In a second aspect, the present application provides a data processing device for crowdsourcing map construction, and the device includes:

[0023] An obtaining module, configured to obtain a local vectorized semantic map when the vehicle travels on a target road section, and a historical crowdsourcing map corresponding to the target road section, where the local vectorized semantic map is constructed based on sensor data collected by sensors of the vehicle, and the sensor data includes positioning data, perception data, and sensor parameters of the vehicle;

[0024] A processing module, configured to obtain a matching confidence level based on matching and alignment processing of the historical crowdsourcing map and the local vectorized semantic map; where the matching confidence level includes a first confidence level corresponding to the local vectorized semantic map and a second confidence level corresponding to the historical crowdsourcing map;

[0025] A determining module, configured to determine a processing strategy for determining whether to collect target data for constructing a global crowdsourcing map based on the first confidence level and the second confidence level, where the target data includes the local vector semantic map and the positioning data.

[0026] In a third aspect, the present application provides a computer device, and the computer device includes a processor and a memory;

[0027] The memory stores a computer program that can be run on the processor, and when the processor executes the computer program, it implements the data processing method for crowdsourcing map construction described in the first aspect above.

[0028] In a fourth aspect, the present application provides a storage medium, and the storage medium stores one or more computer programs, and the one or more computer programs can be executed by one or more processors to implement some or all of the steps in the data processing method for crowdsourcing map construction described above. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] The drawings here are incorporated into the specification and form a part of this specification. These drawings show embodiments that conform to the present application and are used together with the specification to explain the technical solutions of the present application.

[0030] Figure 1Schematic flowchart of an optional data processing method for crowdsourcing map construction provided by an embodiment of the present application;

[0031] Figure 2 Schematic flowchart of an optional data processing method for crowdsourcing map construction provided by an embodiment of the present application;

[0032] Figure 3 Schematic flowchart of an optional data processing method for crowdsourcing map construction provided by an embodiment of the present application;

[0033] Figure 4 Schematic flowchart of an optional data processing method for crowdsourcing map construction provided by an embodiment of the present application;

[0034] Figure 5 Schematic flowchart of an optional data processing method for crowdsourcing map construction provided by an embodiment of the present application;

[0035] Figure 6 Schematic structural diagram of an optional data processing device for crowdsourcing map construction provided by an embodiment of the present application;

[0036] Figure 7 Schematic structural diagram of a computer device provided by an embodiment of the present application. Detailed implementation manners

[0037] In order to make the objectives, technical solutions, and advantages of the present application clearer, the technical solutions of the present application will be further elaborated in detail below with reference to the accompanying drawings and embodiments. The described embodiments should not be construed as limitations on the present application. All other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the scope of protection of the present application.

[0038] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments. However, it can be understood that "some embodiments" can be the same subset or different subsets of all possible embodiments, and can be combined with each other without conflict. The terms "first / second / third" involved are only used to distinguish similar objects and do not represent a specific order for the objects. It can be understood that "first / second / third" can be interchanged with a specific order or sequence when permitted, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein.

[0039] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which this application belongs. The terms used herein are only for the purpose of describing this application and are not intended to limit this application.

[0040] An embodiment of the present application provides a data processing method for crowdsourcing map construction, which can be executed by a processor of a computer device. Herein, the computer device may refer to a device with data processing capabilities such as a server, a laptop, a tablet computer, a desktop computer, a smart TV, a set-top box, a mobile device (such as a mobile phone, a portable video player, a personal digital assistant, a dedicated messaging device, a portable game device), etc. In some embodiments, the computer device may be an in-vehicle terminal device. Herein, the in-vehicle terminal device may be a terminal device deployed on a vehicle, which is communicatively connected to the vehicle and can be used independently of the vehicle or integrated into the vehicle control system. The present application does not make any limitation thereto.

[0041] Referring to Figure 1 , Figure 1 FIG. is a schematic flowchart of the implementation of a data processing method for crowdsourcing map construction provided by an embodiment of the present application. This method can be executed by a controller of a vehicle. Herein, the steps shown in Figure 1 will be described:

[0042] Step 101: Obtain a local vectorized semantic map when the vehicle is driving on a target road section, and a historical crowdsourcing map corresponding to the target road section. Herein, the local vectorized semantic map is constructed based on sensor data collected by the vehicle's sensors, and the sensor data includes the vehicle's positioning data, perception data, and sensor parameters.

[0043] In an embodiment of the present application, the local vectorized semantic map can be at the vehicle end. The local vectorized semantic map can be constructed based on sensor data collected by the vehicle's sensors. It should be noted that when constructing the local vectorized semantic map, the computer device can be based on the vehicle's own coordinate system and draw and construct based on the sensor data.

[0044] In an embodiment of the present application, the vehicle's sensors include but are not limited to mainly including the Global Navigation Satellite System (GNSS), Inertial Measurement Unit (IMU), wheel speed meter or vehicle speed meter, image acquisition sensors such as cameras, Light Detection and Ranging (LiDAR), or other sensors and their combinations.

[0045] In an embodiment of the present application, the positioning data may include but is not limited to the vehicle's spatial position, the vehicle's attitude information, the satellite positioning status information for obtaining the vehicle's spatial position, and the vehicle's wheel speed information.

[0046] Among them, the vehicle's spatial position can be the position of the vehicle obtained by the vehicle through GNSS.

[0047] Among them, the attitude information of the vehicle can be the attitude of the vehicle obtained by the IMU, such as the head orientation information. Based on the attitude information of the vehicle, the computer device can calculate the spatial positions of the various objects included in the image data collected by the image sensor, so as to construct a local vectorized semantic map.

[0048] Among them, the satellite positioning status information is used to characterize the quality of the state of the GNSS satellites. Exemplarily, if the satellite signal of the GNSS is relatively good, the spatial position accuracy of the vehicle located by the GNSS is accurate; if the satellite signal of the GNSS is not good, the spatial position of the vehicle located by the GNSS is inaccurate.

[0049] Among them, the wheel speed information is the rotational speed or driving speed of the vehicle obtained by the vehicle through a wheel speed meter or a vehicle speed meter, and the wheel speed information is used to determine the spatial position of the vehicle. Exemplarily, when the image data collected by the vehicle's image sensor is not synchronized with the positioning data of the vehicle obtained by the GNSS, it is necessary to determine the positioning data of the vehicle at the first moment according to the wheel speed information of the vehicle, the image data collected by the image sensor at the first moment, and the positioning data of the vehicle obtained by the GNSS at two adjacent second moments, where the first moment is located between the two adjacent second moments.

[0050] In the embodiments of the present application, the perception data includes the semantic segmentation result obtained by performing semantic segmentation processing on the image data collected by the image sensor, and the target detection result obtained by performing target detection on the image data.

[0051] Among them, performing semantic segmentation processing on the image data collected by the image sensor can be understood as obtaining a map element classification result through a deep learning method for semantic segmentation.

[0052] Exemplarily, the computer device inputs the image data collected by each camera into the map feature perception model. The map feature perception model adopts a multi-task deep neural network, which mainly consists of a backbone, a neck, and a task head. The backbone mainly completes the extraction of general features, including but not limited to using Convolutional Neural Networks (CNN), Transformer, or a combination of both as the basic units for feature extraction. Through the serial and parallel stacking of the basic units, a backbone network with stronger feature abstraction and learning capabilities is obtained; the task head part is mainly divided into object detection and semantic segmentation tasks for different map features. Through the division of different tasks, different map features can be detected more precisely. The neck part mainly extracts corresponding finer features through different feature pyramids according to the differences of the task head. Here, the map feature perception results include the map features obtained through semantic segmentation and the map features obtained through object detection; among them, the map features or map elements obtained through semantic segmentation include but not limited to: road boundaries, road surface areas, lane lines, crosswalks, road surface arrows, and poles; it should be noted that the road surface area includes but not limited to the diversion area, stop line area, and drivable area, etc.; lane lines include but not limited to single solid lines, single dashed lines, double solid lines, double dashed lines, etc.; road surface arrows include but not limited to straight arrows, left-turn arrows, right-turn arrows, etc.; poles include street lamp poles, traffic light poles, and sign poles, etc. Among them, the map features obtained through object detection include but not limited to signboards, signal lights, and road surface arrows; it should be noted that signboards include but not limited to traffic signboards, speed limit signs, height limit signs; signal lights include traffic signal lights; road surface arrows include but not limited to straight arrows, left-turn arrows, right-turn arrows, etc.

[0053] Among them, the sensor parameters include the installation parameters of the combined inertial navigation, the internal and external parameters of the image sensor for collecting image data. The internal parameters of the image sensor include but not limited to the field of view angle, focal length, and aperture, etc. The external parameters of the image sensor can be understood as the fixed conversion relationship between the image sensor coordinate system and the vehicle coordinate system; it should be noted that the image data obtained by the sensor is all based on the sensor coordinate system. Usually, the sensor is fixed on the carrier, and the carrier is located on the vehicle and moves as a rigid body. Therefore, there is a fixed conversion relationship between the sensor coordinate system and the vehicle coordinate system, that is, the external parameter.

[0054] In the embodiments of the present application, the local vectorized semantic map can be constructed based on the sensor data collected by the vehicle's sensors. For example, the computer device can use the Simultaneous Localization and Mapping (SLAM) technology to perform semantic mapping based on the above-mentioned sensor parameters including positioning data, perception data, and sensor parameters to obtain the local vectorized semantic map. Of course, it can also be that the computer device uses the Bird’s-eye-view (BEV) mapping technology to perform semantic mapping based on the above-mentioned sensor parameters including positioning data, perception data, and sensor parameters to obtain the local vectorized semantic map. In this regard, the present application does not make specific limitations.

[0055] In the embodiments of the present application, the historical crowdsourced map can be the crowdsourced map corresponding to the target road section. It should be noted that the historical crowdsourced map can be the crowdsourced map corresponding to the target road section in the latest crowdsourced map.

[0056] In the embodiments of the present application, the vehicle can be an autonomous vehicle. Here, the type of the autonomous vehicle is not limited, such as vehicle type, size, and brand. The types of autonomous vehicles include but are not limited to cars, small trucks, large trucks, small buses, medium buses, and large buses.

[0057] In the embodiments of the present application, during the driving process of the vehicle, after the vehicle's sensors collect the sensor parameters including the vehicle's positioning data, perception data, and sensor parameters in real time, the computer device obtains the sensor parameters collected by the sensors when the vehicle is driving on the target road section, and uses the SLAM mapping technology or the BEV mapping technology to perform semantic mapping based on the positioning data, perception data, and sensor parameters to construct a local vectorized semantic map, thereby obtaining the local vectorized semantic map when the vehicle is driving on the target road section. Further, the computer device obtains the historical crowdsourced map corresponding to the target road section.

[0058] In some embodiments, if there is no historical crowdsourced map corresponding to the target road section in the historical crowdsourced map, the local vectorized semantic map and the positioning data are directly determined as the target data for constructing the crowdsourced map. In this way, when it is necessary to construct a crowdsourced map on the vehicle side, the encrypted target data is stored locally, and when it is necessary to construct a crowdsourced map on the cloud side, the encrypted target data is sent to the cloud. In this way, the data volume is reduced, the data transmission efficiency is improved, and the data security is ensured through encryption.

[0059] Step 102: Obtain the matching confidence based on the matching and alignment processing of the historical crowdsourced map and the local vectorized semantic map; wherein, the matching confidence includes the first confidence corresponding to the local vectorized semantic map and the second confidence corresponding to the historical crowdsourced map.

[0060] In the embodiments of the present application, the configured confidence includes a first confidence corresponding to the local vectorized semantic map and a second confidence corresponding to the historical crowdsourced map, and the first confidence is different from the second confidence.

[0061] In the embodiments of the present application, when the computer device obtains the local vectorized semantic map of the vehicle driving on the target road section and the historical crowdsourced map corresponding to the target road section, based on the matching and alignment processing of the historical crowdsourced map and the local vectorized semantic map, the first confidence corresponding to the local vectorized semantic map and the second confidence corresponding to the historical crowdsourced map are obtained, so as to determine the processing strategy for determining whether to collect the target data for constructing the crowdsourced map based on the first confidence and the second confidence.

[0062] Step 103: Determine the processing strategy for determining whether to collect the target data for constructing the global crowdsourced map based on the first confidence and the second confidence, where the target data includes the local vector semantic map and the positioning data.

[0063] In the embodiments of the present application, the processing strategy for determining whether to collect the target data for constructing the global crowdsourced map based on the first confidence and the second confidence can be implemented through the following steps:

[0064] If the first confidence meets the first confidence condition and the second confidence meets the second confidence condition, determine the first processing strategy of not collecting the target data;

[0065] If the first confidence does not meet the first confidence condition, and / or the second confidence does not meet the second confidence condition, determine the second processing strategy for the target data, where the second processing strategy includes:

[0066] Encrypt the target data and store the encrypted target data.

[0067] In the embodiments of the present application, the first confidence condition may be that the first confidence is greater than or equal to the first confidence threshold, and the first confidence threshold may be set based on an empirical value; the second confidence condition may be that the second confidence is greater than or equal to the second confidence threshold, and the second confidence threshold may be set based on an empirical value. It should be noted that the first confidence threshold and the second confidence threshold may be the same or different, and the present application does not make specific limitations on this.

[0068] In the embodiments of the present application, after the computer device obtains the first confidence level and the second confidence level through the matching and alignment process of the historical crowdsourced map and the local vectorized semantic map, if the first confidence level is greater than or equal to the first confidence level threshold and the second confidence level is greater than or equal to the second confidence level threshold, it is determined that the first confidence level meets the first confidence level condition and the second confidence level meets the second confidence level condition, indicating that the map similarity between the historical crowdsourced map and the local vectorized semantic map is relatively high, and there is no need to collect target data to update the historical crowdsourced map corresponding to the target road section; further, the computer device determines the first processing strategy of not collecting target data, thus reducing data transmission.

[0069] The embodiments of the present application provide a data processing method for crowdsourced map construction, which obtains a local vectorized semantic map when a vehicle travels on a target road section and a historical crowdsourced map corresponding to the target road section. The local vectorized semantic map is constructed based on sensor data collected by the vehicle's sensors, and the sensor data includes the vehicle's positioning data, perception data, and sensor parameters; through the matching and alignment process of the historical crowdsourced map and the local vectorized semantic map, a matching confidence level is obtained; the matching confidence level includes a first confidence level corresponding to the local vectorized semantic map and a second confidence level corresponding to the historical crowdsourced map; based on the first confidence level and the second confidence level, a processing strategy for determining whether to collect target data for constructing a global crowdsourced map is determined, where the target data includes the local vector semantic map and the positioning data; that is, after obtaining the local vectorized semantic map corresponding to the target road section and the historical crowdsourced map corresponding to the target road section, through the matching and alignment process of the historical crowdsourced map and the local vectorized semantic map, the first confidence level corresponding to the local vectorized semantic map and the second confidence level corresponding to the historical crowdsourced map are obtained; then, based on the first confidence level and the second confidence level, a processing strategy for determining whether to collect target data for constructing a global crowdsourced map is determined; thus, based on the first confidence level and the second confidence level, data for crowdsourced map construction can be accurately, economically, and efficiently collected, reducing the amount of data, lowering the cost of crowdsourced map construction, and improving the efficiency of crowdsourced map construction.

[0070] In some embodiments, the process of obtaining the matching confidence level through the matching and alignment process of the historical crowdsourced map and the local vectorized semantic map in step 102 is combined with Figure 2 described as follows:

[0071] Step 201: Determine the mapping relationship between the vehicle's coordinate system and the world coordinate system. The local vectorized semantic map is constructed based on the vehicle's coordinate system, and the historical crowdsourced map is constructed based on the world coordinate system.

[0072] In the embodiments of the present application, the local vectorized semantic map may be a map drawn based on the vehicle's ego coordinate system at the vehicle end, and the local vectorized semantic map is drawn based on the vehicle's ego coordinate system at the vehicle end.

[0073] In the embodiments of the present application, the historical crowdsourced map is drawn based on the world coordinate system, and the world coordinate system maintains a fixed relationship with the actual geographical location. For example, the Earth-Centered, Earth-Fixed (ECEF) coordinate system can be adopted.

[0074] In the embodiments of the present application, since the vehicle moves in the world, the relationship between the ego coordinate system and the world coordinate system changes over time. To perform a transformation between these two coordinate systems, a transformation matrix or transformation (usually consisting of rotation and translation), that is, a mapping relationship, is generally required. The mapping relationship between the ego coordinate system and the world coordinate system can be obtained through various sensors (such as GPS, IMU, lidar) and algorithms (such as SLAM).

[0075] Step 202: Based on the mapping relationship, convert the first map into a reference map with the same coordinate system as the coordinate system for drawing the second map, where the first map is one of the historical crowdsourced map and the local vectorized semantic map, and the second map is the other of the historical crowdsourced map and the local vectorized semantic map.

[0076] In the embodiments of the present application, converting the first map into a reference map with the same coordinate system as the coordinate system for drawing the second map based on the mapping relationship can be understood as: converting the historical crowdsourced map into a map in the ego coordinate system based on the mapping relationship, so that the converted historical crowdsourced map, that is, the reference map, is unified with the local vectorized semantic map in the ego coordinate system; or, converting the local vectorized semantic map into a map in the world coordinate system based on the mapping relationship, so that the converted local vectorized semantic map, that is, the reference map, is unified with the historical crowdsourced map in the world coordinate system.

[0077] In the embodiments of the present application, after the computer device determines the mapping relationship between the ego coordinate system and the world coordinate system, based on the mapping relationship, it converts one of the historical crowdsourced map and the local vectorized semantic map into a reference map with the same coordinate system as the other map, thereby unifying the data in the historical crowdsourced map and the local vectorized semantic map into the same coordinate reference system.

[0078] Step 203: Obtain the first object information of the first semantic map object in the reference map, and the second object information of the second semantic map object in the second map.

[0079] In the embodiments of the present application, the semantic map object may be a map feature or a map element, and the object information of the semantic map object includes the spatial position, geometric attributes, and semantic attributes of the semantic map object. Among them, the spatial position may be the position of the semantic map object in the reference map, the geometric attributes may be the geometric area and / or geometric length of the map object, and the semantic attributes may be the classification result of the map element or map feature.

[0080] In the embodiments of the present application, the reference map contains multiple first semantic map objects, and the first object information of each first semantic map object includes a first spatial position, first geometric attributes, and first semantic attributes.

[0081] In the embodiments of the present application, the second map contains multiple second semantic map objects, and the second object information of each second semantic map object includes a second spatial position, second geometric attributes, and second semantic attributes.

[0082] Step 204: Based on the first object information and the second object information, perform an association matching process on the first semantic map objects in the reference map and the second semantic map objects in the second map to obtain a matching result.

[0083] In the embodiments of the present application, the matching result includes the result of successful association matching between the first semantic map objects in the reference map and the second semantic map objects in the second map. The matching result also includes: the result of failed or unassociated matching between the first semantic map objects in the reference map and the second semantic map objects in the second map.

[0084] Step 205: Based on the matching result, determine a first confidence level and a second confidence level.

[0085] In the embodiments of the present application, after the computer device obtains the first object information of the first semantic map objects in the reference map and the second object information of the second semantic map objects in the second map, based on the first object information of the first semantic map objects and the second object information of the second semantic map objects, perform an association matching process on the first semantic map objects in the reference map and the second semantic map objects in the second map to obtain a matching result. Then, based on the matching result, determine the first confidence level corresponding to the local vectorized semantic map and the second confidence level corresponding to the historical crowdsourced map. Finally, based on the first confidence level and the second confidence level, determine the processing strategy for whether to collect the target data for constructing the global crowdsourced map.

[0086] As can be seen from the above, in the embodiments of the present application, the local vectorized semantic map and the historical crowdsourced map are unified under the same coordinates. Under the same coordinate system, the first object information of the first semantic map object in the reference map and the second object information of the second semantic map object in the second map are obtained. Further, based on the first object information and the second object information, the first semantic map object and the second semantic map object are associated and matched to obtain the first confidence level and the second confidence level. In this way, the accuracy of the object information and the confidence level is ensured, so as to more accurately determine whether to collect data for constructing the crowdsourced map.

[0087] In some embodiments, step 204 is described in conjunction with the process of performing an association matching process on the first semantic map object in the reference map and the second semantic map object in the second map based on the first object information and the second object information to obtain a matching result. Figure 3 The description is as follows:

[0088] Step 301: Based on the first object information and the second object information, determine the determination result of whether there are first semantic map objects and second semantic map objects with the same semantic attributes.

[0089] In the embodiments of the present application, the determination result includes that the first semantic attribute of the first semantic map object is the same as the second semantic attribute of the second semantic map object, and the determination result includes that the first semantic attribute of the first semantic map object is different from the second semantic attribute of the second semantic map object.

[0090] Step 302: If the determination result indicates that there are first semantic map objects and second semantic map objects with the same semantic attributes, based on the first geometric attribute and the first spatial position of the first semantic map object with the same semantic attributes, and the second geometric attribute and the second spatial position of the second semantic map object with the same semantic attributes, perform an association matching process on the first semantic map object and the second semantic map object with the same semantic attributes to obtain a matching result.

[0091] In the embodiments of the present application, if the computer device determines that there are two semantic map objects with the same semantic attributes in the reference map and the second map, it means that there may be an association between the first semantic map object in the reference map and the second semantic map object in the second map. Further, the computer device performs an association matching process on the first semantic map object and the second semantic map object with the same semantic attributes based on the first geometric attribute and the first spatial position of the first semantic map object with the same semantic attributes, and the second geometric attribute and the second spatial position of the second semantic map object with the same semantic attributes to obtain a matching result.

[0092] In some embodiments, in step 302, based on the first geometric attribute and the first spatial position of the first semantic map object with the same semantic attribute, and the second geometric attribute and the second spatial position of the second semantic map object with the same semantic attribute, the association matching process is performed on the first semantic map object and the second semantic map object with the same semantic attribute to obtain a matching result, which can be achieved through the following steps.

[0093] For the first semantic map object and the second semantic map object with the same semantic attribute, if the relationship between the first spatial position and the second spatial position, and the relationship between the first geometric attribute and the second geometric attribute satisfy the association matching condition, then a matching result indicating successful association matching between the first semantic map object and the second semantic map object is obtained.

[0094] If the relationship between the first spatial position and the second spatial position, and the relationship between the first geometric attribute and the second geometric attribute do not satisfy the association matching condition, then a matching result indicating failed association matching between the first semantic map object and the second semantic map object is obtained.

[0095] In the embodiments of the present application, the association matching condition includes: the spatial position distance is less than the first set threshold, and the geometric area difference and / or the geometric length difference is less than the second set threshold, where the geometric attribute includes geometric area and / or geometric difference; here, the first set threshold and the second set threshold can be set based on the empirical values of the semantic map object, that is, the first set threshold and the second set threshold corresponding to different semantic map objects are different.

[0096] In the embodiments of the present application, for the first semantic map object and the second semantic map object with the same semantic attribute, calculate the spatial position distance between the first spatial position of the first semantic map object and the second spatial position of the second semantic map object, and calculate the geometric attribute difference (geometric area difference and / or geometric length difference) between the first geometric attribute (geometric area and / or geometric length) and the second geometric attribute (geometric area and / or geometric length) of the second semantic map object. If the spatial position distance is less than the first set threshold and the geometric attribute difference is less than the second set threshold, it is determined that the relationship between the first spatial position and the second spatial position, and the relationship between the first geometric attribute and the second geometric attribute satisfy the association matching condition; further, a matching result indicating successful association matching between the first semantic map object and the second semantic map object with the same semantic attribute is determined. Of course, if the spatial position distance is not less than the first set threshold and / or the geometric attribute difference is not less than the second set threshold, it is determined that the relationship between the first spatial position and the second spatial position, and the relationship between the first geometric attribute and the second geometric attribute do not satisfy the association matching condition; further, a matching result indicating failed association matching between the first semantic map object and the second semantic map object with the same semantic attribute is determined.

[0097] As described above, in the embodiment of the present application, when the semantic attributes of two semantic map objects are the same, based on the spatial position and geometric attributes in the object information, the association between the first semantic map object and the second semantic map object can be determined again. When the spatial position and geometric attributes of the first semantic map object and the second semantic map object meet the association matching conditions, it is determined that the first semantic map object and the second semantic map object are associated and matched. In this way, by setting the association matching conditions based on the object information, the accuracy of the map element attributes and the spatial position accuracy are ensured.

[0098] Step 303: If the determination result indicates that there are no first semantic map object and second semantic map object with the same semantic attributes, a matching result of failed association matching of the first semantic map object or the second semantic map object is obtained.

[0099] In the embodiment of the present application, if the determination result indicates that there are no first semantic map object and second semantic map object with the same semantic attributes, it means that there is no map object in the second map with the same first semantic attribute as the first semantic map object in the reference map, and / or there is no map object in the reference map with the same second semantic attribute as the second semantic map object in the second map.

[0100] In the embodiment of the present application, when the computer device determines that there are no semantic map objects with the same semantic attributes in the reference map and the second map, it further determines that the first semantic map object cannot be associated and matched, and / or the second semantic map object cannot be associated and matched, and then obtains a matching result of failed association matching of the first semantic map object or the second semantic map object.

[0101] As described above, the embodiment of the present application determines based on whether the semantic attributes of the semantic map objects in the reference map and the second map are the same. When the semantic attributes are the same, the association matching of the first semantic map object and the second semantic map object is determined, improving the matching accuracy; when the semantic attributes are different, it indicates that the association matching of the first semantic map object and the second semantic map object fails, and there is no need to perform association matching judgment based on other parameters in the object information. In this way, the processing efficiency is improved.

[0102] In some embodiments, the process of determining the first confidence level and the second confidence level based on the matching result in step 205 is combined Figure 4 for description:

[0103] Step 401: According to the matching result, determine the first quantity of the first semantic map object in the reference map, the second quantity of the second semantic map object in the second map, and the third quantity of the semantic map objects that are successfully associated and matched among the first semantic map object and the second semantic map object;

[0104] Step 402: Determine a first confidence level and a second confidence level based on a first quantity, a second quantity, and a third quantity.

[0105] Here, determining the first confidence level and the second confidence level based on the first quantity, the second quantity, and the third quantity includes: determining the ratio of the third quantity to the first quantity as the first confidence level; determining the ratio of the third quantity to the second quantity as the second confidence level.

[0106] Here, the first confidence level can be represented by formula (1), and the second confidence level can be represented by formula (2).

[0107]

[0108]

[0109] Among them, P1 represents the first confidence level corresponding to a reference map having a conversion relationship with the first map, P2 represents the second confidence level corresponding to the second map, N represents the third quantity of semantic map objects that are successfully associated and matched between the first semantic map object and the second semantic map object, N1 represents the first quantity of the first semantic map object in the reference map, and N2 represents the second quantity of the second semantic map object in the second map.

[0110] In the embodiment of the present application, after the semantic map objects in the reference map and the semantic map objects in the second map are successfully associated and matched, obtain the first quantity of the first semantic map object in the reference map, the second quantity of the second semantic map object in the second map, and the third quantity of the semantic map objects that are successfully associated and matched between the first semantic map object and the second semantic map object. Further, determine the ratio of the third quantity to the first quantity as the first confidence level, determine the ratio of the third quantity to the second quantity as the second confidence level, and finally, based on the first confidence level and the second confidence level, determine a processing strategy for whether to collect target data for constructing a global crowdsourcing map. In this way, the embodiment of the present application can, based on the quantity of the successfully associated and matched semantic map objects, the quantity of the first semantic map object in the reference map, and the quantity of the second semantic map object in the second map, obtain the first confidence level and the second confidence level by calculating the proportion of the successfully associated and matched semantic map objects in the reference image and the second image, so as to more accurately determine whether to collect the target data.

[0111] In an implementable scenario, taking the first map as a local vectorized semantic map and the second map as a historical crowdsourced map as an example, first, after determining the mapping relationship between the ego vehicle coordinate system and the world coordinate system, based on the mapping relationship, the local vectorized semantic map is converted into a reference map in the world coordinate system. Secondly, obtain the first object information of the first semantic map object in the reference map and the second object information of the second semantic map object in the historical crowdsourced map, and based on the first object information and the second object information, determine the determination result of whether there are first and second semantic map objects with the same semantic attributes; if the determination result indicates that there are first and second semantic map objects with the same semantic attributes, and the spatial position distance between the first spatial position and the second spatial position is less than the first set threshold, and the geometric attribute difference between the first geometric attribute and the second geometric attribute is less than the second set threshold, then it is determined that the association matching between the first semantic map object and the second semantic map object is successful. Further, determine the first quantity of the first semantic map object in the reference map, the second quantity of the second semantic map object in the historical crowdsourced map, and the third quantity of the semantic map objects that are successfully associated and matched between the first semantic map object and the second semantic map object; determine the ratio of the third quantity to the first quantity as the first confidence level corresponding to the local vectorized semantic map, and determine the ratio of the third quantity to the second quantity as the second confidence level corresponding to the historical crowdsourced map. Finally, if the computer device determines that both the first confidence level and the second confidence level reach the set threshold, then determine the processing strategy of not collecting the target data; if it is determined that at least one of the first confidence level and the second confidence level does not reach the set threshold, then determine the processing strategy of collecting the target data, encrypt the target data and store it in the vehicle terminal. When it is necessary to build a crowdsourced map at the vehicle terminal, the target data is stored in the vehicle terminal; when it is necessary to build a crowdsourced map in the cloud, the target data is sent to the cloud. In this way, the embodiments of the present application can accurately, economically and efficiently collect the target data for building the crowdsourced map, reduce the cost of building the crowdsourced map, and improve the efficiency of building the crowdsourced map.

[0112] To better understand the present application, the technical solutions of the above embodiments of the present application will be further described in detail below in combination with a specific application example.

[0113] A crowdsourced map is a high-precision map produced by using low-cost sensor hardware installed on different vehicles to collect a large amount of road information collected by different vehicles at different times for semantic mapping and data fusion, and then performing data aggregation. Since the data fusion and aggregation process needs to process a large amount of data, this process is usually executed in a cloud server, and it is necessary to transmit the large amount of collected data to the cloud. Limited by the transmission channel bandwidth and traffic cost, we expect to only transmit the necessary data to reduce the cost of building the crowdsourced map.

[0114] A method for collecting autonomous driving data is disclosed in a Chinese patent with the patent application number CN202111181088.8. This method collects corresponding data in stages by setting different data collection trigger mechanisms, screens the collected data, stores the data locally after screening, and then uploads it to the cloud platform as needed. This method triggers map data collection based on scenarios during driving, sensor limitations, driving behaviors, etc., which can effectively reduce the amount of collected data, but the data collection efficiency for crowdsourced map construction using this method is relatively low.

[0115] A method for collecting and processing autonomous driving positioning data is disclosed in a Chinese patent with the patent application number CN202210700731.1. This method first sequences the real-time positioning results to obtain multiple positioning frame data, then performs anomaly checks on the data, collects the real-time positioning results and corresponding original positioning data within the time windows before and after the abnormal data, and finally uses them to assist autonomous driving. This method can only be used to collect positioning data and cannot meet the requirements of crowdsourced map construction.

[0116] Based on the above problems, this application provides a data processing method for crowdsourced map construction. Referring to Figure 5 as shown, it can be implemented through the following steps:

[0117] Step 501: Obtain sensor data collected in real time during driving.

[0118] Here, the sensor data includes positioning data, perception data, and sensor parameters. The positioning data includes coordinate information, attitude information, positioning status, and wheel speed information. The perception data includes semantic segmentation maps and target detection results. The sensor parameters include combined inertial navigation installation parameters and camera internal and external parameters.

[0119] Step 502: Construct a local vectorized semantic map based on the sensor data.

[0120] In the embodiments of this application, the SLAM mapping technology or BEV mapping technology is used to construct the local vectorized semantic map.

[0121] Step 503: Determine whether there is a historical crowdsourced map of the area where the local vectorized semantic map is located.

[0122] In the embodiments of this application, the historical crowdsourced map can be understood as the existing crowdsourced high-precision map within a certain range near the current position of the vehicle. The crowdsourced high-precision map can be stored in the cloud, or it can also be stored on the vehicle side.

[0123] Here, if the crowdsourced high-precision map stored in the vehicle does not include the historical crowdsourced map of the area where the local vectorized semantic map is located, step 506 is executed; if the crowdsourced high-precision map stored in the vehicle includes the historical crowdsourced map of the area where the local vectorized semantic map is located, step 504 is executed.

[0124] Step 504: Match and align the historical crowdsourced map with the local vectorized semantic map.

[0125] Here, match and align can be understood as unifying the two map data of the historical crowdsourced map and the local vectorized semantic map into the same coordinate system, and then associating and matching the semantic map objects in the historical crowdsourced map and the local vectorized semantic map.

[0126] Here, unify the historical crowdsourced map and the local vectorized semantic map into the same coordinate system, such as the world coordinate system; further, obtain the object information of the semantic map objects in the local vectorized semantic map and the historical crowdsourced map, where the object information includes the spatial position, geometric attributes, and semantic attributes of the semantic map objects; based on the object information, associate and match the semantic map objects in the historical crowdsourced map and the local vectorized semantic map; exemplarily, when the semantic attributes of two semantic map objects in the historical crowdsourced map and the local vectorized semantic map are the same, the difference in geometric area or geometric length is less than the first set threshold, and the spatial position distance is less than the second set threshold, it is considered that the association and matching are successful.

[0127] Step 505: Calculate the precision-recall rate and determine whether to collect data according to the precision-recall rate.

[0128] Here, the precision-recall rate (matching confidence) includes the precision rate (first confidence) and the recall rate (second confidence).

[0129] Here, calculating the precision-recall rate can be understood as calculating the precision-recall rate in the constructed local vectorized semantic map with the obtained historical crowdsourced map as a reference, so as to determine whether to collect data according to the precision-recall rate.

[0130] Here, calculate the precision rate P1 and the recall rate P2, and the calculation formulas (3) and (4) are as follows:

[0131]

[0132]

[0133] Where N is the number of successfully matched and associated semantic map objects, N3 is the number of semantic map objects in the converted local vectorized semantic map, and N4 is the number of semantic map objects in the historical crowdsourced map data of the corresponding road section.

[0134] Here, it is determined whether to collect data according to the precision-recall rate, including: when both the precision rate P1 and the recall rate P2 reach the set indicators, data is not collected; when either the precision rate P1 or the recall rate P2 does not reach the set indicators, data is collected. In this way, determining whether to collect data through the precision-recall rate further realizes the accuracy of map element attributes and spatial positions.

[0135] Step 506: Collect data, encrypt it, and store the encrypted data.

[0136] Here, the collected data includes local vectorized semantic map data and positioning data. After collecting the data, the collected data is encrypted.

[0137] Here, when it is necessary to locally construct a crowdsourced map, the data is stored locally; when it is necessary to construct a crowdsourced map in the cloud, the data is sent to the cloud. In this way, through the above method, the embodiments of the present application can accurately, economically, and efficiently collect target data for constructing a crowdsourced map, reduce the cost of constructing a crowdsourced map, and improve the efficiency of constructing a crowdsourced map.

[0138] The embodiments of the present application provide a data processing device for constructing a crowdsourced map. Refer to Figure 6 as shown in Figure 6 which is a schematic structural diagram of a data processing device for constructing a crowdsourced map provided by the embodiments of the present application. The data processing device 6 for constructing a crowdsourced map includes, among which,

[0139] An acquisition module 601, configured to acquire a local vectorized semantic map when the vehicle is driving on a target road section, and a historical crowdsourced map corresponding to the target road section. Among them, the local vectorized semantic map is constructed based on sensor data collected by the vehicle's sensors, and the sensor data includes the vehicle's positioning data, perception data, and sensor parameters;

[0140] A processing module 602, configured to obtain a matching confidence level based on the matching and alignment processing of the historical crowdsourced map and the local vectorized semantic map; among them, the matching confidence level includes a first confidence level corresponding to the local vectorized semantic map and a second confidence level corresponding to the historical crowdsourced map;

[0141] A determination module 603, configured to determine a processing strategy for whether to collect target data for constructing a global crowdsourced map based on the first confidence level and the second confidence level, where the target data includes local vector semantic maps and positioning data.

[0142] The embodiments of the present application provide a computer device. Refer to Figure 7 as shown in Figure 7A schematic structural diagram of a computer device provided by an embodiment of the present application. The computer device 7 includes a memory 701 and a processor 702. The memory 701 stores a computer program that can run on the processor 702. When the processor 702 executes the computer program, the following steps are implemented:

[0143] Obtain a local vectorized semantic map when the vehicle is driving on a target road section, and a historical crowdsourcing map corresponding to the target road section. Among them, the local vectorized semantic map is constructed based on sensor data collected by the vehicle's sensors, and the sensor data includes the vehicle's positioning data, perception data, and sensor parameters;

[0144] Based on the matching and alignment processing of the historical crowdsourcing map and the local vectorized semantic map, obtain a matching confidence level; among them, the matching confidence level includes a first confidence level corresponding to the local vectorized semantic map and a second confidence level corresponding to the historical crowdsourcing map;

[0145] Based on the first confidence level and the second confidence level, determine a processing strategy for whether to collect target data for constructing a global crowdsourcing map, where the target data includes a local vector semantic map and positioning data.

[0146] An embodiment of the present application provides a storage medium that stores one or more computer programs. The one or more computer programs can be executed by one or more processors to implement some or all of the steps in the above method. The storage medium can be transient or non-transient.

[0147] An embodiment of the present application provides a computer program, including computer-readable code. When the computer-readable code runs on a computer device, the processor in the computer device executes to implement some or all of the steps in the above method.

[0148] An embodiment of the present application provides a computer program product. The computer program product includes a non-transient computer-readable storage medium storing a computer program. When the computer program is read and executed by a computer, some or all of the steps in the above method are implemented. The computer program product can be specifically implemented in a manner of hardware, software, or a combination thereof. In some embodiments, the computer program product is specifically embodied as a computer storage medium. In other embodiments, the computer program product is specifically embodied as a software product, such as a Software Development Kit (SDK), etc.

[0149] It should be noted here that the descriptions of the above embodiments tend to emphasize the differences between the embodiments, and their similarities can be referred to each other. The descriptions of the above embodiments of the device, storage medium, computer program, and computer program product are similar to the descriptions of the above method embodiments and have beneficial effects similar to those of the method embodiments. For the technical details not disclosed in the embodiments of the device, storage medium, computer program, and computer program product of the present application, please refer to the descriptions of the method embodiments of the present application for understanding.

[0150] It should be noted here that the descriptions of the above storage medium and device embodiments are similar to the descriptions of the above method embodiments and have beneficial effects similar to those of the method embodiments. For the technical details not disclosed in the embodiments of the storage medium and device of the present application, please refer to the descriptions of the method embodiments of the present application for understanding.

[0151] The above processor may be at least one of an application specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field programmable gate array (FPGA), a central processing unit (CPU), a controller, a microcontroller, and a microprocessor. It can be understood that other electronic devices for implementing the functions of the above processor may also exist, and the embodiments of the present application do not make specific limitations.

[0152] The above computer storage medium / memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a ferromagnetic random access memory (FRAM), a flash memory, a magnetic surface memory, an optical disc, or a compact disc read-only memory (CD-ROM), etc.; or it may be various terminals including one or any combination of the above memories, such as a mobile phone, a computer, a tablet device, a personal digital assistant, etc.

[0153] It should be understood that the phrase "one embodiment" or "an embodiment" mentioned throughout the specification means that a specific feature, structure, or characteristic related to the embodiment is included in at least one embodiment of the present application. Therefore, the appearances of "in one embodiment" or "in an embodiment" throughout the specification do not necessarily refer to the same embodiment. In addition, these specific features, structures, or characteristics may be combined in one or more embodiments in any suitable manner. It should be understood that in various embodiments of the present application, the magnitudes of the serial numbers of the above steps / processes do not mean the order of execution is prior or subsequent. The order of execution of each step / process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application. The serial numbers of the embodiments of the present application are only for description and do not represent the superiority or inferiority of the embodiments.

[0154] It should be noted that in this text, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements but also includes other elements not explicitly listed, or further includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of another identical element in the process, method, article or device including that element.

[0155] In several embodiments provided by the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined, or can be integrated into another system, or some features can be ignored, or not executed. In addition, the couplings, direct couplings, or communication connections between the various components shown or discussed may be through some interfaces. The indirect couplings or communication connections of devices or units can be electrical, mechanical, or other forms.

[0156] The units described above as separate components may or may not be physically separated. The components shown as units may or may not be physical units. They can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0157] In addition, each functional unit in the embodiments of the present application can be all integrated in a processing unit, or each unit can be separately used as a unit, or two or more units can be integrated in a unit. The above-mentioned integrated units can be implemented in the form of hardware, or in the form of a combination of hardware and software functional units.

[0158] Those of ordinary skill in the art can understand that all or part of the steps for implementing the above method embodiments can be completed by hardware related to program instructions. The foregoing program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps including the above method embodiments. The foregoing storage media include various media that can store program codes, such as removable storage devices, read-only memory (ROM), magnetic disks, or optical discs.

[0159] Alternatively, if the above-mentioned integrated units of the present application are implemented in the form of software function modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present application, in essence, or the part that contributes to the related technology, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions to enable a vehicle-mounted terminal (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in the embodiments of the present application. The foregoing storage media include various media that can store program codes, such as removable storage devices, ROM, magnetic disks, or optical discs.

[0160] As described above, it is only the implementation mode of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered within the protection scope of the present application.

Claims

1. A data processing method for crowdsourcing map construction, characterized in that, The method includes: Obtaining a local vectorized semantic map when the vehicle is driving on a target road section, and a corresponding historical crowdsourcing map of the target road section, where the local vectorized semantic map is constructed based on sensor data collected by sensors of the vehicle, and the sensor data includes positioning data, perception data, and sensor parameters of the vehicle; Obtaining a matching confidence based on matching and alignment processing of the historical crowdsourcing map and the local vectorized semantic map; where the matching confidence includes a first confidence corresponding to the local vectorized semantic map and a second confidence corresponding to the historical crowdsourcing map; Determining a processing strategy for whether to collect target data for constructing a global crowdsourcing map based on the first confidence and the second confidence, where the target data includes the local vector semantic map and the positioning data.

2. The method according to claim 1, characterized in that, The local vectorized semantic map is constructed based on the vehicle's own coordinate system, and the historical crowdsourcing map is constructed based on the world coordinate system. The obtaining of the matching confidence based on the matching and alignment processing of the historical crowdsourcing map and the local vectorized semantic map includes: Determining a mapping relationship between the vehicle's own coordinate system and the world coordinate system; Based on the mapping relationship, converting a first map into a reference map with the same coordinate system as the second map for drawing, where the first map is one of the historical crowdsourcing map and the local vectorized semantic map, and the second map is the other of the historical crowdsourcing map and the local vectorized semantic map; Obtaining first object information of a first semantic map object in the reference map and second object information of a second semantic map object in the second map; Performing associated matching processing on the first semantic map object in the reference map and the second semantic map object in the second map based on the first object information and the second object information to obtain a matching result; Determining the first confidence and the second confidence based on the matching result.

3. The method according to claim 2, characterized in that, The object information includes the spatial position, geometric attribute, and semantic attribute of the semantic map object. The performing of the associated matching processing on the semantic map object in the reference map and the semantic map object in the second map based on the first object information and the second object information to obtain a matching result includes: Determining a determination result of whether there are first and second semantic map objects with the same semantic attribute based on the first object information and the second object information; If the determination result indicates that there are first and second semantic map objects with the same semantic attribute, performing associated matching processing on the first and second semantic map objects with the same semantic attribute based on the first geometric attribute and first spatial position of the first semantic map object with the same semantic attribute, and the second geometric attribute and second spatial position of the second semantic map object with the same semantic attribute to obtain the matching result; If the determination result indicates that there are no first semantic map objects and second semantic map objects with the same semantic attributes, a matching result of failed associated matching of the first semantic map object or the second semantic map object is obtained.

4. The method according to claim 2, characterized in that, The performing, based on the first geometric attribute and the first spatial position of the first semantic map object with the same semantic attributes, and the second geometric attribute and the second spatial position of the second semantic map object with the same semantic attributes, an associated matching process on the first semantic map object and the second semantic map object with the same semantic attributes to obtain the matching result includes: For the first semantic map object and the second semantic map object with the same semantic attributes, if the relationship between the first spatial position and the second spatial position, and the relationship between the first geometric attribute and the second geometric attribute satisfy the associated matching condition, a matching result of successful associated matching between the first semantic map object and the second semantic map object is obtained; If the relationship between the first spatial position and the second spatial position, and the relationship between the first geometric attribute and the second geometric attribute do not satisfy the associated matching condition, a matching result of failed associated matching between the first semantic map object and the second semantic map object is obtained.

5. The method according to claim 2, characterized in that, The determining, according to the matching result, the first confidence level and the second confidence level includes: According to the matching result, determining the first quantity of the first semantic map objects in the reference map, the second quantity of the second semantic map objects in the second map, and the third quantity of the semantic map objects with successful associated matching among the first semantic map objects and the second semantic map objects; Based on the first quantity, the second quantity, and the third quantity, determining the first confidence level and the second confidence level.

6. The method according to claim 5, characterized in that, The determining, based on the first quantity, the second quantity, and the third quantity, the first confidence level and the second confidence level includes: Determining the ratio of the third quantity to the first quantity as the first confidence level; Determining the ratio of the third quantity to the second quantity as the second confidence level.

7. The method according to any one of claims 1 to 6, characterized in that, The processing strategy for determining whether to collect target data for constructing a global crowdsourcing map based on the first confidence level and the second confidence level includes: If the first confidence level meets the first confidence level condition and the second confidence level meets the second confidence level condition, determining a first processing strategy of not collecting the target data; If the first confidence level does not meet the first confidence level condition, and / or, the second confidence level does not meet the second confidence level condition, determining a second processing strategy of collecting the target data, where the second processing strategy includes: Collecting the target data for encryption processing and storing the encrypted target data.

8. A data processing device for crowdsourcing map construction, characterized in that, The device includes: An acquisition module, configured to acquire a local vectorized semantic map when the vehicle is traveling on a target road section, and a historical crowdsourcing map corresponding to the target road section, where the local vectorized semantic map is constructed based on sensor data collected by sensors of the vehicle, and the sensor data includes positioning data, perception data, and sensor parameters of the vehicle; A processing module, configured to obtain a matching confidence level based on matching and alignment processing of the historical crowdsourcing map and the local vectorized semantic map; where the matching confidence level includes a first confidence level corresponding to the local vectorized semantic map and a second confidence level corresponding to the historical crowdsourcing map; A determination module, configured to determine a processing strategy for collecting target data for constructing a global crowdsourcing map based on the first confidence level and the second confidence level, where the target data includes the local vector semantic map and the positioning data.

9. A computer device, characterized in that, The computer device includes: a memory and a processor, The memory stores a computer program that can be run on the processor, and when the processor executes the computer program, it implements the data processing method for crowdsourcing map construction according to any one of claims 1 to 7.

10. A storage medium, characterized in that, The storage medium stores one or more computer programs, and the one or more computer programs can be executed by one or more processors to implement the data processing method for crowdsourcing map construction according to any one of claims 1 to 7.

Images