Data processing method and computing device

Abstract

Embodiments of the present application provide a data processing method and a computing device. The method comprises: obtaining road picture data, the road picture data comprising one or more road pictures and collection information of the road pictures; identifying a candidate ground object in each road picture and determining a category of the candidate ground object; determining a position of the candidate ground object according to a collection position in the collection information of each road picture; counting a cumulative identification number of the candidate ground object with the same category and position, and taking the candidate ground object with a cumulative identification number satisfying a first set condition as a target object according to the category and the position; and comparing the target object with a map object with the same position in an electronic map to determine whether the target object has changed. The technical solution provided by the embodiments of the present application realizes accurate detection of map data.

Description

Technical Field

[0001] This application relates to the field of high-precision mapping technology, and in particular to a data processing method and computing device. Background Technology

[0002] With the continuous improvement of network technology, high-precision maps have emerged. Compared with ordinary map data, high-precision map data has higher accuracy and more detailed representation of the real world.

[0003] High-precision map data is typically created based on data collected using specialized equipment. Due to the frequent changes in the real world, map data needs frequent updates. However, the inventors discovered during their research into existing technologies that:

[0004] If specialized equipment is still used to collect data and update maps, the cost will be very high, and the update frequency will not meet the update requirements. One way to reduce costs and increase the update frequency is:

[0005] When discrepancies are found between the information of a real-world feature and its recorded information in high-precision map data—that is, when changes are detected in the real-world feature—specialized equipment is used to collect data more specifically. For this method, accurately detecting changes in real-world features is a problem that those skilled in the art need to optimize and solve. Summary of the Invention

[0006] This application provides a data processing method and computing device for achieving accurate detection of map data.

[0007] In a first aspect, embodiments of this application provide a data processing method, including:

[0008] Acquire road image data, which includes: one or more road images and acquisition information of the road images;

[0009] Identify candidate ground features in each road image and determine the object category of the candidate ground features;

[0010] Based on the collection location information in each road image, the object location of the candidate ground feature is determined;

[0011] Based on the object category and object location of candidate ground objects identified from different road images, the cumulative number of identifications of candidate ground objects with the same object category and object location is counted, and the candidate ground objects whose cumulative number of identifications meets the first set condition are taken as target objects;

[0012] Based on the object location of the target object, obtain a map object that matches the object location from the map objects recorded in the electronic map database;

[0013] The target object is compared with the map object to determine whether the target object has changed.

[0014] Secondly, embodiments of this application provide a computing device, including a processing component and a storage component;

[0015] The storage component stores one or more computer instructions; the one or more computer instructions are invoked and executed by the processing component to implement the data processing method as described in the first aspect above.

[0016] In this embodiment, road image data is acquired, including one or more road images and acquisition information of the acquired road images. Candidate ground features are identified in each road image, and their object categories are determined. Based on the acquisition location in the acquisition information of each road image, the object locations of the candidate ground features are determined. Based on the object categories and locations of the candidate ground features identified from different road images, the cumulative number of identifications of candidate ground features with the same object category and location is counted, and candidate ground features whose cumulative identification count meets a first preset condition are designated as target objects. Based on the object location, a map object matching the object location is obtained from map objects recorded in an electronic map database. The target object is compared with the map object to determine whether the target object has changed. This embodiment can accurately determine whether map objects have changed, achieving accurate map data detection.

[0017] These or other aspects of this application will become more apparent in the following description of the embodiments. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 A flowchart of one embodiment of a data processing method provided in this application is shown;

[0020] Figure 2 A flowchart of yet another embodiment of a data processing method provided in this application is shown;

[0021] Figure 3This illustration shows a scenario interaction diagram of an embodiment of this application in a practical application;

[0022] Figure 4 A flowchart of yet another embodiment of a data processing method provided in this application is shown;

[0023] Figure 5 This invention provides a schematic diagram of the structure of one embodiment of a data processing apparatus.

[0024] Figure 6 A schematic diagram of one embodiment of a computing device provided in this application is shown. Detailed Implementation

[0025] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0026] In some of the processes described in the specification, claims, and accompanying drawings of this application, multiple operations appearing in a specific order are included. However, it should be clearly understood that these operations may not be executed in the order they appear herein, or may be executed in parallel. The operation numbers, such as 101, 102, etc., are merely used to distinguish different operations and do not themselves represent any execution order. Furthermore, these processes may include more or fewer operations, and these operations may be executed sequentially or in parallel. It should be noted that the descriptions such as "first," "second," etc., in this document are used to distinguish different messages, devices, modules, etc., and do not represent a chronological order, nor do they limit "first" and "second" to different types.

[0027] The technical solution in this application is applicable to scenarios involving the updating of electronic maps, such as high-precision maps. Electronic maps are maps stored and viewed digitally using computer technology; they are maps displayed in a paperless manner using collected map data. High-precision maps, on the other hand, are electronic maps with higher data accuracy and more detailed attributes. Scenarios such as intelligent transportation and autonomous driving require high-precision maps to provide corresponding services.

[0028] High-precision maps are typically created based on surveying data collected by professionals using specialized equipment. This requires high precision in data collection and sophisticated hardware, often involving data collection vehicles equipped with lasers, radar, cameras, and GPS. Due to the dynamic nature of geographic space changes, such as shifts in road traffic rules, map data updates are inevitable. If specialized equipment is still used for data collection and map updates are then performed, the cost would be prohibitively high. The inventors conceived of a solution: could data be collected only when changes are observed in the geographic space, and the map updated based on that data? This would significantly reduce costs.

[0029] Accordingly, the inventors, after a series of studies, proposed the technical solution of this application. In the embodiments of this application, road image data is first acquired, including one or more road images and their acquisition information. Candidate ground features are identified in each road image, and their object categories are determined. Based on the acquisition location in the acquisition information of each road image, the object location of the candidate ground features is determined. Based on the object categories and locations of the candidate ground features identified from different road images, the cumulative number of identifications of candidate ground features with the same object category and location is counted, and candidate ground features whose cumulative identification count meets a first predetermined condition are designated as target objects. Based on the object location, a map object matching the object location is obtained from map objects recorded in the electronic map database. The target object is compared with the map object to determine whether the map object has changed. This application mainly addresses changes in road data, using road images to identify candidate ground features. When the cumulative identification count of the same candidate ground feature meets a first predetermined condition, it is determined that the candidate ground feature exists on the road. Then, it is used as a target object and compared with ground features whose object locations match in the electronic map to determine whether the target object has changed. By setting constraints under the first condition, we can ensure accurate identification of objects and then determine whether the objects have changed. This enables fast and accurate detection of map data in geospatial space. Based on the changes in the map data, we can trigger a data collection task to collect more accurate data. Based on this, we can update the map, which can ensure timely map updates while reducing map update costs.

[0030] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0031] Figure 1 A flowchart of one embodiment of a data processing method provided in this application is shown. The method may include the following steps:

[0032] 101: Obtain road image data.

[0033] The road image data includes one or more road images and the information collected from those images.

[0034] Among them, road image data can be obtained by collecting data at a precision lower than the target acquisition precision.

[0035] The road image data is collected using cameras and other acquisition devices, while the location information can be obtained using GPS positioning.

[0036] Target acquisition accuracy can refer to the acquisition accuracy corresponding to the professional equipment used to acquire data, which may include location acquisition accuracy and / or image resolution accuracy, etc. Location acquisition accuracy may refer to GPS accuracy.

[0037] This application can initially acquire road image data using a lower acquisition precision than the target precision. The target precision is typically at the decimeter or centimeter level, while a lower precision, such as at the meter level, is sufficient. For example, the target precision requires GPS accuracy of approximately 20-50 centimeters, while a lower precision, such as GPS accuracy, can be around 50 meters, or even less than or equal to 50 meters.

[0038] The target acquisition precision usually requires a high image resolution, while the image resolution for acquisition precision lower than the target acquisition precision can be set according to the actual situation, such as lower than or equal to the preset resolution.

[0039] In practical applications, a large number of road images can be acquired at a precision lower than the target acquisition precision. It should be noted that the acquisition precision can vary for different road images.

[0040] Each road image can be acquired by a low-precision camera and GPS mounted on a specialized data collection vehicle while it is in motion.

[0041] Of course, data can also be collected using non-professional equipment. For example, non-professionals can use the low-precision cameras and GPS installed in their vehicles to collect data. Specifically, the electronic map client can utilize the vehicle's camera and GPS, or the vehicle's camera and the GPS in the user's device equipped with the client, to collect data. The user device can be a portable mobile terminal such as a smartphone. When driving, the user can carry the user device, launch the electronic map client, and establish a connection with the camera, allowing the client to collect data. The client can also interact with the user by providing relevant prompts and obtaining their consent before data collection. In practical applications, road images and information collected by non-professionals using non-professional equipment can be considered crowdsourced data. Non-users can use the electronic map client to receive corresponding data collection tasks and trigger data collection operations. Using crowdsourced data collection can significantly reduce data collection costs.

[0042] 102: Identify candidate ground features in each road image and determine the object category of the candidate ground features.

[0043] In this embodiment of the application, the object can specifically refer to road elements, such as traffic signs or traffic markings, which carry corresponding road attribute data, such as traffic rules indicated by traffic signs or traffic markings. Taking speed limit signs as an example, the road speed limit information they indicate, such as minimum speed limit information, maximum speed limit information, and section speed limit information, etc. When the road attribute data is updated, the map data of the electronic map needs to be updated in a timely manner.

[0044] This application performs object recognition on each road image and uses the identified objects as candidate ground features for further verification.

[0045] The specific recognition method can be implemented using image recognition algorithms, such as object detection algorithms. This application does not specify any particular method and will not elaborate further.

[0046] 103: Determine the location of candidate ground features based on the location information in each road image.

[0047] 104: Based on the object category and object location of candidate ground objects identified from different road images, count the cumulative number of identifications of candidate ground objects with the same object category and object location, and take the candidate ground objects whose cumulative number of identifications meets the first set condition as the target objects.

[0048] By recognizing a large number of road images, the number of times the same candidate ground feature is recognized can be counted. Whether they are the same candidate ground feature can be determined based on the object category and the object location.

[0049] The location of an object can be determined by combining the collected information, which includes the location of the road image, and this location can be used as the object's location. Alternatively, it can be determined by combining the image location of the candidate ground object in the road image, the collection location, and the collection angle. This application does not limit the specific implementation method of location determination.

[0050] Object categories can be used to distinguish different objects, such as different traffic signs or different traffic markings. Furthermore, object categories can also be divided based on object content. Taking speed limit signs as an example, speed limit signs are typically divided into three categories: minimum speed limit, maximum speed limit, and section speed limit. Optionally, object categories can be obtained during the image recognition process in step 102.

[0051] Based on object category and location, it can be determined whether the candidate feature actually exists on the road, thus avoiding false detections. If a candidate feature is identified, it can be treated as the target object for further processing to facilitate descriptive differentiation.

[0052] The first setting condition could be, for example, that the cumulative number of recognitions is greater than or equal to the target number.

[0053] 105: Based on the object location of the target object, retrieve the map object that matches the object location from the map objects recorded in the electronic map database.

[0054] 106: Compare the target object with the map object to determine if the object's content has changed.

[0055] Once the target object is identified, it can be compared with other map objects in the electronic map whose locations match the target object's location to determine if the target object has changed. However, due to limitations in the accuracy of road image data acquisition, matching object locations may mean that the distance difference between the object locations is within a certain range. For example, if the acquisition accuracy is 50 meters, this distance range could mean less than or equal to 50 meters.

[0056] Furthermore, as yet another embodiment, the method may also include:

[0057] If the target object changes, a data collection task can be generated; otherwise, the process can return to step 101 and continue execution.

[0058] This data collection task can be used to instruct the collection of relevant data about the target object based on the target collection precision, such as collecting road images of the road where the target object is located and more accurate data such as the object's location, in order to update electronic maps, etc. More professional equipment, such as professional data collection vehicles, can be used for data collection.

[0059] Of course, data collection tasks can also be generated based on the electronic map update request type. For example, if the update type is a regular electronic map, a data collection task for a regular electronic map can be generated; if the update type is a high-precision electronic map, a data collection task for the high-precision map can be generated.

[0060] Optionally, based on the comparison results, if the target object changes, a target road matching the target object can be determined; and a data collection task can be generated for the target road.

[0061] Optionally, there may not be a map object in the electronic map whose location matches the location of the target object. In this case, a data collection task can be generated directly.

[0062] Specifically, data collection tasks can be generated based on road signs of the target road. Alternatively, the collection area can be determined based on the location of the target object, and the data collection task can be generated based on road signs and this collection area. A data collection task can specifically instruct the collection of the target object within a specified collection area on the target road, obtaining the object's content, and updating the electronic map based on this content.

[0063] In this embodiment, candidate ground features are identified from each road image. If the cumulative number of identifications of the same candidate ground feature meets a first predetermined condition, the presence of that candidate ground feature in the road is confirmed. This candidate ground feature is then compared with a map object whose location matches the target object in the electronic map to determine if the target object has changed. The constraint of the first predetermined condition ensures accurate object identification, and then the determination of whether the object has changed, achieving fast and accurate map data detection. Furthermore, road images can be acquired using low-precision acquisition, reducing acquisition costs.

[0064] To facilitate object statistics, in some embodiments, based on the object category and location of candidate ground features identified from different road images, the cumulative number of identifications of candidate ground features with the same object category and location is counted, and candidate ground features whose cumulative identification count meets a first preset condition are selected as target objects. This may include:

[0065] For any candidate ground feature object obtained from identification, search the saved data for similar objects whose object location and object category both match the candidate ground feature object;

[0066] If not, add record information for the candidate land cover object to the saved data to record the number of times the candidate land cover object has been identified;

[0067] If so, accumulate the number of times similar objects are identified;

[0068] Objects whose cumulative recognition count in the saved data meets the first set condition are selected as target objects.

[0069] For each identified candidate ground feature, the process is performed as described above. The recorded information may include the object's category, location, and cumulative identification count. If a similar object matching the target object exists in the saved data, the identification count for that similar object can be increased; for example, the identification count can be incremented from 1 for each identification. Finally, objects whose cumulative identification count meets the first predefined condition can be counted from the saved data. Objects whose cumulative identification count does not meet the first predefined condition can be left unprocessed and observation can continue.

[0070] Furthermore, when the object is a road element, it also possesses directional information; therefore, in some embodiments, the method may further include:

[0071] Based on the acquisition direction in the acquisition information of each road image, determine the object orientation of the candidate ground features;

[0072] The search for similar objects in the saved data that match both the object location and object category of the candidate land cover object can include:

[0073] Find multiple candidate objects in the saved data whose distance difference between the object location and the candidate ground object location is within a predetermined range, and whose object orientation matches the object orientation of the candidate ground object.

[0074] Filter from multiple candidate objects to find similar objects that belong to the same object category as the candidate ground object.

[0075] Among them, the object orientation can be determined based on the collection direction in the collected information, and the object orientation can specifically refer to the collection direction of its corresponding road image.

[0076] The predetermined range can be determined based on the accuracy of the road images and location information collected; for example, it can be less than or equal to 50 meters.

[0077] Matching the object's orientation can mean, for example, that the object's orientation deviates from the orientation by an angle of less than or equal to 30 degrees.

[0078] Since it may not be possible to identify candidate ground features from any road image, there is a possibility that objects in the road may be removed. In this case, if no candidate ground features are identified from any road image, we can search the electronic map for map objects in the roads that match the collection location of the road images. If so, we can set the first marker information for the road.

[0079] The system collects the first marker information of roads. For roads where the cumulative number of times the first marker information is set exceeds a first predetermined number, a data collection task is generated.

[0080] The first marker information is used to indicate that the map object in the road may no longer exist. The first predetermined number of constraints is used to further verify whether the map object in the road no longer exists. If it is determined that the map object in the road no longer exists, a new data collection task is generated to update the electronic map based on the road data collected again by the data collection task.

[0081] Furthermore, since there may not be a map object matching the location of the target object in the electronic map, meaning an object may have been added to the road, as an alternative, if no map object exists at the location matching the target object in the electronic map, a data collection task can be generated directly.

[0082] Of course, in order to further improve detection accuracy and reduce data collection costs, if there is no map object at the location that matches the target object in the electronic map, a road that matches the location information corresponding to the road image can be found, and a second marker information can be set for that road; for roads where the cumulative number of times the second marker information is set is greater than a second predetermined number, a data collection task will be generated again.

[0083] The constraint of a second predetermined number of times can be used to further verify whether new objects have been added to the road.

[0084] For ease of understanding, see [link to relevant documentation]. Figure 2 The flowchart shown below is a flowchart of yet another embodiment of a data processing method provided in this application. The method may include the following steps:

[0085] 201: Obtain road image data.

[0086] The road image data includes one or more road images and corresponding acquisition information. Acquisition information may include the acquisition location and direction, among other things.

[0087] 202: Identify whether there are candidate ground features in each road image. If yes, proceed to step 203; otherwise, proceed to step 208.

[0088] 203: Determine the object location of the candidate land cover based on the collection location information in each road image.

[0089] 204: Based on the object category and object location of candidate ground objects identified from different road images, count the cumulative number of identifications of candidate ground objects with the same object category and object location, and take the candidate ground objects whose cumulative number of identifications meets the first set condition as the target objects.

[0090] 205: Based on the object's location, search the electronic map database records to see if a map object exists at a location that matches the object's location. If yes, proceed to step 206; otherwise, proceed to step 207.

[0091] 206: Compare the target object with the map object to determine if the target object has changed. If yes, proceed to step 207; otherwise, return to step 201 to continue execution.

[0092] 207: Generate a data collection task based on the target object.

[0093] 208: Check if there is a map object in the road that matches the collection location in the electronic map database record. If yes, proceed to step 209; otherwise, return to step 201 to continue.

[0094] 209: Set the first marking information for the road.

[0095] 210: Calculate the first marking information of the road, and generate a data collection task for roads where the cumulative number of times the first marking information is set is greater than the first predetermined number of times.

[0096] As described above, generating a data collection task may include: determining the target road that matches the target object; and generating a data collection task for that target road.

[0097] Because road images and collected information may not have been professionally acquired, and due to the complexity of roads, in order to accurately determine the target road matching the target object, in some embodiments, determining the target road matching the target object may include:

[0098] The target road matching the target object is determined based at least on the matching results of the road location and the object location, and the matching results of the road orientation and the object orientation.

[0099] In addition, the target road can be determined by combining factors such as road elevation, road grade, and road relationships.

[0100] Road grades can be determined based on road type. For example, a common classification method currently divides roads into five grades, from highest to lowest:

[0101] Highways: These are highways of particular political and economic importance. They have four or more lanes, a central median, grade-separated intersections, and comprehensive traffic safety, management, and service facilities. Access is fully controlled, and they are dedicated highways for high-speed motor vehicle travel. They can accommodate an average daily traffic volume (ADT) of over 25,000 vehicles per year.

[0102] Class I highways are highways that connect important political, economic and cultural centers and have some grade-separated intersections. They can generally accommodate 10,000 to 25,000 vehicles.

[0103] Secondary highways are trunk roads connecting political and economic centers or large industrial and mining areas, or busy suburban roads, capable of accommodating ADTs of 2,000 to 10,000 vehicles.

[0104] Class III highways are branch roads connecting counties or cities above the county level, capable of accommodating ADT=200 to 2000 vehicles.

[0105] Class IV highways are branch roads that connect counties, towns, or townships and can accommodate less than 200 vehicles per hour.

[0106] Road relationships can indicate the complexity of a road system. One way to classify a road is to consider it complex if multiple roads exist within a certain distance from it (e.g., more than a certain number of roads); otherwise, it can be considered simple. Alternatively, it can be classified as complex based on the presence of interchanges, intersections, more than a certain number of roads, and / or auxiliary roads within a certain distance. If these exist, it can be identified as complex; otherwise, it is considered simple.

[0107] Road elevation can refer to the distance from the road along the vertical line to the horizontal plane, or it can refer to the road's altitude.

[0108] Based on practical applications, target roads that match the target object can be selected according to multiple factors such as road location, road orientation, road elevation, road grade, and road relationships.

[0109] Alternatively, for example, a road whose location matches the object's location, whose orientation matches the object's orientation, and whose elevation matches the object's elevation can be used as the target road, and the target object can be assumed to have been photographed on that target road.

[0110] Matching road location with object location can mean that the road is within a certain distance of the object, such as within 50 meters. Matching road orientation with object orientation can mean that the road orientation deviates from the object orientation by less than 160 degrees. Matching road elevation with object elevation can mean that the difference between the road elevation and the object elevation is within a certain range, such as 5 meters.

[0111] If there are multiple roads that match the location of the road with the location of the object, the orientation of the road with the orientation of the object, and the elevation of the road with the elevation of the object, you can choose the road closest to the target object as the target road.

[0112] Of course, further filtering can be performed by combining other factors such as road grade and road relationships. For example, speed limit information will also affect road filtering for speed limit signs. Therefore, for objects that are speed limit signs, the target roads that match the target object can include:

[0113] 1) Filter candidate roads whose location matches the object's location, road orientation matches the object's orientation, and road elevation matches the object's elevation;

[0114] 2) If there is a candidate road, determine that candidate road as the target road.

[0115] 3) If there are two or more candidate roads, compare the speed limit information of the target speed limit sign with the speed limit value, and determine the target road from the two or more candidate roads based on the speed limit value comparison result.

[0116] As an optional approach, determining the target road from two or more candidate roads based on the speed limit comparison results may include:

[0117] If the speed limit information of the target speed limit sign is greater than the first speed limit value, the highway that is closest to the target object among the two or more candidate roads shall be taken as the target road.

[0118] Based on the current speed limits in China, the first speed limit could be, for example, 80 km / h. Since the road images were taken on highways, the highway closest to the target can be selected from two or more candidate roads as the target road.

[0119] As another optional method, determining the target road from the two or more candidate roads based on the speed limit comparison results may include: if the speed limit information of the target speed limit sign is less than the second speed limit value, selecting the expressway ramps and non-expressways from the two or more candidate roads; if the selection result is one road, that road is taken as the target road; if the selection result is multiple roads and there are complex roads, the candidate road that is closest to the target object and has a lower level is marked as a pending road; if the selection result is multiple roads and there are no complex roads, the candidate road that is closest to the target object is taken as the target road.

[0120] The second speed limit could be, for example, 40 km / h.

[0121] In cases of complex road conditions, speed limit signs on adjacent roads may be captured, allowing for further filtering based on road classification. If the speed limit information is less than the second speed limit value, the road image can be assumed to have been captured on a highway ramp or a non-highway. Therefore, from two or more candidate roads, selecting highway ramps and non-highways, if the result is a single road, that road can be used as the target road. If the result is multiple roads, and all are simple roads, the candidate road closest to the target can be selected as the target road; otherwise, the candidate road closest to the target and of lower classification can be marked as a pending road. Since lower road classifications correspond to smaller speed limits, the candidate road closest to the target and of lower classification can be marked as a pending road.

[0122] As another alternative, determining the target road from two or more candidate roads based on the speed limit comparison results may include:

[0123] If the speed limit information of the target speed limit sign is between the first speed limit value and the second speed limit value, select the high-speed ramp and non-highway from the two or more candidate roads. If the result is a single road, take that road as the target road. If the result is multiple roads and there are complex roads, mark the candidate road closest to the target as a pending road. If the result is multiple roads and there are no complex roads, take the candidate road closest to the target as the target road.

[0124] When the speed limit information is between the first and second speed limit values, the target speed limit sign may be of the section speed limit type. The speed limit information is specifically the speed limit section. You can select the candidate road that is closest to the target object and mark it as the road to be determined, without having to consider the road level.

[0125] In addition, in some embodiments, the method may further include: outputting a first prompt message for a pending road whose cumulative marking count is greater than a third predetermined count.

[0126] As described above, some roads will be marked as pending roads. For pending roads whose cumulative marking count exceeds the third predetermined count, the first prompt message will be output.

[0127] The first prompt message can be used to prompt manual verification of whether the candidate road needs to be collected. Manual verification can be performed by viewing road images or conducting on-site surveys to determine whether the candidate road needs to be collected, and then decide whether to generate a data collection task.

[0128] To further ensure the accuracy of map data detection and the accuracy of data collection task generation, in some embodiments, generating a data collection task for the target road may include:

[0129] Determine the number of target objects that match each target road; for target roads where the number of matched target objects meets the second set condition, generate a data collection task.

[0130] The second condition could be, for example, that the number of matched target objects exceeds a predetermined number. That is, for a target road, if multiple target objects match it, a data collection task is generated for that target road. This task specifically instructs data collection to be performed on the target road according to the target collection precision. The object categories of the multiple target objects matching the target road can be the same or different. By accurately detecting whether the objects and roads have changed, and determining when map data has changed before generating a data collection task, the accuracy of the data collection task can be guaranteed, thus ensuring the effectiveness of high-precision data collection and reducing map update costs.

[0131] As described above, in a practical application, the technical solution of this application embodiment can be applied to scenarios involving the rapid and accurate detection of changes in the current status of speed limit signs on roads, where the object specifically refers to the speed limit sign. For ease of understanding, the technical solution of this application will be described below using a speed limit sign as an example, combined with... Figure 3 The illustrated scenario shows an interactive diagram. In this embodiment, the technical solution can be executed by the electronic map server 301. The electronic map client 302 can collect road image data and upload it to the server 301. The client 302 can be configured in a user device, allowing the user to carry the device and collect data while driving (motorized vehicle or non-motorized vehicle, etc.) or walking. The server 301 can collect a large amount of road image data collected by the client 302. For ease of understanding, Figure 3This section uses a user device as an example. The client is configured on the user device, and the user device can be used in scenarios such as driving a car, riding a bicycle, or walking. Figure 3 The examples provided are merely illustrations of a few possible scenarios, and this application is not limited to these.

[0132] Then, the server 301 can identify each road image in the road image data, identify candidate speed limit signs 31 in the road images, and identify the same candidate speed limit sign according to the speed limit sign category and speed limit sign location, and take the candidate speed limit sign whose cumulative identification number meets the first set condition as the target speed limit sign 32; then, compare the target speed limit sign with the original speed limit sign that matches the object location in the electronic map 33; based on the comparison result, determine whether to generate a data collection task 34.

[0133] The data collection task generated by the server 301 can be assigned to relevant professionals (through any form of communication, such as email, instant messaging, etc., this application does not specifically limit this). The professionals will use professional equipment (driving a professional data collection vehicle) to perform the data collection task and obtain professional data. The server 301 can then update the electronic map based on this professional data.

[0134] As yet another example, such as Figure 4 The data processing method flowchart shown below may include the following steps:

[0135] 401: Get road image data.

[0136] The road image data includes one or more road images and the acquisition information of the road images.

[0137] 402: Identify candidate ground features in each road image and determine the object category of the candidate ground features.

[0138] 403: Determine the object location of the candidate land cover based on the collection location information in each road image.

[0139] 404: Based on the object category and object location of candidate ground objects identified from different road images, the cumulative number of identifications of candidate ground objects with the same object category and object location is counted, and the candidate ground objects whose cumulative number of identifications meets the first set condition are taken as target objects.

[0140] 405: Based on the object location of the target object, retrieve the map object that matches the object location from the map objects recorded in the electronic map database.

[0141] 406: Compare the target object with the map object to determine whether the target object has changed.

[0142] 407: If the target object changes, determine the target road that matches the target object.

[0143] The specific method for determining the road that matches the target object is the same as or similar to the method for determining the target road that matches the target object in the previous embodiment. For example, multiple factors such as road location, road orientation, road elevation, road grade, road relationship, and speed limit information of the target speed limit sign can be combined for road screening, which will not be elaborated here.

[0144] 408: Determine the number of target objects matched for each target road, and generate a data collection task for target roads where the number of matched target objects exceeds a predetermined number.

[0145] In this embodiment, target objects identified in road images can be directly compared with map objects whose locations match in the electronic map. If the target object changes, the road matching the target object can be determined. Roads with more than a predetermined number of matching target objects can be designated as target roads, and data collection tasks are generated for these target roads. By constraining the predetermined number of times, it can be verified whether the target objects on the road have changed, thereby determining whether to generate a data collection task. This embodiment achieves accurate map data detection, ensuring that data collection tasks are only generated when map data changes, which can improve the accuracy of data collection task generation and reduce map update costs.

[0146] Figure 5 A schematic diagram of the structure of a data processing apparatus according to an embodiment of this application may include:

[0147] The first acquisition module 501 is used to acquire road image data, which includes: one or more road images and acquisition information for acquiring the road images.

[0148] The first identification module 502 is used to identify candidate ground features in each road image and determine the object category of the candidate ground features.

[0149] The first determining module 503 is used to determine the object location of the candidate ground object based on the collection location in the collection information of each road image; based on the object category and object location of the candidate ground object identified from different road images, to count the cumulative number of recognitions of candidate ground objects with the same object category and object location, and to take the candidate ground object whose cumulative recognition count meets the first set condition as the target object;

[0150] The first comparison module 504 is used to obtain a map object matching the object location from the map objects recorded in the electronic map database based on the object location of the target object; and compare the target object with the map object matching the object location in the electronic map to determine whether the object content has changed.

[0151] In some embodiments, the device may further include:

[0152] The first processing module is used to determine the target road that matches the target object if the target object changes; and to generate a data collection task for the target road.

[0153] In some embodiments, the first determining module is further configured to determine the object orientation of the candidate ground feature based on the acquisition direction in the acquisition information of each road image;

[0154] The first determining module can specifically, for any identified candidate ground feature object, search the saved data for multiple candidate objects whose distance difference from the object's location to the candidate ground feature object's location is within a predetermined range, and whose object orientation matches the candidate ground feature object's object orientation; and filter from the multiple candidate objects for similar objects belonging to the same object category as the candidate ground feature object; if no similar object exists in the saved data, add a record of the candidate ground feature object to the saved data, recording the number of times the candidate ground feature object was identified; if a similar object exists in the saved data, accumulate the number of times the similar object was identified; and count the objects in the saved data whose cumulative number of identifications meets a first set condition as target objects.

[0155] In some embodiments, the device may further include:

[0156] The second processing module is used to, if no candidate ground features are identified from any road image, search the electronic map for a road whose collection location matches that of the road image; if so, set first marker information for the road; count the first marker information of the road, and generate a data collection task for roads whose cumulative number of first marker settings is greater than a first predetermined number.

[0157] In some embodiments, the object is a speed limit sign, and the first processing module may determine the target road that matches the target object by: at least based on the matching result of the road position and the object position, and the matching result of the road orientation and the object orientation, to determine the target road that matches the target object.

[0158] In some embodiments, the first processing module determines, at least based on the matching results of road location and object location, and the matching results of road orientation and object orientation, that the target road matching the target object may be:

[0159] Filter candidate roads whose location matches the object's location, orientation matches the object's orientation, and elevation matches the object's elevation; if a candidate road exists, determine that candidate road as the target road;

[0160] If there are two candidate roads, the speed limit information at the target speed limit sign is compared with the speed limit value; based on the comparison result of the speed limit value, the target road is determined from the two or more candidate roads.

[0161] The first processing module determines the target road from the two or more candidate roads based on the speed limit comparison results. This may include: if the speed limit information of the target speed limit sign is greater than the first speed limit value, the highway that is closest to the target object among the two or more candidate roads shall be taken as the target road.

[0162] The first processing module determines the target road from two or more candidate roads based on the speed limit comparison results. This can include: if the speed limit information of the target speed limit sign is less than the second speed limit value, selecting the expressway ramps and non-expressways from the two or more candidate roads; if the selection result is one road, that road is taken as the target road; if the selection result is multiple roads and there are complex roads, the candidate road that is closest to the target object and has a lower level is marked as a pending road; if the selection result is multiple roads and there are no complex roads, the candidate road that is closest to the target object is taken as the target road.

[0163] The first processing module determines the target road from two or more candidate roads based on the speed limit comparison results. This can include: if the speed limit information of the target speed limit sign is between the first speed limit value and the second speed limit value, selecting the expressway ramps and non-expressways from the two or more candidate roads; if the selection result is one road, that road is taken as the target road; if the selection result is multiple roads and there are complex roads, the candidate road closest to the target object is marked as a pending road; if the selection result is multiple roads and there are no complex roads, the candidate road closest to the target object is taken as the target road.

[0164] In addition, the first processing module can also be used to output a first prompt message for pending roads whose cumulative marking count is greater than a third predetermined count.

[0165] In some embodiments, the first processing module may generate a data collection task for the target road by: determining the number of target objects matched for each target road; and generating a data collection task for target roads where the number of matched target objects is greater than a predetermined number.

[0166] Figure 5 The data processing device can perform Figure 1The implementation principle and technical effects of the data processing method described in the illustrated embodiments will not be repeated here. The specific methods by which each module and unit of the data processing device in the above embodiments performs its operations have been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0167] In one possible design, Figure 5 The data processing apparatus shown in the embodiment can be implemented as a computing device, such as... Figure 6 As shown, the computing device may include a storage component 601 and a processing component 602;

[0168] Storage component 601 stores one or more computer instructions, wherein the one or more computer instructions are invoked and executed by processing component 602 to achieve [the desired result]. Figure 1 The data processing method shown.

[0169] Of course, computing devices may also include other components, such as input / output interfaces and communication components.

[0170] Input / output interfaces provide interfaces between processing components and peripheral interface modules, which can be output devices, input devices, etc.

[0171] The communication components are configured to facilitate wired or wireless communication between computing devices and other devices.

[0172] The computing device can be a physical device or an elastic computing host provided by a cloud computing platform. In this case, the computing device can refer to a cloud server, and the aforementioned processing components, storage components, etc., can be basic server resources rented or purchased from the cloud computing platform.

[0173] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a computer, can perform the above-described functions. Figure 1 The data processing method of the embodiment shown.

[0174] The processing components involved in the corresponding embodiments described above may include one or more processors to execute computer instructions to complete all or part of the steps in the methods described above. Alternatively, the processing components may be implemented as one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to execute the methods described above.

[0175] Storage components are configured to store various types of data to support operations on the terminal. Storage components can be implemented from any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0176] Display component 903 may be an electroluminescent (EL) element, a liquid crystal display or a microdisplay with a similar structure, or a retina-direct display or a similar laser scanning display.

[0177] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0178] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0179] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0180] 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 of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A data processing method, wherein, include: Acquire road image data, which includes: one or more road images and acquisition information of the road images; Identify candidate ground features in each road image and determine the object category of the candidate ground features; Based on the collection location information in each road image, the object location of the candidate ground feature is determined; Based on the object category and object location of candidate ground objects identified from different road images, the cumulative number of identifications of candidate ground objects with the same object category and object location is counted, and the candidate ground objects whose cumulative number of identifications meets the first set condition are taken as target objects; Based on the object location of the target object, obtain a map object that matches the object location from the map objects recorded in the electronic map database; The target object is compared with the map object to determine whether the target object has changed; If the target object changes, determine the target road that matches the target object; A data collection task is generated for the target road.

2. The method according to claim 1, wherein, Also includes: Based on the acquisition direction in the acquisition information of each road image, determine the object orientation of the candidate ground features; The step of calculating the cumulative number of recognitions of candidate ground features with the same object category and location based on the object category and location identified from different road images, and selecting candidate ground features whose cumulative recognition count meets a first set condition as target objects includes: For any candidate ground feature object obtained by identification, search the saved data for multiple candidate objects whose distance difference between the object location and the object location of the candidate ground feature object is within a predetermined range and whose object orientation matches the object orientation of the candidate ground feature object; and filter similar objects belonging to the same object category as the candidate ground feature object from the multiple candidate objects. If the similar object does not exist in the saved data, add the record information of the candidate land cover object to the saved data and record the number of times the candidate land cover object is identified; If the similar object exists in the saved data, the number of times the similar object has been identified is accumulated. Objects whose cumulative recognition count in the saved data meets the first set condition are selected as target objects.

3. The method according to claim 1, wherein, Also includes: If no candidate ground features are identified from any road image, search the electronic map for any map objects in the roads whose acquisition location matches that of any road image. If so, set the first marking information for the road; The first marker information of the road is statistically analyzed. For roads where the cumulative number of times the first marker information is set is greater than a first predetermined number, a data collection task is generated.

4. The method according to claim 1, wherein, The object is a speed limit sign, and the method further includes: Based on the acquisition direction in the acquisition information of each road image, determine the object orientation of the candidate ground features; Determining the target road that matches the target object includes: The target road matching the target object is determined based at least on the matching results of the road location and the object location, and the matching results of the road orientation and the object orientation.

5. The method according to claim 4, wherein, Based at least on the matching results of road location and object location, and the matching results of road orientation and object orientation, the target road matching the target object is determined to include: Filter candidate roads whose location matches the object's location, road orientation matches the object's orientation, and road elevation matches the object's elevation; If a candidate road exists, the candidate road is determined as the target road; If there are two or more candidate roads, the speed limit information of the target speed limit sign is compared with the speed limit value. Based on the comparison of speed limit values, the target road is determined from the two or more candidate roads.

6. The method according to claim 5, wherein, The step of determining the target road from the two or more candidate roads based on the speed limit comparison results includes: If the speed limit information of the target speed limit sign is greater than the first speed limit value, the highway that is closest to the target object among the two or more candidate roads shall be taken as the target road.

7. The method according to claim 5, wherein, The step of determining the target road from the two or more candidate roads based on the speed limit comparison results includes: If the speed limit information of the target speed limit sign is less than the second speed limit value, select the high-speed ramp and non-highway from the two or more candidate roads. If the selected result is a single road, that road is taken as the target road. If the selected result is multiple roads and there are complex roads, mark the candidate road that is closest to the target object and has a lower level as a pending road. If the selected result is multiple roads and there are no complex roads, take the candidate road that is closest to the target object as the target road.

8. The method according to claim 5, wherein, The step of determining the target road from the two or more candidate roads based on the speed limit comparison results includes: If the speed limit information of the target speed limit sign is between the first speed limit value and the second speed limit value, select the high-speed ramp and non-highway from the two or more candidate roads. If the selected result is a single road, that road is taken as the target road. If the selected result is multiple roads and there are complex roads, mark the candidate road closest to the target as a road to be determined. If the selected result is multiple roads and there are no complex roads, take the candidate road closest to the target as the target road.

9. The method according to claim 7 or 8, wherein, Also includes: For undetermined roads whose cumulative marking count exceeds the third predetermined count, output the first prompt message.

10. The method according to claim 1, wherein, The data collection task generated for the target road includes: Determine the number of target objects that match each target road; For target roads where the number of matched target objects exceeds a predetermined number, a data collection task is generated.

11. A computing device, wherein, This includes processing components and storage components; The storage component stores one or more computer instructions; the one or more computer instructions are to be invoked and executed by the processing component to implement the data processing method as described in any one of claims 1-10.

Images