Image recommendation methods, apparatus, electronic devices, and computer-readable storage media

Abstract

This disclosure relates to an image recommendation method, apparatus, electronic device, and computer-readable storage medium. In at least one embodiment of this disclosure, a problem point corresponding to a target map element to be processed is obtained, and at least one candidate image acquisition trajectory adjacent to the problem point is determined. This realizes a trajectory comparison-based approach to replace candidate trajectories with similar single trajectory points, improving recommendation accuracy. Furthermore, based on the problem point and the positions of each trajectory point in the candidate image acquisition trajectory, a secondary screening of the candidate image trajectories is performed using trajectory features to obtain the target image acquisition trajectory as the basis for image recommendation to map operators. Thus, at least one image corresponding to a trajectory point is selected from the target image acquisition trajectory for recommendation. This realizes a multi-feature fusion matching image recommendation method combining trajectory features and image features, improving recommendation accuracy and assisting map operators in finding suitable images.

Description

Technical Field

[0001] This disclosure relates to the field of map data production technology, specifically to an image recommendation method, apparatus, electronic device, and computer-readable storage medium. Background Technology

[0002] With technological advancements and rapid development in the real world, road mileage is increasing, and data richness is rising. To accurately represent the real world, it is necessary to improve the quality of map data, including high accuracy and high timeliness. High accuracy refers to the map's ability to reproduce the real world, measured by precision; high timeliness refers to the map's ability to display changes in the real world in real time, measured by update speed. To provide users with high-quality map data, essential steps in the map data production process include: identifying corresponding map features by examining related images, and verifying and updating the map data.

[0003] However, as more and more images are collected, the number of images adjacent to map features also increases, increasing the workload of searching and verifying. Therefore, there is an urgent need to provide an image recommendation scheme to help map workers find suitable images. Summary of the Invention

[0004] At least one embodiment of this disclosure provides an image recommendation method, apparatus, electronic device, and computer-readable storage medium.

[0005] In a first aspect, embodiments of this disclosure propose an image recommendation method, the method comprising:

[0006] Obtain the problem locations corresponding to the target map features to be processed;

[0007] Identify at least one candidate image acquisition trajectory adjacent to the problem location;

[0008] Based on the problem location and the position of each trajectory point in each candidate image acquisition trajectory, the target image acquisition trajectory is selected from at least one candidate image acquisition trajectory.

[0009] Recommend images by selecting at least one trajectory point from the target image acquisition trajectory.

[0010] Secondly, embodiments of this disclosure also provide an image recommendation device, which includes:

[0011] The acquisition unit is used to acquire the problem locations corresponding to the target map features to be processed.

[0012] The determining unit is used to determine at least one candidate image acquisition trajectory adjacent to the problem point;

[0013] The selection unit is used to select the target image acquisition trajectory from at least one candidate image acquisition trajectory based on the problem location and the position of each trajectory point in each candidate image acquisition trajectory.

[0014] The recommendation unit is used to select at least one image corresponding to a trajectory point from the target image acquisition trajectory for recommendation.

[0015] Thirdly, embodiments of this disclosure also provide an electronic device, which includes a memory, a processor, and a computer program stored on the memory, wherein the processor executes the computer program to implement the steps of the image recommendation method as described in the first aspect.

[0016] Fourthly, embodiments of this disclosure also provide a computer-readable storage medium, wherein the computer-readable storage medium stores a program or instructions that cause a computer to perform the steps of the image recommendation method as described in the first aspect.

[0017] Fifthly, embodiments of this disclosure also provide a computer program product, wherein the computer program product includes a computer program stored in a computer-readable storage medium, and at least one processor of a computer reads from and executes the computer program from the computer-readable storage medium, causing the computer to perform the steps of the image recommendation method as described in the first aspect.

[0018] As can be seen, in at least one embodiment of this disclosure, by obtaining the problem point corresponding to the target map element to be processed, at least one candidate image acquisition trajectory adjacent to the problem point is determined. This realizes the candidate approach of replacing the single trajectory point proximity (i.e., determining whether the trajectory is adjacent to the problem point) with trajectory comparison (i.e., determining whether two points (trajectory point and problem point) are close) and improves the recommendation accuracy. Furthermore, based on the problem point and the position of each trajectory point in the candidate image acquisition trajectory, the candidate image trajectory is further filtered using trajectory features (including the position of trajectory points) to obtain the target image acquisition trajectory as the basis for recommending images to map operators. Thus, at least one image corresponding to a trajectory point is selected from the target image acquisition trajectory for recommendation. This realizes a multi-feature fusion matching image recommendation method that combines trajectory features and image features, improving the recommendation accuracy and assisting map operators in finding suitable images to verify and update map data. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of this disclosure, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this disclosure. For those skilled in the art, other drawings can be obtained based on these drawings.

[0020] Figure 1 A flowchart illustrating an image recommendation method provided in an embodiment of this disclosure;

[0021] Figure 2 This is a flowchart illustrating a process for obtaining the problem locations corresponding to the target map features to be processed, provided by an embodiment of this disclosure.

[0022] Figure 3 This is a schematic flowchart illustrating a process for determining at least one candidate image acquisition trajectory adjacent to a problem location, as provided in an embodiment of this disclosure.

[0023] Figure 4 This is a schematic diagram of a process for selecting a target image acquisition trajectory from at least one candidate image acquisition trajectory, provided in an embodiment of this disclosure.

[0024] Figure 5 This is a schematic diagram illustrating how to determine the relative distance between a trajectory point and a problem point, and the relative angle between the line connecting the trajectory point and the problem point and the image acquisition direction, as provided in this embodiment of the disclosure.

[0025] Figure 6 This is a schematic diagram of a process for determining whether a candidate image acquisition trajectory is a target image acquisition trajectory based on the relative distance and relative angle of a trajectory point relative to a problem point, as provided in an embodiment of this disclosure.

[0026] Figure 7 This is a schematic diagram of a process for recommending images corresponding to at least one trajectory point from a target image acquisition trajectory, provided by an embodiment of the present disclosure.

[0027] Figure 8 This is a schematic flowchart illustrating a process for determining whether a trajectory representative point exists in each target image acquisition trajectory, as provided in an embodiment of this disclosure.

[0028] Figure 9 This is a schematic diagram of a process for recommending images corresponding to at least one trajectory point from a target image acquisition trajectory, provided by an embodiment of the present disclosure.

[0029] Figure 10 This is a schematic diagram illustrating a scenario where at least one image corresponding to a trajectory point is selected from a target image acquisition trajectory for recommendation, according to an embodiment of this disclosure.

[0030] Figure 11 A flowchart illustrating an image recommendation method provided in an embodiment of this disclosure;

[0031] Figure 12 A schematic diagram of an image recommendation device provided in an embodiment of this disclosure;

[0032] Figure 13This is an exemplary block diagram of an electronic device provided in an embodiment of the present disclosure. Detailed Implementation

[0033] To better understand the above-described objectives, features, and advantages of this disclosure, the present disclosure will be further described in detail below with reference to the accompanying drawings and embodiments. It is to be understood that the described embodiments are only some, not all, of the embodiments of this disclosure. The specific embodiments described herein are merely for explaining this disclosure and are not intended to limit it. All other embodiments obtained by those skilled in the art based on the described embodiments of this disclosure are within the scope of protection of this disclosure.

[0034] It should be noted that in this article, relational terms such as “first” and “second” are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations.

[0035] In some related embodiments, one method for map operators to find images is to sort them according to their entry time into the database and then search for images based on the sorting results. This method has the problem of not meeting the requirement of timeliness, where timeliness means that the geospatial information provided by the map should reflect the most up-to-date situation as much as possible. For example, there might be a speed limit sign on a road, but it has been removed due to construction or other reasons. Historically acquired images might include the speed limit sign, while the most recently acquired images might not. This means that the searched image might still include the speed limit sign, causing the map data used to generate that speed limit sign to be inconsistent with reality and fail to reflect the latest situation, thus failing to meet the requirement of timeliness.

[0036] In other related embodiments, a second method for map operators to find images involves sorting images based on their spatial proximity to the target map feature, and then searching for images based on the sorting results. However, this second method still suffers from a lack of timeliness. For example, a speed limit sign might exist on a road, but it has been removed due to construction. Historically acquired images might include the sign, while the most recently acquired images do not. If the historically acquired images are closer to the target map feature, map operators will prioritize searching from those images, potentially resulting in images that still contain the speed limit sign. This would lead to map data for the speed limit sign that is inconsistent with reality and fails to reflect the latest situation, thus failing to meet timeliness requirements. Furthermore, sorting by spatial proximity does not screen for the inclusion of the target map feature, meaning that images ranked higher might not include the target map feature, resulting in the inability to find it.

[0037] To recommend suitable images to map workers, this disclosure provides an image recommendation method, apparatus, electronic device, and computer-readable storage medium. By acquiring the problem point corresponding to the target map element to be processed, at least one candidate image acquisition trajectory adjacent to the problem point is determined. This replaces the single trajectory point proximity (i.e., determining whether a trajectory is adjacent to the problem point) candidate approach with trajectory comparison (i.e., determining whether two points (trajectory point and problem point) are close), improving recommendation accuracy. Furthermore, based on the problem point and the positions of each trajectory point in the candidate image acquisition trajectory, a secondary screening of the candidate image trajectories is performed using trajectory features (including the positions of trajectory points), resulting in the target image acquisition trajectory used as the basis for image recommendation to map workers. Thus, at least one image corresponding to a trajectory point is selected from the target image acquisition trajectory for recommendation. This achieves a multi-feature fusion matching image recommendation method combining trajectory features and image features, improving recommendation accuracy and assisting map workers in finding suitable images to verify and update map data.

[0038] Figure 1 This is a flowchart illustrating an image data recommendation method provided in an embodiment of the present disclosure. The execution subject of the image data recommendation method is an electronic device, including but not limited to in-vehicle devices, smartphones, PDAs, tablets, wearable devices with displays, desktop computers, laptops, all-in-one computers, smart home devices, servers, etc. The server can be an independent server or a cluster of multiple servers, and can include servers built locally and servers set up in the cloud.

[0039] like Figure 1 As shown, the image data recommendation method may include, but is not limited to, steps 101 to 104:

[0040] In step 101, the problem points corresponding to the target map features to be processed are obtained.

[0041] In this embodiment, the target map element to be processed is a map element that may cause or has already caused problems. Map elements are elements used to construct high-precision maps. For example, map elements include, but are not limited to, ground elements such as directional arrows, lane lines, stop lines, and ground text, as well as ground elements such as signs and speed cameras outside the ground, and also include intersections and paths related to intersections.

[0042] In this embodiment, problem points can be understood as coordinate locations used to analyze whether there are problems with the target map elements. It is evident that problem points are related to the target map elements. Therefore, after obtaining the target map elements to be processed, the problem points corresponding to the target map elements can be further determined. Optional implementation methods for determining the corresponding problem points based on the target map elements are described below.

[0043] In step 102, at least one candidate image acquisition trajectory adjacent to the problem location is determined.

[0044] In this embodiment, in order to recommend images to map operators, images need to be selected from images collected in multiple image acquisition operations. During each image acquisition operation, the trajectory traveled by the acquisition vehicle or the manual handheld acquisition device forms one or more image acquisition trajectories. Each trajectory point in the image acquisition trajectory is the location of the acquired image, that is, each trajectory point in the image acquisition trajectory corresponds to one image.

[0045] In this embodiment, from multiple image acquisition trajectories formed by multiple image acquisition operations, at least one candidate image acquisition trajectory adjacent to the problem point is determined. "Adjacent to the problem point" can be understood as being within a preset range around the problem point, indicating a connection to the problem point and serving as a reference for map operators to find suitable images. Therefore, the at least one candidate image acquisition trajectory adjacent to the problem point can be used as a set of candidate trajectories for recommending images to map operators. Non-adjacent image acquisition trajectories are redundant trajectories that are unrelated to the problem point or have a weak correlation, and should be eliminated; otherwise, they will interfere with map operators' image search. Optional implementation methods for determining at least one candidate image acquisition trajectory adjacent to the problem point are described below.

[0046] As can be seen, this embodiment realizes the candidate approach of replacing the single trajectory point proximity (i.e., determining whether the trajectory is adjacent to the problem point) with trajectory comparison (i.e., determining whether two points (the trajectory point and the problem point) are close) to improve the recommendation accuracy.

[0047] In step 103, based on the problem location and the position of each trajectory point in each candidate image acquisition trajectory, the target image acquisition trajectory is selected from at least one candidate image acquisition trajectory.

[0048] In this embodiment, although at least one candidate image acquisition trajectory adjacent to the problem point is determined, by combining the position of the problem point and the position of each trajectory point in the candidate image acquisition trajectory, the candidate image acquisition trajectory under abnormal acquisition scenario can be further eliminated. The abnormal acquisition scenario is abnormal relative to the problem point.

[0049] An abnormal acquisition scenario is as follows: The image acquisition device (e.g., a camera) installed on the acquisition vehicle acquires images of the front and sides of the vehicle, but cannot acquire images of the rear of the vehicle. Therefore, based on the position of each trajectory point in the candidate image acquisition trajectory of the problem point, it can be determined whether the image corresponding to each trajectory point is an image of the rear of the vehicle. If the image corresponding to a certain trajectory point is an image of the rear of the vehicle, it means that the image corresponding to that trajectory point requires changing the position and orientation of the camera on the vehicle to capture the image. However, in reality, the position of the camera on the vehicle is usually fixed. Therefore, the image corresponding to that trajectory point is an abnormally captured image relative to the problem point, and belongs to the interference image. It means that the candidate image acquisition trajectory where the trajectory point is located is irrelevant to the problem point and should be eliminated.

[0050] In this embodiment, based on the problem location and the positions of each trajectory point in each candidate image acquisition trajectory, candidate image acquisition trajectories unrelated to the problem location are eliminated. The remaining candidate image acquisition trajectories are related to the problem location, and therefore, the remaining candidate image acquisition trajectories are used as the target image acquisition trajectories for image recommendation to map operators. An alternative implementation method for selecting the target image acquisition trajectory from at least one candidate image acquisition trajectory based on the problem location and the positions of each trajectory point in each candidate image acquisition trajectory is described below.

[0051] It is evident that by determining the candidate image acquisition trajectory and the target image acquisition trajectory, the image acquisition trajectory is filtered twice using trajectory features (including the location of trajectory points), eliminating redundant trajectories that may interfere with map operators. Therefore, it can help map operators find suitable images more quickly.

[0052] In step 104, at least one image corresponding to a trajectory point is selected from the target image acquisition trajectory for recommendation.

[0053] In this embodiment, considering that the images corresponding to each trajectory point in the target image acquisition trajectory are different, although they are all related to the problem location, the degree of correlation is different. For example, the image corresponding to a certain trajectory point may contain the entire target map feature corresponding to the problem location, while the image corresponding to another trajectory point may only contain a part of the target map feature. Therefore, the image corresponding to the former is selected for priority recommendation, and then the image corresponding to the latter is recommended. An optional implementation method for selecting at least one image corresponding to a trajectory point from the target image acquisition trajectory for recommendation is described below.

[0054] As can be seen, in the above embodiments, by obtaining the problem point corresponding to the target map element to be processed, at least one candidate image acquisition trajectory adjacent to the problem point is determined. This realizes the candidate approach of replacing the single trajectory point proximity (i.e., determining whether the trajectory is adjacent to the problem point) with trajectory comparison (i.e., determining whether two points (trajectory point and problem point) are close) and improves the recommendation accuracy. Furthermore, based on the problem point and the position of each trajectory point in the candidate image acquisition trajectory, the candidate image trajectory is further filtered using trajectory features (including the position of trajectory points) to obtain the target image acquisition trajectory as the basis for recommending images to map operators. Thus, at least one image corresponding to a trajectory point is selected from the target image acquisition trajectory for recommendation. This realizes a multi-feature fusion matching image recommendation method that combines trajectory features and image features, improving the recommendation accuracy and assisting map operators in finding suitable images to verify and update map data.

[0055] Based on the above embodiments, Figure 1 Step 101 shown, "obtaining the problem points corresponding to the target map features to be processed", includes, for example, Figure 2 Steps 201 to 203 are shown below:

[0056] In step 201, the error location reported by the navigation terminal is received.

[0057] In this embodiment, there are several ways for the navigation terminal to report error locations. The navigation terminal can be a vehicle-side application (APP) installed on the vehicle, containing an electronic map (e.g., a high-precision map or a standard-precision map). It can also be a user-side application, i.e., an application containing an electronic map installed on the user's mobile device. For example, if a user finds a discrepancy between the real world and the map at a certain location, the user can actively report an error, and the error location reported by the navigation terminal becomes the user-reported error location. As another example, if a deviation occurs during navigation (i.e., a departure from the planned driving path), the navigation terminal can report an error location, which becomes the deviation location.

[0058] In step 202, the type of the target map feature corresponding to the error location is determined based on the error location.

[0059] In this embodiment, the target map element corresponding to the erroneous location can be the map element closest to the erroneous location. The types of target map elements include: point elements, line elements, intersection types, and path types associated with intersections. Point elements include, but are not limited to: speed limit signs, speed cameras, and points of interest (POIs). A POI is an abstract representation of any meaningful object in the real world (e.g., a shop, restaurant, bar, gas station, hospital, station, etc.) on an electronic map. Line elements include, but are not limited to: roads, lane lines, and stop lines. Map elements corresponding to intersection types include, but are not limited to: forked intersections and crossroads. Map elements corresponding to path types associated with intersections include, for example, the main road and side roads of a forked intersection, and the entrance and exit roads of a crossroads.

[0060] In step 203, the problem locations corresponding to the target map features are determined based on the type of the target map features.

[0061] In this embodiment, the problem locations correspond to different types of target map features, specifically:

[0062] If the target map feature is a point feature, its coordinates are used as the problem location. The coordinates are considered attribute information of the target map feature. The coordinates can be determined simultaneously with the acquisition of the target map feature. If the target map feature is a point feature, for example, a speed limit sign, its coordinates are directly used as the problem location; and / or,

[0063] If the target map feature is a linear feature, then select a point on the target map feature and use the coordinates of that point as the problem point. For example, if the target map feature is a lane line, then use the coordinates of one endpoint of the lane line as the problem point; and / or,

[0064] If the target map feature is of the intersection type or the path type associated with the intersection, then the coordinates of the intersection point corresponding to the target map feature will be used as the problem location.

[0065] An intersection is formed by the intersection of at least two roads. An intersection point can be understood as the location where at least two roads intersect. In the real world, an intersection has a certain shape and range and is not a single point. However, in a map, roads are represented by lines, and the intersection of two roads can be regarded as a point, which is the intersection point. The path associated with the intersection can be understood as the entry or exit route of the intersection, and its corresponding intersection point is the intersection point of the path associated with the intersection.

[0066] As can be seen, this embodiment provides an optional implementation for obtaining the problem locations corresponding to the target map features to be processed. By receiving the error locations reported by the navigation terminal, the type of the target map feature corresponding to the error location can be determined. Based on the type of the target map feature, the problem locations corresponding to the target map feature can be determined, so that map operators can use the problem locations to analyze whether there are problems with the target map feature or to analyze the problem locations through... Figure 1 Steps 102 to 104 show images recommended to map operators to help them find suitable images, verify and update map data.

[0067] Based on the above embodiments, Figure 1 Step 102, as shown, "determine at least one candidate image acquisition trajectory adjacent to the problem point," includes, for example: Figure 3 Steps 301 and 302 are shown below:

[0068] In step 301, the trajectory selection range is determined based on the problem location.

[0069] If there is only one problem point, the trajectory selection range is a circular area centered on the problem point with a preset distance as its radius. The preset distance, for example, is 100 meters. This preset distance can be set as a configurable parameter for map operators. In some embodiments, the trajectory selection range may not be a circular area and can be configured according to actual needs; and / or,

[0070] If two related problem points exist, the line connecting the two points is taken as the center line. The rectangular area extending outwards from this center line by a preset distance is the trajectory selection range. This preset distance, for example, is 100 meters. This preset distance can be set as a configurable parameter for map operators. In some embodiments, the trajectory selection range may not be a circular area; it can be configured according to actual needs. An example of a scenario with two related problem points is as follows:

[0071] For example, the starting point and the ending point of a speed camera's speed measurement zone are two related points. If a vehicle reports an error in this speed camera's speed measurement zone, the starting point and the ending point can be identified as two related problem points.

[0072] For example, if there is a first sign prohibiting overtaking and a second sign canceling the prohibition on overtaking on a certain road section, and a vehicle reports an error on that road section, the coordinates of the first and second signs can be determined as two related problem points.

[0073] For example, if a vehicle's navigation destination is a merchant in a shopping mall, and the vehicle reports an error at the destination, the coordinates of the shopping mall (as a map point of interest) and the merchant (as another map point of interest) can be identified as two related problem locations.

[0074] In step 302, the image acquisition trajectory that falls within the trajectory selection range is taken as the candidate image acquisition trajectory.

[0075] In this embodiment, the problem location is within the trajectory selection range, which can be understood as the area surrounding and adjacent to the problem location. Therefore, image acquisition trajectories falling within the trajectory selection range can be understood as trajectories adjacent to the problem location, indicating their relevance and ability to serve as a reference for map operators to find suitable images. Thus, image acquisition trajectories falling within the trajectory selection range are considered candidate image acquisition trajectories, forming a set of candidate trajectories for image recommendation to map operators. Conversely, non-adjacent image acquisition trajectories (i.e., trajectories not falling within the trajectory selection range) are redundant trajectories unrelated to or with low relevance to the problem location and should be eliminated; otherwise, they would interfere with map operators' image search.

[0076] As can be seen, this embodiment realizes the alternative to the single trajectory point proximity (i.e., determining whether two points (trajectory point and problem point) are close) candidate approach based on trajectory comparison (i.e., using the problem point to determine the trajectory selection range, judging whether the trajectory falls within the trajectory selection range, and thus determining whether the trajectory is adjacent to the problem point) to improve the recommendation accuracy.

[0077] Based on the above embodiments, Figure 1 Step 103, as shown, "based on the problem location and the positions of each trajectory point in each candidate image acquisition trajectory, select the target image acquisition trajectory from at least one candidate image acquisition trajectory," includes, for example: Figure 4 Steps 401 and 402 are shown below:

[0078] For any candidate image acquisition trajectory:

[0079] 401. Based on the problem location and the positions of each trajectory point in the candidate image acquisition trajectory, determine the relative distance of each trajectory point to the problem location, and the relative angle between the line connecting each trajectory point to the problem location and the image acquisition direction.

[0080] The relative distance between the trajectory point and the problem point is the length of the line connecting the trajectory point and the problem point. Figure 5 This is a schematic diagram illustrating how to determine the relative distance between a trajectory point and a problem point, and the relative angle between the line connecting the trajectory point and the problem point and the image acquisition direction, as provided in this embodiment of the disclosure. Figure 5In the image, the relative distance between trajectory point A and problem point B is the length of line segment AB, and the relative angle between the line connecting trajectory point A and problem point B and the image acquisition direction is θ.

[0081] 402. Based on the relative distance and relative angle of each trajectory point, determine whether the candidate image acquisition trajectory is the target image acquisition trajectory.

[0082] In this embodiment, the relative distance of each trajectory point is the relative distance of each trajectory point to the problem point, and the relative angle of each trajectory point is the relative angle between the line connecting each trajectory point and the problem point and the image acquisition direction. It can be seen that the relative distance and relative angle constitute the trajectory features associated with the problem point. The trajectory features are used to filter candidate image acquisition trajectories, eliminating candidate image acquisition trajectories that are irrelevant to the problem point in abnormal acquisition scenarios. The remaining candidate image acquisition trajectories are related to the problem point, so the remaining candidate image acquisition trajectories are used as the target image acquisition trajectories for image recommendation to map operators. Here, abnormal acquisition scenarios are abnormal relative to the problem point.

[0083] An abnormal acquisition scenario is as follows: The image acquisition device (e.g., a camera) installed on the acquisition vehicle acquires images of the front and sides of the vehicle, but cannot acquire images of the rear of the vehicle. Therefore, based on the angle between the line connecting each trajectory point and the problem point and the image acquisition direction, it can be determined whether the image corresponding to each trajectory point is an image of the rear of the vehicle. If the image corresponding to a certain trajectory point is an image of the rear of the vehicle, it means that the image corresponding to that trajectory point requires changing the position and orientation of the camera on the vehicle to capture the image. However, in reality, the position of the camera on the vehicle is usually fixed. Therefore, the image corresponding to that trajectory point is an abnormally captured image relative to the problem point, and belongs to the interference image. It means that the candidate image acquisition trajectory where the trajectory point is located is irrelevant to the problem point and should be eliminated.

[0084] As can be seen, in this embodiment, by determining the candidate image acquisition trajectory and the target image acquisition trajectory, the image acquisition trajectory is screened twice using trajectory features (including the relative distance of each trajectory point to the problem point, and the relative angle between the line connecting each trajectory point to the problem point and the image acquisition direction). This eliminates redundant trajectories that may interfere with map operators, thus helping map operators find suitable images more quickly.

[0085] In some embodiments, step 402, "determining whether a candidate image acquisition trajectory is a target image acquisition trajectory based on the relative distance and relative angle corresponding to each trajectory point," includes, for example: Figure 6 Steps 601 to 603 are shown below:

[0086] In step 601, a distance threshold is determined based on the relative distances corresponding to each trajectory point.

[0087] In this embodiment, considering the significant differences in the relative distances between trajectory points and problem points in different candidate image acquisition trajectories, using a fixed distance threshold for trajectory filtering might result in the rejection of useful trajectories. Therefore, this embodiment dynamically determines the distance threshold for each image acquisition trajectory by utilizing the relative distances between each trajectory point and the problem point. One possible implementation method for determining the distance threshold is as follows:

[0088] For any candidate image acquisition trajectory, the average distance and standard deviation of the distance are determined based on the relative distances of each trajectory point to the problem point. Based on the average distance and standard deviation, a distance threshold corresponding to the candidate image acquisition trajectory is determined. For example, if the distance threshold is denoted as len_max, then len_max = k1 × average distance + k2 × standard deviation, where k1 and k2 are positive constants. In some embodiments, k1 is 1 and k2 is 0.5, then len_max = average distance + 0.5 × standard deviation. Those skilled in the art can configure the values ​​of k1 and k2 according to actual needs; this embodiment does not limit the specific values ​​of k1 and k2.

[0089] As can be seen, in this embodiment, by refining the distance average and distance standard deviation, the distance threshold corresponding to each image acquisition trajectory is dynamically determined, avoiding the problem of rejecting useful trajectories due to using a fixed distance threshold for trajectory filtering, thus improving the accuracy of recommendations.

[0090] In some embodiments, the distance threshold is a configurable item and can be pre-configured manually. Those skilled in the art can configure the specific value of the distance threshold according to the actual scenario. This embodiment does not limit the specific value.

[0091] In step 602, the target distance from the problem point to the candidate image acquisition trajectory is determined.

[0092] In this embodiment, the problem point is projected onto the candidate image acquisition trajectory to obtain the projection position (i.e., the foot of the perpendicular) of the problem point on the candidate image acquisition trajectory. Then, the length of the line connecting the problem point and the projection position is taken as the target distance from the problem point to the candidate image acquisition trajectory.

[0093] In step 603, if the target distance is less than or equal to the distance threshold, and the relative angles corresponding to each trajectory point are all less than or equal to the preset angle threshold, then the candidate image acquisition trajectory is determined to be the target image acquisition trajectory.

[0094] In this embodiment, for any candidate image acquisition trajectory, it is determined whether the target distance from the problem point to the candidate image acquisition trajectory is greater than the distance threshold corresponding to the candidate image acquisition trajectory. If the target distance is greater than the distance threshold, it indicates that the candidate image acquisition trajectory has little correlation with the problem point and should be removed from the candidate set; otherwise, it will interfere with the map operators' image search. If the target distance is less than or equal to the distance threshold, it is further determined whether the relative angle between the line connecting each trajectory point and the problem point and the image acquisition direction is less than or equal to a preset angle threshold. If they are all less than or equal to the preset angle threshold, it indicates that the problem point is within the field of view of the image acquisition device relative to each trajectory point, and the problem point can be normally acquired by the image acquisition device. Therefore, the candidate image acquisition trajectory can be determined as the target image acquisition trajectory, which serves as the basis for recommending images to map operators. The angle threshold is determined by the field of view of the image acquisition device. For example, if the field of view is 180°, the angle threshold is half of the field of view, i.e., 90°.

[0095] As can be seen, in this embodiment, by determining the candidate image acquisition trajectory and the target image acquisition trajectory, the image acquisition trajectory is screened twice using trajectory features (including the relative distance of each trajectory point to the problem point, and the relative angle between the line connecting each trajectory point to the problem point and the image acquisition direction). This eliminates redundant trajectories that may interfere with map operators, thus helping map operators find suitable images more quickly.

[0096] In some embodiments, Figure 4 Before step 401, which involves determining the relative distance between each trajectory point and the problem point based on the location of the problem point and the position of each trajectory point in the candidate image acquisition trajectory, and the relative angle between the line connecting each trajectory point and the problem point and the image acquisition direction, the image recommendation method further includes: determining the time interval between the acquisition time of the candidate image acquisition trajectory and the current time. If the time interval is less than or equal to a preset time interval threshold, then step 401 is executed: determining the relative distance between each trajectory point and the problem point based on the location of the problem point and the position of each trajectory point in the candidate image acquisition trajectory, and the relative angle between the line connecting each trajectory point and the problem point and the image acquisition direction.

[0097] In this embodiment, the preset time interval threshold can be understood as the lifespan of the candidate image acquisition trajectory. Exceeding the lifespan only indicates that the candidate image acquisition trajectory is outdated and unsuitable as a basis for recommending images to map operators, but it does not mean that the trajectory should be deleted. For example, the preset time interval threshold is 2 years. If the time interval between the acquisition time of the candidate image acquisition trajectory and the current time is greater than 2 years, the candidate image acquisition trajectory will be removed from the candidate set; otherwise, it will interfere with map operators' image search.

[0098] Based on the above embodiments, Figure 1 Step 104, "selecting at least one image corresponding to a trajectory point from the target image acquisition trajectory for recommendation," includes, for example: Figure 7 Steps 701 to 703 are shown below:

[0099] In step 701, if there are multiple target image acquisition trajectories, it is determined whether there is a trajectory representative point in each target image acquisition trajectory, and / or the acquisition time of each target image acquisition trajectory, and / or the clarity of the acquired image corresponding to each target image acquisition trajectory.

[0100] In this embodiment, for any target image acquisition trajectory with a representative trajectory point, the representative trajectory point can be understood as the trajectory point in the target image acquisition trajectory with the highest matching degree between the acquired image and the target map elements. How the representative trajectory point is determined is described below.

[0101] In step 702, the recommended priority of multiple target image acquisition trajectories is determined based on whether there is a trajectory representative point in each target image acquisition trajectory, and / or the acquisition time of each target image acquisition trajectory, and / or the clarity of the acquired image corresponding to each target image acquisition trajectory.

[0102] In this embodiment, the recommendation priority of target image acquisition trajectories with representative trajectory points is higher than that of target image acquisition trajectories without representative trajectory points; and / or, the recommendation priority of target image acquisition trajectories with updated acquisition time is higher; and / or, the recommendation priority of target image acquisition trajectories with higher image clarity is higher.

[0103] For example, firstly, the recommendation priority of target image acquisition trajectories with representative trajectory points is determined to be higher than that of target image acquisition trajectories without representative trajectory points. Then, for multiple target image acquisition trajectories with representative trajectory points, the target image acquisition trajectory with the more recent acquisition time has a higher recommendation priority. Similarly, for multiple target image acquisition trajectories without representative trajectory points, the target image acquisition trajectory with the more recent acquisition time has a higher recommendation priority. Finally, for multiple target image acquisition trajectories with representative trajectory points, if the recommendation priority cannot be determined based on the acquisition time (e.g., the acquisition time is the same), then the target image acquisition trajectory with higher image clarity has a higher recommendation priority. Similarly, for multiple target image acquisition trajectories without representative trajectory points, if the acquisition time is the same, then the target image acquisition trajectory with higher image clarity has a higher recommendation priority.

[0104] In step 703, at least one image corresponding to a trajectory point is selected from each target image acquisition trajectory for recommendation in descending order of recommendation priority.

[0105] In this embodiment, there are two target image acquisition trajectories, denoted as Trajectory 1 and Trajectory 2. Trajectory 1 has a higher recommendation priority than Trajectory 2. Therefore, at least one image corresponding to a trajectory point is selected from Trajectory 1 for recommendation, and then at least one image corresponding to a trajectory point is selected from Trajectory 2 for recommendation. It is evident that, for both Trajectory 1 and Trajectory 2, not all images corresponding to trajectory points are recommended; instead, images corresponding to a subset of trajectory points are selected for recommendation, thus improving recommendation accuracy.

[0106] Based on the above embodiments, Figure 7 Step 701, "determining whether there are representative points in each target image acquisition trajectory," includes, for example: Figure 8 Steps 801 to 805 are shown below:

[0107] For any target image acquisition trajectory:

[0108] 801. Determine the projection position of the problem point on the target image acquisition trajectory.

[0109] 802. Determine the relative distance between each trajectory point in the target image acquisition trajectory and the projection position.

[0110] 803. Determine multiple candidate trajectory points in order of increasing relative distance.

[0111] 804. Perform map feature recognition on the images corresponding to multiple candidate trajectory points to obtain the map features corresponding to each candidate trajectory point, and match the map features corresponding to each candidate trajectory point with the target map features.

[0112] Map feature recognition is a routine operation in the field of map technology, while map feature matching belongs to image matching and is a mature technology in the field of image processing. To avoid repetition, it will not be elaborated further.

[0113] 805. If at least one candidate trajectory point corresponds to a map feature that matches the target map feature, then select the candidate trajectory point with the highest matching degree from at least one candidate trajectory point as the trajectory representative point.

[0114] It should be noted that if the matching confidence of the candidate trajectory point with the highest matching degree is less than the preset matching confidence threshold, it is determined that the target image acquisition trajectory has no representative trajectory point. The matching confidence is automatically output by the matching algorithm during the matching process in step 804. A matching confidence score less than the preset matching confidence threshold indicates that the matching confidence is unreliable.

[0115] As can be seen, in this embodiment, the trajectory representative point is determined by image matching. The trajectory representative point can be understood as the trajectory point in the target image acquisition trajectory where the map elements included in the acquired image have the highest matching degree with the target map elements.

[0116] Based on the above embodiments, Figure 1 Step 104, "selecting at least one image corresponding to a trajectory point from the target image acquisition trajectory for recommendation," includes, for example: Figure 9 Steps 901 and 902 are shown below:

[0117] In step 901, the images corresponding to the representative points of the trajectory in the target image acquisition trajectory are recommended as the first priority.

[0118] Among them, the map features included in the images collected at the trajectory representative points had the highest matching degree with the target map features.

[0119] In step 902, along the image acquisition direction of the target image acquisition trajectory, the image corresponding to the next trajectory point of the trajectory representative point is recommended as the second order, the image corresponding to the previous trajectory point of the trajectory representative point is recommended as the third order, the images corresponding to the next two trajectory points of the trajectory representative point are recommended as the fourth order, the images corresponding to the previous two trajectory points of the trajectory representative point are recommended as the fifth order, and so on, the images corresponding to the trajectory points before and after the trajectory representative point are recommended alternately, and the number of recommended images is a preset number, which is a positive integer.

[0120] For example, Figure 10 This is a schematic diagram illustrating a scenario where at least one image corresponding to a trajectory point is selected from a target image acquisition trajectory for recommendation, as provided in this embodiment of the disclosure. Figure 10In the process, the image corresponding to the trajectory representative point (1) in the target image acquisition trajectory is recommended as the first order. Along the image acquisition direction of the target image acquisition trajectory, the image corresponding to the next trajectory point (2) of the trajectory representative point is recommended as the second order, the image corresponding to the previous trajectory point (3) of the trajectory representative point is recommended as the third order, the images corresponding to the next two trajectory points (4) of the trajectory representative point are recommended as the fourth order, the images corresponding to the previous two trajectory points (5) of the trajectory representative point are recommended as the fifth order, and so on. The images corresponding to the trajectory points before and after the trajectory representative point are recommended alternately, and the number of recommended images is a preset number. The preset number is a configurable parameter that is manually pre-configured.

[0121] In this embodiment, alternating recommendations do not require waiting for map operators to confirm the image content; the recommended images can be displayed synchronously in the recommended order. Alternatively, alternating recommendations can display only one image at a time, and after the map operators confirm the image content, if the map operators do not select the image, the next image is displayed in the recommended order.

[0122] As can be seen, in this embodiment, not all images corresponding to all trajectory points in the target image acquisition trajectory are recommended. Instead, images corresponding to at least one trajectory point with a high degree of matching with the target map elements are selected for recommendation. Furthermore, the recommendation is performed in an alternating manner to prevent the problem of cumulative error caused by unidirectional sequential recommendation and to improve the accuracy of recommendation.

[0123] Based on the above embodiments, Figure 11 This is a flowchart illustrating an image recommendation method provided in an embodiment of the present disclosure, including the following steps 1101 to 1106:

[0124] 1101. Selection of problem location.

[0125] In this embodiment, the system receives erroneous locations reported by the navigation terminal, such as user-reported locations or vehicle deviation locations. Then, based on the erroneous location, it determines the type of the target map element corresponding to the erroneous location. The types of target map elements include: point elements, line elements, intersection types, and path types associated with intersections. Therefore, based on the type of the target map element, it determines the corresponding problem location. If the target map element is a point element, its coordinates are used as the problem location. If the target map element is a line element, a point is selected from the target map element, and its coordinates are used as the problem location. If the target map element is an intersection type or a path type associated with an intersection, the coordinates of the intersection point corresponding to the target map element are used as the problem location.

[0126] 1102. Determination of candidate image acquisition trajectory.

[0127] In this embodiment, the trajectory selection range is determined based on the problem location. For example, if there is only one problem location, the trajectory selection range is a circular area centered on the problem location and with a preset distance as the radius. If there are two related problem locations, the line connecting the two related problem locations is used as the center line, and the rectangular area formed by extending a preset distance outward from both sides of the center line is the trajectory selection range. Image acquisition trajectories falling within the trajectory selection range are used as candidate image acquisition trajectories.

[0128] 1103. Calculation of relative angles and relative distances.

[0129] In this embodiment, the time interval between the acquisition time of the candidate image acquisition trajectory and the current time is determined. If the time interval is less than or equal to a preset time interval threshold, then based on the problem point and the position of each trajectory point in the candidate image acquisition trajectory, the relative distance of each trajectory point to the problem point, and the relative angle between the line connecting each trajectory point to the problem point and the image acquisition direction are determined. For example... Figure 5 In the image, the relative distance between trajectory point A and problem point B is the length of line segment AB, and the relative angle between the line connecting trajectory point A and problem point B and the image acquisition direction is θ.

[0130] 1104. Determination of the target image acquisition trajectory.

[0131] In this embodiment, the candidate image acquisition trajectory is determined to be the target image acquisition trajectory based on the relative distance of each trajectory point to the problem point and the relative angle between the line connecting each trajectory point and the problem point and the image acquisition direction. Specifically, a distance threshold is determined based on the relative distance of each trajectory point to the problem point; a target distance from the problem point to the candidate image acquisition trajectory is determined; if the target distance is less than or equal to the distance threshold, and the relative angle between the line connecting each trajectory point and the problem point and the image acquisition direction is less than or equal to a preset angle threshold, then the candidate image acquisition trajectory is determined to be the target image acquisition trajectory.

[0132] 1105. Sorting of target image acquisition trajectories.

[0133] In this embodiment, firstly, the recommendation priority of target image acquisition trajectories with representative trajectory points is determined to be higher than that of target image acquisition trajectories without representative trajectory points. Then, for multiple target image acquisition trajectories with representative trajectory points, the target image acquisition trajectory with the more recent acquisition time has a higher recommendation priority. Similarly, for multiple target image acquisition trajectories without representative trajectory points, the target image acquisition trajectory with the more recent acquisition time has a higher recommendation priority. Finally, for multiple target image acquisition trajectories with representative trajectory points, if the recommendation priority cannot be determined based on the acquisition time (e.g., the acquisition times are all the same), the target image acquisition trajectory with higher image clarity has a higher recommendation priority. Similarly, for multiple target image acquisition trajectories without representative trajectory points, if the acquisition times are all the same, the target image acquisition trajectory with higher image clarity has a higher recommendation priority.

[0134] 1106. Extraction of the target image acquisition trajectory.

[0135] For example Figure 10 In the process, the image corresponding to the trajectory representative point (1) in the target image acquisition trajectory is recommended as the first order. Along the image acquisition direction of the target image acquisition trajectory, the image corresponding to the next trajectory point (2) of the trajectory representative point is recommended as the second order, the image corresponding to the previous trajectory point (3) of the trajectory representative point is recommended as the third order, the images corresponding to the next two trajectory points (4) of the trajectory representative point are recommended as the fourth order, the images corresponding to the previous two trajectory points (5) of the trajectory representative point are recommended as the fifth order, and so on, and the images corresponding to the trajectory points before and after the trajectory representative point are recommended alternately.

[0136] As can be seen, in at least one embodiment of this disclosure, by obtaining the problem point corresponding to the target map element to be processed, at least one candidate image acquisition trajectory adjacent to the problem point is determined. This realizes the candidate approach of replacing the single trajectory point proximity (i.e., determining whether the trajectory is adjacent to the problem point) with trajectory comparison (i.e., determining whether two points (trajectory point and problem point) are close) and improves the recommendation accuracy. Furthermore, based on the problem point and the position of each trajectory point in the candidate image acquisition trajectory, the candidate image trajectory is further filtered using trajectory features (including the position of trajectory points) to obtain the target image acquisition trajectory as the basis for recommending images to map operators. Thus, at least one image corresponding to a trajectory point is selected from the target image acquisition trajectory for recommendation. This realizes a multi-feature fusion matching image recommendation method that combines trajectory features and image features, improving the recommendation accuracy and assisting map operators in finding suitable images to verify and update map data.

[0137] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art will understand that the embodiments of this disclosure are not limited to the described order of actions, because according to the embodiments of this disclosure, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art will understand that the embodiments described in the specification are all optional embodiments.

[0138] Figure 12 This is a schematic diagram of an image recommendation device provided in an embodiment of this disclosure. The image recommendation device can be applied to electronic devices, including but not limited to in-vehicle devices, smartphones, PDAs, tablets, wearable devices with displays, desktop computers, laptops, all-in-one computers, smart home devices, servers, etc. The server can be a standalone server or a cluster of multiple servers, and can include locally located servers and cloud-based servers. The image recommendation device provided in this embodiment of the disclosure can execute the processing flow provided in various embodiments of the image recommendation method, such as... Figure 12 As shown, the image recommendation device includes, but is not limited to: an acquisition unit 1201, a determination unit 1202, a selection unit 1203, and a recommendation unit 1204. The functions of each unit are described below:

[0139] Acquisition unit 1201 is used to acquire the problem points corresponding to the target map features to be processed;

[0140] The determining unit 1202 is used to determine at least one candidate image acquisition trajectory adjacent to the problem point;

[0141] Selection unit 1203 is used to select a target image acquisition trajectory from at least one candidate image acquisition trajectory based on the problem point location and the position of each trajectory point in each candidate image acquisition trajectory.

[0142] The recommendation unit 1204 is used to select at least one image corresponding to a trajectory point from the target image acquisition trajectory for recommendation.

[0143] In some embodiments, the obtaining unit 1201 is configured to:

[0144] Receive error location reports from the navigation system;

[0145] Based on the error location, determine the type of the target map feature corresponding to the error location;

[0146] Based on the type of the target map features, determine the corresponding problem locations.

[0147] In some embodiments, the acquisition unit 1201 determines the problem location corresponding to the target map feature based on the type of the target map feature, including:

[0148] If the target map feature is a point feature, then the coordinates of the target map feature will be used as the problem point; and / or,

[0149] If the target map feature is a linear feature, then select a point on the target map feature and use the coordinates of that point as the problem point; and / or,

[0150] If the target map feature is of the intersection type or the path type associated with the intersection, then the coordinates of the intersection point corresponding to the target map feature will be used as the problem location.

[0151] In some embodiments, the determining unit 1202 is configured to:

[0152] Based on the problem locations, determine the range of trajectory selection;

[0153] Image acquisition trajectories that fall within the trajectory selection range are used as candidate image acquisition trajectories.

[0154] In some embodiments, the determining unit 1202 determines the trajectory selection range based on the problem location, including:

[0155] If there is only one problem point, then the trajectory selection range is a circular area centered on the problem point with a preset distance as the radius; and / or,

[0156] If there are two related problem points, the line connecting the two related problem points will be the center line, and the rectangular area formed by extending a preset distance on both sides of the center line will be the trajectory selection range.

[0157] In some embodiments, the selection unit 1203 is used for:

[0158] For any candidate image acquisition trajectory:

[0159] Based on the problem location and the positions of each trajectory point in the candidate image acquisition trajectory, determine the relative distance of each trajectory point to the problem location, as well as the relative angle between the line connecting each trajectory point to the problem location and the image acquisition direction.

[0160] Based on the relative distance and relative angle of each trajectory point, it is determined whether the candidate image acquisition trajectory is the target image acquisition trajectory.

[0161] In some embodiments, the selection unit 1203 determines whether a candidate image acquisition trajectory is a target image acquisition trajectory based on the distance of each trajectory point relative to the problem point, and the angle between the line connecting each trajectory point and the problem point and the image acquisition direction, including:

[0162] The distance threshold is determined based on the relative distances between each trajectory point;

[0163] Determine the target distance from the problem location to the candidate image acquisition trajectory;

[0164] If the target distance is less than or equal to the distance threshold, and the relative angles corresponding to each trajectory point are all less than or equal to the preset angle threshold, then the candidate image acquisition trajectory is determined to be the target image acquisition trajectory.

[0165] In some embodiments, before the selection unit 1203 determines the distance of each trajectory point relative to the problem point based on the problem point location and the positions of each trajectory point in the candidate image acquisition trajectory, and before determining the angle between the line connecting each trajectory point and the problem point location and the image acquisition direction, it is further configured to:

[0166] Determine the time interval between the acquisition time of the candidate image acquisition trajectory and the current time;

[0167] If the time interval is less than or equal to the preset time interval threshold, then based on the problem location and the position of each trajectory point in the candidate image acquisition trajectory, the relative distance of each trajectory point to the problem location and the relative angle between the line connecting each trajectory point to the problem location and the image acquisition direction are determined.

[0168] In some embodiments, the recommendation unit 1204 is configured to:

[0169] If there are multiple target image acquisition trajectories, determine whether there is a trajectory representative point in each target image acquisition trajectory, and / or the acquisition time of each target image acquisition trajectory, and / or the clarity of the acquired image corresponding to each target image acquisition trajectory;

[0170] The recommendation priority of multiple target image acquisition trajectories is determined based on whether there are representative points in each target image acquisition trajectory, and / or the acquisition time of each target image acquisition trajectory, and / or the clarity of the acquired image corresponding to each target image acquisition trajectory.

[0171] Based on the recommendation priority from high to low, at least one image corresponding to each trajectory point is selected from each target image acquisition trajectory for recommendation.

[0172] In some embodiments, the recommendation unit 1204 determines whether there is a trajectory representative point in each target image acquisition trajectory, including:

[0173] For any target image acquisition trajectory:

[0174] Determine the projection position of the problem point on the target image acquisition trajectory;

[0175] Determine the relative distance between each trajectory point in the target image acquisition trajectory and the projection position;

[0176] Multiple candidate trajectory points were determined in ascending order of their relative distances;

[0177] Map feature recognition is performed on the images corresponding to multiple candidate trajectory points to obtain the map features corresponding to each candidate trajectory point, and the map features corresponding to each candidate trajectory point are matched with the target map features.

[0178] If at least one candidate trajectory point corresponds to a map feature that matches the target map feature, then the candidate trajectory point with the highest matching degree is selected from at least one candidate trajectory point as the trajectory representative point.

[0179] In some embodiments, the recommendation unit 1204 determines the recommendation priority of multiple target image acquisition trajectories based on whether there is a trajectory representative point in each target image acquisition trajectory, and / or, the acquisition time of each target image acquisition trajectory, and / or, the clarity of the acquired image corresponding to each target image acquisition trajectory, including:

[0180] The recommended priority for target image acquisition trajectories with representative trajectory points is higher than that for target image acquisition trajectories without representative trajectory points; and / or,

[0181] The recommended acquisition trajectory for the target image, updated at the acquisition time, has a higher priority; and / or,

[0182] The acquisition trajectory of the target image with higher image clarity has a higher recommendation priority.

[0183] In some embodiments, the recommendation unit 1204 is configured to:

[0184] The images corresponding to the representative points of the target image acquisition trajectory are recommended in the first order, among which the map elements included in the images acquired at the representative points of the trajectory have the highest matching degree with the target map elements.

[0185] Along the image acquisition direction of the target image acquisition trajectory, the image corresponding to the next trajectory point of the trajectory representative point is recommended as the second order, the image corresponding to the previous trajectory point of the trajectory representative point is recommended as the third order, the images corresponding to the next two trajectory points of the trajectory representative point are recommended as the fourth order, the images corresponding to the previous two trajectory points of the trajectory representative point are recommended as the fifth order, and so on, alternately recommending the images corresponding to the trajectory points before and after the trajectory representative point, and the number of recommended images is a preset number, which is a positive integer.

[0186] As can be seen, in at least one embodiment of the image recommendation device of this disclosure, by acquiring the problem point corresponding to the target map element to be processed, at least one candidate image acquisition trajectory adjacent to the problem point is determined. This realizes the candidate approach of replacing the single trajectory point proximity (i.e., determining whether the trajectory is adjacent to the problem point) with trajectory comparison (i.e., determining whether two points (trajectory point and problem point) are close) and improves the recommendation accuracy. Furthermore, based on the problem point and the position of each trajectory point in the candidate image acquisition trajectory, the candidate image trajectory is further filtered using trajectory features (including the position of trajectory points) to obtain the target image acquisition trajectory as the basis for recommending images to map operators. Thus, multiple images corresponding to multiple trajectory points are selected from the target image acquisition trajectory for recommendation. This realizes a multi-feature fusion matching image recommendation method that combines trajectory features and image features, improving the recommendation accuracy and assisting map operators in finding suitable images to verify and update map data.

[0187] Figure 13 This is an exemplary block diagram of an electronic device provided in an embodiment of this disclosure. Figure 13 As shown, the electronic device includes: a memory 1301, a processor 1302, and a computer program stored on the memory 1301. It is understood that the memory 1301 in this embodiment may be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory.

[0188] In some implementations, memory 1301 stores elements such as executable modules or data structures, or subsets thereof, or extended sets thereof: operating systems and applications.

[0189] The operating system includes various system programs, such as the framework layer, core library layer, and driver layer, used to implement various basic tasks and handle hardware-based tasks. The application programs include various applications, such as media players and browsers, used to implement various application tasks. The program implementing the image recommendation method provided in this disclosure can be included in the application programs.

[0190] In this embodiment of the disclosure, at least one processor 1302 executes the steps of the various embodiments of the image recommendation method provided in this disclosure by calling a program or instruction stored in at least one memory 1301, specifically, a program or instruction stored in an application.

[0191] The image recommendation method provided in this disclosure can be applied to, or implemented by, processor 1302. Processor 1302 can be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method can be completed by integrated logic circuits in the hardware of processor 1302 or by instructions in software form. The processor 1302 can be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. The general-purpose processor can be a microprocessor or any conventional processor.

[0192] The steps of the image recommendation method provided in this disclosure can be directly implemented by a hardware decoding processor, or implemented by a combination of hardware and software modules in the decoding processor. The software modules can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. This storage medium is located in memory 1301, and processor 1302 reads the information in memory 1301 and combines it with hardware to complete the steps of the method.

[0193] This disclosure also proposes a computer-readable storage medium that stores a program or instructions that cause a computer to perform the steps of the various embodiments of the image recommendation method; to avoid repetition, these steps will not be repeated here. The computer-readable storage medium can be a non-transitory computer-readable storage medium.

[0194] This disclosure also proposes a computer program product comprising a computer program stored in a computer-readable storage medium, which may be a non-transitory computer-readable storage medium. At least one processor of a computer reads from and executes the computer program from the computer-readable storage medium, causing the computer to perform the steps of the various embodiments of the image recommendation method, which will not be repeated here to avoid repetition.

[0195] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0196] Those skilled in the art will understand that although some embodiments described herein include certain features included in other embodiments but not others, combinations of features from different embodiments are meant to be within the scope of this disclosure and form different embodiments.

[0197] Those skilled in the art will understand that the descriptions of the various embodiments have different focuses, and for parts not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0198] Although embodiments of the present disclosure have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the present disclosure, and all such modifications and variations fall within the scope defined by the appended claims.

Claims

1. An image recommendation method, the method comprising: Obtain the problem points corresponding to the target map elements to be processed. The problem points are the coordinate locations used to analyze whether there are problems with the target map elements. Identify at least one candidate image acquisition trajectory adjacent to the problem location; For any candidate image acquisition trajectory: based on the problem point and the position of each trajectory point in the candidate image acquisition trajectory, determine the relative distance of each trajectory point relative to the problem point, and the relative angle between the line connecting each trajectory point and the problem point and the image acquisition direction; based on the relative distance and the relative angle corresponding to each trajectory point, determine whether the candidate image acquisition trajectory is the target image acquisition trajectory; Recommend images by selecting at least one trajectory point from the target image acquisition trajectory; The step of obtaining the problem locations corresponding to the target map features to be processed includes: Receive error location reports from the navigation system; Based on the error location, determine the type of the target map feature corresponding to the error location; Based on the type of the target map element, the problem location corresponding to the target map element is determined.

2. The method according to claim 1, wherein, The step of determining the problem location corresponding to the target map feature based on the type of the target map feature includes: If the target map feature is a point feature, then the coordinates of the target map feature are taken as the problem point; and / or, If the target map feature is a linear feature, then select a point from the target map feature and use the coordinates of that point as the problem location; and / or, If the target map element is of the intersection type or the path type associated with the intersection, then the coordinates of the intersection point corresponding to the target map element will be used as the problem point.

3. The method according to claim 1, wherein, Determining at least one candidate image acquisition trajectory adjacent to the problem location includes: Based on the identified problem locations, the trajectory selection range is determined; Image acquisition trajectories falling within the trajectory selection range are selected as candidate image acquisition trajectories.

4. The method according to claim 3, wherein, The process of determining the trajectory selection range based on the problem location includes: If there is only one problem location, then the trajectory selection range is defined by a circular area centered on the problem location and with a preset distance as the radius; and / or, If there are two related problem points, the line connecting the two related problem points is the center line, and the rectangular area formed by extending a preset distance outward on both sides of the center line is the trajectory selection range.

5. The method according to claim 1, wherein, Determining whether the candidate image acquisition trajectory is the target image acquisition trajectory based on the relative distance and relative angle corresponding to each trajectory point includes: Based on the relative distances corresponding to each trajectory point, a distance threshold is determined; Determine the target distance from the problem location to the candidate image acquisition trajectory; If the target distance is less than or equal to the distance threshold, and the relative angles corresponding to each trajectory point are all less than or equal to a preset angle threshold, then the candidate image acquisition trajectory is determined to be the target image acquisition trajectory.

6. The method according to claim 1, wherein, The step of selecting at least one image corresponding to a trajectory point from the target image acquisition trajectory for recommendation includes: If there are multiple target image acquisition trajectories, then determine whether there is a trajectory representative point in each target image acquisition trajectory, and / or the acquisition time of each target image acquisition trajectory, and / or the clarity of the acquired image corresponding to each target image acquisition trajectory; The recommendation priority of the multiple target image acquisition trajectories is determined based on whether there is a trajectory representative point in each target image acquisition trajectory, and / or the acquisition time of each target image acquisition trajectory, and / or the clarity of the acquired image corresponding to each target image acquisition trajectory. According to the recommendation priority from high to low, at least one image corresponding to each trajectory point is selected from each target image acquisition trajectory for recommendation.

7. The method according to claim 6, wherein, Determining whether a trajectory representative point exists in each of the target image acquisition trajectories includes: For any of the target image acquisition trajectories: Determine the projection position of the problem point on the target image acquisition trajectory; Determine the relative distance between each trajectory point in the target image acquisition trajectory and the projection position; Multiple candidate trajectory points are determined according to the order of their relative distances from smallest to largest; Map feature recognition is performed on the images corresponding to the multiple candidate trajectory points to obtain the map features corresponding to each candidate trajectory point, and the map features corresponding to each candidate trajectory point are matched with the target map features; If at least one candidate trajectory point corresponds to a map feature that matches the target map feature, then the candidate trajectory point with the highest matching degree is selected from the at least one candidate trajectory point as the trajectory representative point.

8. The method according to claim 1, wherein, The step of selecting at least one image corresponding to a trajectory point from the target image acquisition trajectory for recommendation includes: The images corresponding to the trajectory representative points in the target image acquisition trajectory are recommended as the first order, wherein the map elements included in the images acquired at the trajectory representative points have the highest matching degree with the target map elements; Along the image acquisition direction of the target image acquisition trajectory, the image corresponding to the next trajectory point of the trajectory representative point is recommended as the second order, the image corresponding to the previous trajectory point of the trajectory representative point is recommended as the third order, the images corresponding to the next two trajectory points of the trajectory representative point are recommended as the fourth order, the images corresponding to the previous two trajectory points of the trajectory representative point are recommended as the fifth order, and so on, alternately recommending the images corresponding to the trajectory points before and after the trajectory representative point, and the number of recommended images is a preset number, which is a positive integer.

9. An image recommendation device, the device comprising: The acquisition unit is used to acquire the problem points corresponding to the target map elements to be processed. The problem points are the coordinate positions used to analyze whether there are problems with the target map elements. The determining unit is used to determine at least one candidate image acquisition trajectory adjacent to the problem location; The selection unit is used for any candidate image acquisition trajectory to: determine the relative distance of each trajectory point relative to the problem point, and the relative angle between the line connecting each trajectory point and the problem point and the image acquisition direction, based on the position of the problem point and the position of each trajectory point in the candidate image acquisition trajectory; Based on the relative distance and relative angle corresponding to each trajectory point, determine whether the candidate image acquisition trajectory is the target image acquisition trajectory; The recommendation unit is used to select at least one image corresponding to a trajectory point from the target image acquisition trajectory for recommendation; The step of obtaining the problem locations corresponding to the target map features to be processed includes: Receive error location reports from the navigation system; Based on the error location, determine the type of the target map feature corresponding to the error location; Based on the type of the target map element, the problem location corresponding to the target map element is determined.

10. An electronic device, wherein, The method includes a memory, a processor, and a computer program stored on the memory, wherein the processor executes the computer program to implement the steps of the image recommendation method as claimed in any one of claims 1 to 9.

11. A computer-readable storage medium, wherein, The computer-readable storage medium stores a program or instructions that cause a computer to perform the steps of the image recommendation method as described in any one of claims 1 to 9.

Images