Map correction method and device, electronic device, storage medium and program product

Abstract

The embodiment of the application discloses a map correction method and device, an electronic device, a storage medium and a program product. The sample road image collected in real time is acquired, the target road image can be corrected in time, and the efficiency of map correction is improved. The first similarity between the first position information and the second position information is used to quickly and accurately determine the target road image which is more matched with the geographical position of the sample road image from the plurality of candidate road images. The second similarity between the current road image and the target road image is introduced to correct the target road image, the deviation between the image of the map in the map database and the actual situation can be accurately repaired, and the accuracy of the map is improved in real time.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to a map correction method and apparatus, electronic device, storage medium, and program product. Background Technology

[0002] Currently, maps are used in an increasing number of scenarios, such as virtual reality games and map navigation. Therefore, map accuracy is particularly important. For example, in map navigation applications, road malfunctions or road construction often lead to navigation errors, necessitating map updates to ensure accuracy. Currently, map updates are typically done manually; however, the efficiency and accuracy of this method need improvement. Summary of the Invention

[0003] The following is an overview of the subject matter described in detail herein. This overview is not intended to limit the scope of the claims.

[0004] This application provides a map correction method and apparatus, electronic device, storage medium, and program product that can improve map accuracy in real time.

[0005] On the one hand, embodiments of this application provide a map correction method, including:

[0006] Acquire sample road images captured in real time by an image acquisition device, and the first location information of the sample road images;

[0007] Obtain the second location information of each candidate road image in the map database;

[0008] A first similarity is determined between the first location information and the second location information, and a target road image is determined from a plurality of candidate road images based on the first similarity;

[0009] A second similarity is determined between the sample road image and the target road image, and the target road image is corrected based on the second similarity.

[0010] On the other hand, embodiments of this application also provide a map correction device, including:

[0011] The first acquisition module is used to acquire real-time sample road images and the first location information of the sample road images;

[0012] The second acquisition module is used to acquire the second location information of each candidate road image in the map database;

[0013] A target road image determination module is used to determine a first similarity between the first location information and the second location information, and to determine a target road image from a plurality of candidate road images based on the first similarity.

[0014] The correction module is used to determine a second similarity between the sample road image and the target road image, and to correct the target road image based on the second similarity.

[0015] Furthermore, the first location information is first location coordinate information, and the second location information is second location coordinate information. The aforementioned target road image determination module is specifically used for:

[0016] Determine the difference in abscissa between the first position coordinate information and the second position coordinate information, and the difference in ordinate between the first position coordinate information and the second position coordinate information;

[0017] The similarity between the horizontal coordinate difference information and the vertical coordinate difference information is used as the first similarity between the first position coordinate information and the second position coordinate information.

[0018] Furthermore, the first location coordinate information includes first longitude information and first latitude information, and the second location coordinate information includes second longitude information and second latitude information. The aforementioned target road image determination module is specifically used for:

[0019] Determine the first difference information between the first longitude information and the second longitude information, and use the first difference information as the abscissa difference information between the first position coordinate information and the second position coordinate information;

[0020] Determine the second difference information between the first latitude information and the second latitude information, and use the second difference information as the ordinate difference information between the first position coordinate information and the second position coordinate information.

[0021] Furthermore, the aforementioned target road image determination module is specifically used for:

[0022] The product of the first longitude information and the first latitude information and the product of the second longitude information and the second latitude information are summed to obtain the first sum value information;

[0023] The product of the first longitude information and the second latitude information, and the product of the second longitude information and the first latitude information are summed to obtain the second sum value information;

[0024] The target difference information is obtained based on the difference between the first sum information and the second sum information, and the first modulus information of the first difference information and the second modulus information of the second difference information are determined.

[0025] Divide the target difference information by the first modulus information and the second modulus information to obtain the first similarity between the first difference information and the second difference information.

[0026] Furthermore, the sample road image is acquired by the image acquisition device of the target vehicle, and the second acquisition module is specifically used for:

[0027] Obtain the vehicle identifier of the target vehicle;

[0028] Obtain multiple candidate road images associated with the vehicle identifier from the map database, and obtain the second location information of each candidate road image.

[0029] Furthermore, the aforementioned correction module is specifically used for:

[0030] The sample road image is input into a pre-trained image feature extraction model to extract image features from the sample road image, thereby obtaining the first image feature information of the sample road image;

[0031] The target road image is input into the image feature extraction model to extract image features from the target road image, thereby obtaining the second image feature information of the target road image;

[0032] A second similarity between the sample road image and the target road image is determined based on the first image feature information and the second image feature information.

[0033] Furthermore, the aforementioned correction module is specifically used for:

[0034] The sample road image is enhanced to obtain an enhanced road image, wherein the enhancement process includes at least one of magnification, reduction, rotation, flipping, and grayscale conversion.

[0035] The enhanced road image is input into a pre-trained image feature extraction model to extract image features from the enhanced road image, thereby obtaining the first image feature information of the enhanced road image.

[0036] Furthermore, the aforementioned correction module is specifically used for:

[0037] The first image feature information and the second image feature information are multiplied to obtain feature product information;

[0038] Determine the third modulus information of the first image feature information and the fourth modulus information of the second image feature information;

[0039] Dividing the feature product information by the third modulus information and the fourth modulus information yields the second similarity between the sample road image and the target road image.

[0040] Furthermore, the aforementioned correction module is specifically used for:

[0041] The target similarity is obtained by multiplying the first similarity and the second similarity.

[0042] The target road image is corrected based on the target similarity.

[0043] Furthermore, the aforementioned correction module is specifically used for:

[0044] When the target similarity is greater than or equal to a preset similarity threshold, the target road image is retained in the map database;

[0045] Alternatively, when the target similarity is less than the similarity threshold, the target road image in the map database is replaced with the sample road image.

[0046] Furthermore, the sample road image is acquired by the image acquisition device of the target vehicle, and the aforementioned map correction device is also used for:

[0047] During the navigation of the target vehicle, a corrected image of the target road is obtained based on the target vehicle's current navigation route;

[0048] The corrected target road image is subjected to road fault identification processing to obtain road fault identification results;

[0049] The navigation route is adjusted based on the road fault identification results.

[0050] On the other hand, embodiments of this application also provide an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the above-described map correction method.

[0051] On the other hand, embodiments of this application also provide a computer-readable storage medium storing a computer program that is executed by a processor to implement the map correction method described above.

[0052] On the other hand, embodiments of this application also provide a computer program product, which includes a computer program stored in a computer-readable storage medium. A processor of a computer device reads the computer program from the computer-readable storage medium and executes the computer program, causing the computer device to perform the map correction method described above.

[0053] The embodiments of this application include at least the following beneficial effects: by acquiring sample road images collected in real time, the target road image can be corrected in a timely manner, improving the efficiency of map correction; furthermore, by introducing a first similarity between the first location information and the second location information, the target road image that is more geographically matched to the sample road image can be quickly and accurately determined from multiple candidate road images; by introducing a second similarity between the current road image and the target road image to correct the target road image, the deviation between the map image in the map database and the actual situation can be accurately repaired, thereby improving the accuracy of the map in real time.

[0054] Other features and advantages of this application will be set forth in the following description and will be apparent in part from the description or may be learned by practicing the application. Attached Figure Description

[0055] The accompanying drawings are used to provide a further understanding of the technical solutions of this application and constitute a part of the specification. They are used together with the embodiments of this application to explain the technical solutions of this application and do not constitute a limitation on the technical solutions of this application.

[0056] Figure 1 A schematic diagram of an implementation environment provided for an embodiment of this application;

[0057] Figure 2 A flowchart illustrating a map correction method provided in this application embodiment;

[0058] Figure 3 This application provides a schematic diagram of a structure for collecting first position coordinate information in an embodiment of the present application.

[0059] Figure 4 This is a schematic diagram of another structure for collecting first position coordinate information provided in an embodiment of this application;

[0060] Figure 5 A schematic diagram illustrating the process of determining the second similarity provided in an embodiment of this application;

[0061] Figure 6 A detailed flowchart illustrating the map correction method provided in this application embodiment;

[0062] Figure 7 A flowchart illustrating another map correction method provided in this application embodiment;

[0063] Figure 8 This is a schematic diagram of an autonomous driving scenario for performing a map correction method, provided in an embodiment of this application.

[0064] Figure 9 This is a schematic diagram of the map correction device provided in the embodiments of this application;

[0065] Figure 10 This is a partial structural block diagram of a terminal provided in an embodiment of this application;

[0066] Figure 11 A partial structural block diagram of the server provided in an embodiment of this application. Detailed Implementation

[0067] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0068] It should be noted that in various specific embodiments of this application, when processing data related to the characteristics of the target object, such as target object attribute information or attribute information sets, is required, the permission or consent of the target object will be obtained first. Furthermore, the collection, use, and processing of this data will comply with the relevant laws, regulations, and standards of the relevant countries and regions. In addition, when embodiments of this application need to obtain target object attribute information, separate permission or consent from the target object will be obtained through pop-up windows or redirection to a confirmation page. Only after obtaining the separate permission or consent of the target object will the necessary target object-related data for the normal operation of the embodiments of this application be obtained.

[0069] To facilitate understanding of the technical solutions provided in the embodiments of this application, some key terms used in the embodiments of this application will be explained below:

[0070] Intelligent Traffic Systems (ITS), also known as Intelligent Transportation Systems, effectively integrate advanced science and technology (information technology, computer technology, data communication technology, sensor technology, electronic control technology, automatic control theory, operations research, artificial intelligence, etc.) into transportation, service control, and vehicle manufacturing. This strengthens the connection between vehicles, roads, and users, thereby forming a comprehensive transportation system that ensures safety, improves efficiency, enhances the environment, and saves energy.

[0071] Intelligent Vehicle Infrastructure Cooperative Systems (IVICS) are a development direction of Intelligent Transportation Systems (ITS). IVICS utilizes advanced wireless communication and next-generation Internet technologies to implement comprehensive, real-time dynamic information exchange between vehicles and infrastructure. Based on the collection and fusion of dynamic traffic information across all times and spaces, it conducts active vehicle safety control and cooperative road management, fully realizing effective collaboration between people, vehicles, and roads. This ensures traffic safety, improves traffic efficiency, and ultimately forms a safe, efficient, and environmentally friendly road traffic system.

[0072] A Convolutional Neural Network (CNN) is a type of feedforward neural network whose artificial neurons can respond to surrounding units within a certain coverage area, making it excellent for large-scale image processing. A CNN consists of one or more convolutional layers and a fully connected layer at the top (corresponding to a classic neural network), as well as associated weights and pooling layers.

[0073] In related technologies, when updating maps to ensure their accuracy, manual methods are generally used to update the maps. However, the efficiency and accuracy of this method need to be improved.

[0074] Based on this, embodiments of this application provide a map correction method and apparatus, electronic device, storage medium, and program product, which can improve the accuracy of maps.

[0075] Reference Figure 1 , Figure 1 This is a schematic diagram of an implementation environment provided in an embodiment of this application. The implementation environment includes a terminal 101 and a server 102, wherein the terminal 101 and the server 102 are connected through a communication network.

[0076] For example, terminal 101 can collect sample road images and first location information of the sample road images in real time, and send the sample road images and first location information to server 102. Server 102 obtains the sample road images and first location information from terminal 101, and then obtains second location information of each candidate road image in the map database. It then determines a first similarity between the first and second location information, and determines a target road image from the multiple candidate road images based on the first similarity. After determining the target road image, server 102 determines a second similarity between the sample road image and the target road image, corrects the target road image based on the second similarity, and then sends the corrected target road image to terminal 101.

[0077] In one possible implementation, the terminal can independently complete the acquisition of sample road images and the correction of target road images. Specifically, the terminal can acquire sample road images and their first location information in real time, and also obtain the second location information of each candidate road image in the map database. Then, it determines the first similarity between the first and second location information, determines the target road image from multiple candidate road images based on the first similarity, and finally determines the second similarity between the sample road image and the target road image, and corrects the target road image based on the second similarity.

[0078] Server 102 can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms. Additionally, server 102 can also be a node server in a blockchain network.

[0079] The aforementioned terminal 101 is equipped with an image acquisition device capable of capturing images. This terminal can be a smartphone, tablet computer, laptop computer, desktop computer, smart speaker, smartwatch, portable wearable device, vehicle terminal, etc., but is not limited to these. The terminal 101 and the server 102 can be directly or indirectly connected via wired or wireless communication, and this embodiment of the application does not impose any limitations.

[0080] The methods provided in this application can be applied to various technical fields, including but not limited to cloud technology, intelligent transportation, intelligent driving, maps, navigation and other technical fields.

[0081] The principle of the map correction method provided in the embodiments of this application is described in detail below.

[0082] Reference Figure 2 , Figure 2 The flowchart is a map correction method provided in the embodiments of this application. The map correction method can be executed by a server, or by a terminal, or by a combination of a server and a terminal. The map correction method includes, but is not limited to, the following steps 201 to 204.

[0083] Step 201: Obtain the real-time sample road image and the first location information of the sample road image.

[0084] The first location information can be the geographical name or location coordinates of the sample road image acquired in real time. When the first location information is the location coordinates of the sample road image acquired in real time, the first location information is the first location coordinate information. This application embodiment does not impose specific limitations on this.

[0085] In one possible implementation, the sample road image can be acquired in real time by an image acquisition device on the terminal. This image acquisition device can be a camera, such as a monocular camera, a binocular camera, a depth camera, or a 3D (3D) camera. Specifically, the camera is activated to scan the target object in its field of view in real time, and a sample road image is generated at a specified frame rate. Alternatively, the image acquisition device can be a radar device such as a lidar or millimeter-wave radar. Specifically, the radar device emits a detection signal towards the target object in real time, receives the echo signal reflected back from the target object, determines the characteristics of the target object based on the difference between the detection signal and the echo signal, and generates a sample road image based on these characteristics. Lidar can detect the position, attitude, and shape of the target object by emitting a laser beam, while millimeter-wave radar can detect the position, attitude, and shape of the target object in the millimeter-wave band. Furthermore, the target object can be traffic signs, road surfaces, etc., which are not specifically limited in this embodiment.

[0086] When the sample road image is acquired by the terminal's image acquisition device, the sample road image can be an image reflecting the current surrounding environment of the image acquisition device or the current surrounding environment of the terminal. For example, traffic sign images, road surface images, and other images can be accurately obtained from the sample road image. That is to say, the sample road image can include traffic sign images, road surface images, and other images. Among them, the traffic sign image can include the category of traffic signs, which are specifically represented by text or symbols and can be used to convey guidance, restriction, warning, or instruction information. The road surface image can include the location, posture, shape, and other features of target objects such as obstacles, road curves, and road speed bumps. It can also include the number of roads and the location, posture, shape, and other features between different roads. This application embodiment does not specifically limit this.

[0087] In one possible implementation, when the first location information is first location coordinate information, the first location coordinate information of the sample road image can be the geographical location of the image acquisition device at the current time of acquiring the sample road image; it can also be the geographical location of the target object in the sample road image at the current time of acquiring the sample road image; it can also be the geographical location of the terminal equipped with the image acquisition device at the current time of acquiring the sample road image; it can also be the geographical location range of the content contained in the sample road image, etc., and this application embodiment does not specifically limit it in this way.

[0088] In one possible implementation, when the first location information is the geographical name of the location of the sample road image acquired in real time, the first location information of the sample road image can be the geographical name of the location of the image acquisition device at the current moment of acquiring the sample road image; it can also be the geographical name of the target object in the sample road image at the current moment of acquiring the sample road image; it can also be the geographical name of the location of the terminal equipped with the image acquisition device at the current moment of acquiring the sample road image; it can also be the geographical location range of the content contained in the sample road image, etc., and this application embodiment does not specifically limit it in this way.

[0089] In one possible implementation, when the sample road image can be acquired by the image acquisition device of the target vehicle, the first position coordinate information can be obtained by the in-vehicle inertial navigation system (INS), or by a map application running on the in-vehicle terminal of the target vehicle, or by a map application running on a smart terminal (such as a smartphone) located on the target vehicle, but is not limited to these. When the target vehicle is an autonomous vehicle, the first position coordinate information can also be calculated by the autonomous vehicle's Global Navigation Satellite System (GNSS) data, Inertial Measurement Unit (IMU) data, and laser point cloud data. Alternatively, the first position coordinate information can also be obtained by locating the target vehicle using other positioning systems, such as the Global Positioning System (GPS) or the BeiDou Navigation Satellite System (BDS), etc.

[0090] For example, refer to Figure 3 Taking Differential Global Positioning System (GPS) data as an example, assuming the target vehicle is an autonomous vehicle, during the autonomous vehicle's operation, at the current moment of acquiring sample road images, GPS data is acquired in real time via the vehicle's onboard GPS system, and current point cloud data is acquired in real time via the autonomous vehicle's LiDAR sensor. Then, the GPS data and multiple current point cloud data can be offline registered using the Iterative Closest Points (ICP) algorithm to obtain the current geographical location of the autonomous vehicle, i.e., the first location coordinate information. Since the positioning accuracy of GPS data can reach the centimeter level, the accuracy of the geographical location determined by offline registration using the ICP algorithm, combining GPS data with multiple current point cloud data, can be better than centimeter-level. This can reduce the slight positional deviations generated during real-time acquisition of sample road images. This application does not specifically limit this aspect.

[0091] For example, refer to Figure 4Taking the first position information as the first position coordinate information, and the first position coordinate information is obtained by using an inertial navigation system to locate the autonomous vehicle, at the current moment of acquiring sample road images, the positioning sensor of the inertial navigation system can deduce the current positioning coordinates of the autonomous vehicle, i.e., the first position coordinate information, from the starting position of the autonomous vehicle based on the continuously measured heading angle and speed of the autonomous vehicle. The positioning sensor may include gyroscopes and accelerometers, etc., and this application embodiment does not specifically limit this.

[0092] In one possible implementation, when the first location information is the geographical name of the location of the sample road image acquired in real time, the first location information of the sample road image is a geographical name when the sample road image is taken as a single image unit; when the sample road image is divided into multiple image units, the first location information of the sample road image is the set of geographical names of all image units in the sample road image. This application does not specifically limit this aspect.

[0093] In one possible implementation, when the first location information is the first location coordinate information, when the sample road image is taken as one image unit, the first location coordinate information of the sample road image is a single location coordinate; when the sample road image is divided into multiple image units, the first location coordinate information of the sample road image is the set of location coordinates of all image units in the sample road image, thereby improving the accuracy of the location coordinates. This application does not specifically limit this aspect.

[0094] Step 202: Obtain the second location information of each candidate road image in the map database.

[0095] The second location information may be the geographical name, location coordinates, etc. of the candidate road image in the map database. When the second location information is the location coordinates of the candidate road image in the map database, the second location information is the second location coordinate information. This application embodiment does not impose specific limitations on this.

[0096] In one possible implementation, when the second location information is the geographical name of the location of the candidate road image in the map database, the second location information of the candidate road image is a geographical name when the candidate road image is taken as a single image unit; when the candidate road image is divided into multiple image units, the second location information of the candidate road image is the set of geographical names of all image units in the candidate road image. This application does not specifically limit this aspect.

[0097] In one possible implementation, the second location information is second location coordinate information, which is the geographical location corresponding to the candidate road image. Specifically, when the candidate road image is considered as a single image unit, the second location coordinate information of that candidate road image is a single location coordinate; when the candidate road image is divided into multiple image units, the second location coordinate information of that candidate road image is the set of location coordinates of all image units in the candidate road image. This application does not impose specific limitations on this aspect.

[0098] In one possible implementation, the map database can be located on a server, such as a cloud server or a regular server (i.e., a server), and the map database can be associated with the identification information of devices such as image acquisition devices or terminals configured with image acquisition devices, so that the server, image acquisition devices, or terminals configured with image acquisition devices can obtain the second location coordinate information of each candidate road image in the map database based on the identification information. The identification information can be vehicle identification (such as license plate number), mobile device identification code, etc., and the identification information that can uniquely identify the device can be selected according to the specific device. This application embodiment does not make specific limitations on this.

[0099] In one possible implementation, when the image acquisition device is set on the target vehicle, the sample road image can be acquired by the image acquisition device of the target vehicle. In this case, the first location coordinate information of the sample road image can be the current time of acquisition of the sample road image and the geographical location of the target vehicle. The image acquisition device can be a dashcam. Furthermore, the vehicle identifier of the target vehicle can be obtained, multiple candidate road images associated with the vehicle identifier can be obtained from the map database, and the second location information of each candidate road image can be obtained. This application embodiment does not specifically limit this aspect.

[0100] Step 203: Determine the first similarity between the first location information and the second location information, and determine the target road image from multiple candidate road images based on the first similarity.

[0101] In one possible implementation, when the first location information is the geographical name of the location of the sample road image acquired in real time, and the second location information is the geographical name of the location of the candidate road image in the map database, a first similarity between the first location information and the second location information can be determined, that is, whether the geographical name of the location of the sample road image acquired in real time is consistent with the geographical names of the locations of each candidate road image in the map database. If there is a candidate road image among the multiple candidate road images in the map database that makes the first similarity between the first location information and the second location information 1 (that is, it is determined that the geographical name of the location of the sample road image acquired in real time is consistent with the geographical name of the location of the candidate road image in the map database), then the candidate road image corresponding to the second location information can be determined as the target road image. This application embodiment does not specifically limit this.

[0102] In one possible implementation, when the first location information is first location coordinate information and the second location information is second location coordinate information, the difference in the horizontal coordinates between the first location coordinate information and the second location coordinate information, as well as the difference in the vertical coordinates between the first location coordinate information and the second location coordinate information, can be determined. Then, the similarity between the difference in the horizontal coordinates and the difference in the vertical coordinates is used as the first similarity between the first location coordinate information and the second location coordinate information. Therefore, by introducing the difference in the horizontal coordinates and the difference in the vertical coordinates to calculate the first similarity, the accuracy of the similarity calculation can be improved, the impact of subtle positional deviations generated when real-time acquisition of sample road images can be reduced, and the positional matching degree between the sample road image and the target road image can be effectively improved.

[0103] The horizontal coordinate difference information represents the difference between the horizontal coordinate of the sample road image and the horizontal coordinate of the candidate road image. When both the first and second location coordinate information are latitude and longitude, the horizontal coordinate difference information represents the difference between the longitude of the sample road image and the longitude of the candidate road image. Similarly, the vertical coordinate difference information represents the difference between the vertical coordinate of the sample road image and the vertical coordinate of the candidate road image. When both the first and second location coordinate information are latitude and longitude, the vertical coordinate difference information represents the difference between the latitude of the sample road image and the latitude of the candidate road image.

[0104] In one possible implementation, the first location coordinate information includes first longitude information and first latitude information, and the second location coordinate information includes second longitude information and second latitude information. The first longitude information is the longitude of the sample road image and the first latitude information is the latitude of the sample road image. Correspondingly, the second longitude information is the longitude of the candidate road image and the second latitude information is the latitude of the candidate road image. A first difference between the first longitude information and the second longitude information can be determined, and this first difference is used as the difference in the horizontal coordinates between the first and second position coordinates. Specifically, when both the first and second longitude information include one longitude, the first difference is a single longitude difference; when both include multiple longitudes, the first difference is a set of multiple longitude differences. Similarly, a second difference between the first latitude information and the second latitude information can be determined, and this second difference is used as the difference in the vertical coordinates between the first and second position coordinates. Specifically, when both the first and second latitude information include one latitude, the second difference is a single latitude difference; when both include multiple latitudes, the second difference is a set of multiple latitude differences. Furthermore, the first difference can be positive, zero, or negative, and the second difference can also be positive, zero, or negative. It is understood that the first difference between the first longitude information and the second longitude information can be obtained by subtracting the second longitude information from the first longitude information, or by subtracting the first longitude information from the second longitude information; similarly, the second difference between the first latitude information and the second latitude information can be obtained by subtracting the second latitude information from the first latitude information, or by subtracting the first latitude information from the second latitude information. This application embodiment does not specifically limit this.

[0105] For example, when both the candidate road image and the sample road image are considered as a single image unit, if the first position coordinate information of the sample road image is (116°E, 40°N), i.e., the first longitude information is 116°E and the first latitude information is 40°N, and the second position coordinate information of the candidate road image is (117°E, 50°N), i.e., the second longitude information is 117°E and the second latitude information is 50°N, then the difference in the horizontal coordinate between the first and second position coordinate information is (116°E-117°E), i.e., -1, or the difference in the horizontal coordinate is (117°E-116°E), i.e., 1; similarly, the difference in the vertical coordinate between the first and second position coordinate information is (40°N-50°N), i.e., -10, or the difference in the vertical coordinate is (50°N-40°N), i.e., 10. This application embodiment does not impose specific limitations on this.

[0106] For example, when both the candidate road image and the sample road image are segmented into multiple image units, and the number of image units corresponding to the candidate road image and the number of image units corresponding to the sample road image are both 3, if the position coordinates of all image units in the sample road image, i.e., the first position coordinate information, are (116°E, 40°N), (127°E, 41°N), and (138°E, 42°N), i.e., the first longitude information is 116°E, 127°E, and 138°E, and the first latitude information is 40°N, 41°N, and 42°N, and all image units in the candidate road image are also segmented into multiple image units, the following conditions apply: If the unit's position coordinates, i.e., the second position coordinate information, are (117°E, 50°N), (128°E, 51°N), and (139°E, 51°N), i.e., the second longitude information is 117°E, 128°E, and 139°E, and the second latitude information is 50°N, 51°N, and 51°N, then the difference in the horizontal coordinate information is either 1 or -1. Similarly, the difference in the vertical coordinate information between the first position coordinate information and the second position coordinate information is either -10 or 10. This application embodiment does not impose specific limitations on this.

[0107] In one possible implementation, the first difference information may include a first difference matrix, and the second difference information may include a second difference matrix. The product of the first longitude information and the first latitude information, and the product of the second longitude information and the second latitude information, can be summed to obtain a first sum information. Similarly, the product of the first longitude information and the second latitude information, and the product of the second longitude information and the first latitude information, can be summed to obtain a second sum information. Then, the target difference information is obtained based on the difference between the first and second sum information. Next, the first modulus information of the first difference matrix and the second modulus information of the second difference matrix are determined. Finally, the target difference information is divided by the first and second modulus information to obtain a first similarity between the first and second difference information.

[0108] In one possible implementation, assuming that both the sample road image and the candidate road image can be divided into multiple image units, each image unit corresponds to a location coordinate information, where an image unit can be a single pixel or a set of multiple pixels, then the first location coordinate information of the sample road image can be represented as follows: And n is a positive integer, where A represents the sample road image. This represents the longitude information corresponding to the i-th image unit in the sample road image. The set of longitude information for all image units in the sample road image is the first longitude information. The first dimension information represents the dimensional information corresponding to the i-th image unit in the sample road image; similarly, the second position coordinate information of the candidate road image can be represented as... And n is a positive integer, where B represents the candidate road image. This represents the longitude information corresponding to the i-th image unit in the candidate road image. The set of longitude information for all image units in the candidate road image is the second longitude information. The first difference matrix Δ represents the dimensional information corresponding to the i-th image unit in the candidate road image. The set of dimensional information of all image units in the candidate road image is the second dimensional information. Based on this, the first difference matrix Δ l It can be represented as:

[0109]

[0110] Where 1 ≤ i ≤ n, and n is a positive integer. It can be understood that Δ l It is the set of differences between the longitude information of all image units in the sample road image and the longitude information of all image units in the candidate road image.

[0111] Similarly, the second difference matrix Δ w It can be represented as:

[0112]

[0113] Where 1 ≤ i ≤ n, and n is a positive integer. It can be understood that Δ w It is the set of differences between the dimensional information of all image units in the sample road image and the dimensional information of all image units in the candidate road image.

[0114] Based on this, the first similarity can be expressed as:

[0115]

[0116] Wherein, cos sim(Δ l Δ w The first similarity score is represented by ), which is the cosine similarity between the sample road image and the candidate road image based on geographical location. This indicates the first sum value information. This indicates the second sum value information; Indicates target difference information. This indicates the first modulus value information. This represents the second modulus value information. It is understandable that... It can be used to eliminate dimensions, thereby achieving standardization.

[0117] Specifically, assuming that both the sample road image and the candidate road image are divided into 3 image units, i.e., n=3, the first position coordinate information of the sample road image can be represented as follows: The second location coordinate information of the candidate road image can be represented as: The first similarity is:

[0118]

[0119] If the coordinate position of the first image unit of the sample road image The coordinates of the second image unit in the sample road image are (22°E, 22°N). The coordinates of the third image unit in the sample road image are (33°E, 33°N). The coordinates of the first image unit in the candidate road image are (42°E, 44°N). The coordinates of the second image unit in the candidate road image are (21°E, 21°N). The coordinates of the third image unit in the candidate road image are (32°E, 34°N). Given (43°E, 43°N), the first difference matrix Δ l Given {1, 1, -1}, the second difference matrix Δ w Given the expression {1, -1, 1}, the first sum is 6799, the second sum is 6800, and the target difference is 1. The first modulus is obtained from the first difference matrix. The second modulus information is obtained from the second difference matrix. Finally, by dividing the target difference information by the first modulus information and the second modulus information, the first similarity between the sample road image and the candidate road image is obtained as 1 / 3. This application embodiment does not impose specific limitations on this.

[0120] In one possible implementation, assuming that both the sample road image and the candidate road image are treated as a single image unit, meaning that each image corresponds to only one location coordinate, the first location coordinate information of the sample road image can be represented as (l A ,w A ), where A represents the sample road image, l A w represents the first longitude information of the sample road image. A The first dimension information of the sample road image is represented by (l); similarly, the second position coordinate information of the candidate road image can be represented as (l B ,w B ), where B represents the candidate road image, l Bw represents the second longitude information of the candidate road image. B This represents the second-dimensional information of the candidate road image. Based on this, the first difference matrix Δ l It can be represented as:

[0121] Δ l =l A -l B

[0122] Where, Δ l The first difference matrix is ​​the difference between the first longitude information of the sample road image and the second longitude information of the candidate road image.

[0123] Similarly, the second difference matrix Δ w It can be represented as:

[0124] Δ w =w A -w B

[0125] Where, Δ w The difference between the first dimension information of the sample road image and the second dimension information of the candidate road image is called the second difference matrix.

[0126] Based on this, the first similarity can be expressed as:

[0127]

[0128] Wherein, cos sim(Δ l ,Δ w ) represents the first similarity, which is the cosine similarity between the sample road image and the candidate road image based on geographical location. A w A +l B w B ) represents the first sum information, (l A w B +l B w A (l) indicates the second sum value information; A w A +l B w B )-(l w w B +l B w A This indicates the target difference information. This indicates the first modulus value information. This represents the second modulus value information. It can be understood that (l A w A +l B w B )-(lA w B +l B w A It can be used to eliminate dimensions, thereby achieving standardization.

[0129] Specifically, assuming the first location coordinate information of the sample road image (l A ,w A The second location coordinate information of the candidate road image is (21°E, 21°N). B ,w B If the coordinates are (22°E, 22°N), then the first difference matrix Δ l The second difference matrix Δ is -1. w The first sum information is 925, the second sum information is 924, the target difference information is 1 obtained from the difference between the first sum information and the second sum information, the first modulus information is 1 obtained from the first difference matrix, the second modulus information is 1 obtained from the second difference matrix, and finally, by dividing the target difference information by the first modulus information and the second modulus information, the first similarity between the sample road image and the candidate road image is 1. This application embodiment does not impose specific limitations on this.

[0130] Since there are multiple candidate road images in the map database, multiple first similarities can be obtained based on the second location coordinates of each candidate road image in the map database and the first location coordinates of the sample road images. These first similarities are cosine similarity sequences based on geographic location, which can be represented as {c}. k |k=1,...,m},where c k com sim(Δ) represents the first similarity of the k-th element. l ,Δ w And 1≤k≤m, where m is a positive integer. Then, the first similarity score with the largest value can be selected from this cosine similarity sequence. Right now

[0131]

[0132] The first similarity The corresponding candidate road image is determined as the target road image, that is, the target road image that best matches the geographical location of the sample road image is determined from multiple candidate road images, where max represents the maximum value in the cosine similarity sequence.

[0133] Therefore, in this embodiment, the target road image that best matches the geographical location of the sample road image can be effectively determined from multiple candidate road images. Furthermore, by introducing horizontal and vertical coordinate difference information to calculate the first similarity, the accuracy of similarity calculation can be improved, and the impact of subtle positional deviations generated during real-time acquisition of sample road images can be reduced, thereby effectively improving the positional matching degree between the sample road image and the target road image.

[0134] It is understood that the above embodiments are not limited, and a first similarity value corresponding to the cosine similarity sequence can be selected according to actual needs. That is, a target road image that meets the requirements can be determined from multiple candidate road images according to actual needs, and the number of target road images is not specifically limited. For example, a preset matching threshold can be set. When the first similarity in the cosine similarity sequence is greater than or equal to the preset matching threshold, the candidate road image corresponding to the first similarity is determined as the target road image. The embodiments of this application do not impose specific limitations in this regard.

[0135] Step 204: Determine the second similarity between the sample road image and the target road image, and correct the target road image based on the second similarity.

[0136] In one possible implementation, there are many ways to determine the second similarity between a sample road image and a target road image. For example, the second similarity between the sample road image and the target road image can be determined by calculation methods such as cosine similarity, Euclidean distance, Pearson correlation coefficient, or Jaccard system. This application does not impose any specific limitations on this.

[0137] In one possible implementation, refer to Figure 5 The sample road image can be input into a pre-trained image feature extraction model to extract image features and obtain the first image feature information of the sample road image. Similarly, the target road image can be input into the image feature extraction model to extract image features and obtain the second image feature information of the target road image. Finally, the second similarity between the sample road image and the target road image is determined based on the first image feature information and the second image feature information. This application embodiment does not impose specific limitations on this.

[0138] The aforementioned image feature extraction model can be implemented based on artificial intelligence technology, which may include computer vision, machine learning, or deep learning. The specific form of the model structure can be various artificial neural networks, such as convolutional neural networks, and this application embodiment does not impose specific limitations on this.

[0139] In one possible implementation, the sample road image can first be enhanced to obtain an enhanced road image. This enhancement process includes at least one of magnification, reduction, rotation, flipping, and grayscale conversion. Then, the enhanced road image is input into a pre-trained image feature extraction model to extract image features, obtaining the first image feature information of the enhanced road image. It is understood that in this embodiment, by enhancing the sample road image, the features in the sample road image can be enhanced, thereby improving the image feature extraction effect.

[0140] In one possible implementation, the target road image can first be enhanced to obtain an enhanced target road image. The enhancement process includes at least one of magnification, reduction, rotation, flipping, and grayscale conversion. Then, the enhanced target road image is input into a pre-trained image feature extraction model to extract image features, obtaining second image feature information of the enhanced target road image. This embodiment does not impose specific limitations on this approach. It is understood that in this embodiment, by enhancing the target road image, the features in the target road image can be enhanced, thereby improving the image feature extraction effect.

[0141] In one possible implementation, the first image feature information may include a first feature matrix, and the second image feature information may include a second feature matrix. First, the first image feature information and the second image feature information can be multiplied to obtain feature product information. Then, the third modulus information of the first feature matrix and the fourth modulus information of the second feature matrix are determined. Finally, the feature product information is divided by the third modulus information and the fourth modulus information to obtain the second similarity between the sample road image and the target road image. The first feature matrix may include either a single image feature from the sample road image or a set of all image features in the sample road image; similarly, the second feature matrix may include either a single image feature from the target road image or a set of all image features in the target road image. This application does not impose specific limitations on this aspect.

[0142] In one possible implementation, assume that the first feature matrix of the sample road image is The second feature matrix of the target road image is And n is a positive integer, where, This represents the i-th image feature of the sample road image. Let i represent the i-th image feature of the candidate road image. Then, the second similarity can be expressed as:

[0143]

[0144] in, Indicates the second similarity. Represents feature product information. This represents the third modulus information of the first characteristic matrix. This represents the fourth modulus information of the second feature matrix, where A represents the sample road image and B represents the candidate road image. This represents the i-th image feature extracted from the sample road image. This represents the i-th image feature extracted from the target road image. This refers to sample road images that are within the same longitude and latitude range as the target road image. This refers to target road images in a map database that are within the same longitude and latitude range as the sample road images. It is understandable that, from... and It can be seen that the longitude and latitude ranges mentioned above refer to the longitude and latitude positions of the sample road image and the target road image.

[0145] In one possible implementation, the first similarity and the second similarity can be multiplied to obtain the target similarity, and then the target road image can be corrected based on the target similarity, thereby effectively improving the accuracy of image-based similarity calculation.

[0146] Based on this, target similarity can be expressed as:

[0147]

[0148] Wherein, cos sim(A,B|(Δ l ,Δ w )) represents the target similarity, cos sim(Δ l ,Δ w () indicates the first similarity. This indicates the second similarity.

[0149] Moreover, cos sim(Δ l ,Δ w The similarity can be expressed in two forms. When both the sample road image and the target road image are divided into multiple image units, the first similarity is cos sim(Δ). l ,Δ w Specifically, it can be expressed as:

[0150]

[0151] When both the sample road image and the candidate road image are treated as a single image unit, the first similarity is cos sim(Δ). l ,Δ w Specifically, it can be expressed as:

[0152]

[0153] Among them, the first similarity is cos sim(Δ l Δ w The principle behind this can be found in the previous explanation, and will not be repeated here.

[0154] In one possible implementation, there are many ways to correct the target road image based on the target similarity. For example, when the target similarity is greater than or equal to a preset similarity threshold, the target road image is retained in the map database; or, when the target similarity is less than the similarity threshold, the target road image in the map database is replaced with a sample road image. The preset similarity threshold can be set to 0.6 or other values ​​greater than 0.6, such as 0.6, 0.61, 0.65, 0.7, 0.8, or 0.9, etc., which will not be listed here.

[0155] Understandably, when the target similarity is greater than or equal to a preset similarity threshold, it indicates that the deviation between the real-time collected sample road image and the target road image in the map database is not significant. Therefore, no correction is needed for the target road image; it is simply retained in the map database. Similarly, when the target similarity is less than the similarity threshold, it indicates that there is a significant deviation between the real-time collected sample road image and the target road image in the map database. Therefore, the target road image needs to be modified, i.e., the target road image in the map database is replaced with the sample road image. Thus, this embodiment of the application corrects the target road image through target similarity, which can accurately correct the deviation between the map image in the map database and the actual situation, thereby improving the accuracy of the map in real time.

[0156] The following details the process of the map correction method provided in the embodiments of this application.

[0157] Reference Figure 6 , Figure 6 The following is a detailed flowchart illustrating the map correction method provided in this application embodiment. First, a real-time sample road image is acquired, along with first location coordinate information of the sample road image. The first location coordinate information includes first longitude information and first latitude information. If the sample road image comprises n image units, then the first longitude information is... The first dimension information is The first position coordinate information can be in, This represents the longitude information corresponding to the i-th image unit in the sample road image. This represents the latitude information corresponding to the i-th image unit in the sample road image;

[0158] Similarly, the second location coordinate information of each candidate road image in the map database is obtained. This second location coordinate information also includes n image units, and includes second longitude information and second latitude information. The second longitude information is then... The second latitude information is The second position coordinate information is: in, This represents the longitude information corresponding to the i-th image unit in the candidate road image. This represents the latitude information corresponding to the i-th image unit in the candidate road image.

[0159] Then, the first similarity is determined between the difference in the horizontal coordinates between the first and second position coordinates and the difference in the vertical coordinates between the first and second position coordinates.

[0160] First, determine the first difference information (i.e., the horizontal coordinate difference information) between the first longitude information and the second longitude information, namely the first difference matrix Δ. l :

[0161]

[0162] Similarly, the second difference information (i.e., the ordinate difference information) between the first position coordinate information and the second position coordinate information is determined, namely the second difference matrix Δ. w :

[0163]

[0164] Determine the first modulus information of the first difference matrix and the second modulus information of the second difference matrix;

[0165] Next, the product of the first longitude information and the first latitude information and the product of the second longitude information and the second latitude information are summed to obtain the first sum value information; the product of the first longitude information and the second latitude information and the product of the second longitude information and the first latitude information are summed to obtain the second sum value information; the target difference value information is obtained based on the difference between the first sum value information and the second sum value information.

[0166] Finally, the target difference information is divided by the first modulus information and the second modulus information to obtain the first similarity between the first difference information and the second difference information, i.e.

[0167]

[0168] Suppose there are 5 candidate road images in the map database, and the first similarity between the 5 candidate road images and the sample road image is C = {c1, c2, c3, c4, c5}. Then, the first similarity with the largest value is selected from these 5 first similarity values. Right now

[0169]

[0170] like Then the candidate road image corresponding to the first similarity c5 is determined as the target road image.

[0171] Next, the second similarity between the target road image and the sample road image is determined. First, the sample road image is enhanced by at least one of the following processes: magnification, reduction, rotation, flipping, and grayscale conversion, resulting in an enhanced road image. Then, the enhanced road image is input into a pre-trained image feature extraction model to extract image features, obtaining the first image feature information of the enhanced road image, i.e., the first feature matrix. in, Let represent the i-th image feature of the sample road image, where n is a positive integer and represents the total number of image features of the sample road image;

[0172] Similarly, the target road image is input into the image feature extraction model to extract image features, resulting in the second image feature information of the target road image, i.e., the second feature matrix. n is a positive integer, and n represents the total number of image features of the target road image, where, This represents the i-th image feature of the candidate road image;

[0173] Then, the feature information of the first image and the feature information of the second image are multiplied to obtain the feature product information. The third modulus of the first feature matrix and the fourth modulus of the second feature matrix are then determined. Finally, the feature product information is divided by the third and fourth modulus information to determine the second similarity between the sample road image and the target road image. The second similarity can be:

[0174]

[0175] Next, the first similarity and the second similarity are multiplied to obtain the target similarity, i.e.

[0176]

[0177] Finally, the target road map is corrected. If the preset similarity threshold is 0.6, assuming a second similarity... First similarity cos sim(Δ l ,Δ w If c5 = 0.7, then the target similarity is:

[0178] cos sim(A,B|(Δ l ,Δ w =0.7 × 0.8 = 0.56

[0179] Therefore, since the target similarity is less than the similarity threshold, it is necessary to replace the target road images in the map database with sample road images; or, assume a second similarity... If the first similarity c5 = 0.8, then the target similarity is 0.72. Since the target similarity is greater than the similarity threshold, the target road image is retained in the map database. This application embodiment does not impose specific limitations on this.

[0180] In steps 201 to 204, by acquiring sample road images collected in real time, the target road image can be corrected in a timely manner, improving the efficiency of map correction. Furthermore, by introducing the first similarity between the first location information and the second location information, the target road image that is more geographically matched to the sample road image can be quickly and accurately determined from multiple candidate road images. By introducing the second similarity between the current road image and the target road image to correct the target road image, the deviation between the map image in the map database and the actual situation can be accurately repaired, thereby improving the accuracy of the map in real time.

[0181] In one possible implementation, refer to Figure 7Assuming the sample road image is a traffic sign image, firstly, in the traffic sign image acquisition stage, the traffic sign can be captured in real time using the camera device of the dashcam on the target vehicle, obtaining a traffic sign image, which is then synchronized to the cloud backend system. The first position coordinate information of the traffic sign image is then determined. Secondly, in the real-time image processing stage, the traffic sign image can be enhanced using at least one of the following methods: magnification, reduction, rotation, flipping, and grayscale conversion, resulting in an enhanced road image. This enhanced road image is then input into a pre-trained image feature extraction module to extract image features, obtaining the first image feature information of the enhanced road image. Next, in the map database image processing stage, candidate road images from the map database can be input into an image feature extraction model to extract image features, obtaining the second image feature information of the candidate road images. Finally, in the location-based cosine similarity model construction stage, the second position coordinate information of each candidate road image in the map database can be determined first, and a location-based cosine similarity model can be constructed using the first and second position coordinate information of the traffic sign image. A cosine similarity model is used to determine the location, and a first similarity is calculated based on this model. This involves determining the difference in the horizontal coordinates between the first and second location coordinates, as well as the difference in the vertical coordinates between them, and then determining the first similarity between these two differences. In the geographical location matching stage between traffic sign images and candidate road images, the target road image with the highest geographical relevance to the real-time acquired traffic sign image can be determined from multiple candidate road images based on the first similarity. Then, in the image cosine similarity model... In the initial phase, a second similarity (i.e., image cosine similarity) can be calculated using the first image feature information of the traffic sign image and the second image feature information of the target road image. Next, in the phase of determining the paired cosine similarity, the first similarity and the image cosine similarity can be multiplied to obtain the paired cosine similarity (i.e., target similarity). Finally, in the real-time image editing phase, if the preset similarity threshold is 0.6, the target road image is retained in the map database if the target similarity is greater than or equal to 0.6; otherwise, the target road image in the map database is replaced with a traffic sign image. This cloud-based backend system can be an intelligent vehicle-road cooperative system or other systems; this application embodiment does not impose specific limitations on it.

[0182] Exemplary, the map correction method provided in this application can be applied to autonomous driving scenarios, vehicle navigation scenarios, map data collection scenarios, and road data collection scenarios, etc., and this application does not limit the scope. In the above application scenarios, it is usually necessary to collect sample road images and the first position coordinate information of the sample road images, and then perform subsequent operations based on the sample road images and their first position coordinate information. Specifically, the target vehicle can take pictures and capture images according to the current road conditions to obtain sample road images. Then, based on the position coordinate information, the first similarity between the sample road image and candidate road images in the map database is determined. Based on the first similarity, the target road image is determined from multiple candidate road images. Next, an image feature extraction model pre-trained using convolutional neural networks and other algorithms is used to extract image features from the target road image and the sample road image to obtain the first image feature information of the sample road image and the second image feature information of the target road image. Then, the correlation between the first image feature information and the second image feature information is calculated to determine the second similarity between the sample road image and the target road image. Finally, the target road image in the map database with low correlation to the real-time collected sample road image is replaced with the real-time collected sample road image. The sample road image can be an image reflecting the environment around the vehicle.

[0183] For example, refer to Figure 8 Assuming the sample road images are acquired by the image acquisition device of an autonomous vehicle, in an autonomous driving scenario, the image acquisition device (such as a dashcam camera) acquires the sample road images and their first position coordinates in real time. By analyzing the sample road images and their first position coordinates, the categories and position coordinates of traffic signs are obtained. Based on this information, the target road images in the map database are corrected. Then, the autonomous vehicle can obtain the corrected target road images according to its current navigation route. Road fault identification processing is performed on the corrected target road images to obtain road fault identification results. The navigation route is adjusted according to these results, ensuring the autonomous vehicle's safety. Therefore, this embodiment of the application, through real-time acquisition and processing of traffic signs, can effectively improve the traffic signal processing capabilities of autonomous vehicles during driving.

[0184] For example, assuming the sample road image is acquired by the target vehicle's image acquisition device, during navigation, the device continuously collects the sample road image and its initial coordinates. By analyzing this information, the types and coordinates of traffic signs, road curves, speed bumps, and other features are identified. Based on this information, the target road image in the map database is corrected. The autonomous vehicle can then obtain the corrected target road image based on its current navigation route. This corrected image is then used to identify road faults, and the navigation route is adjusted accordingly. The navigation route is output in voice, text, and image formats to guide the target vehicle. Therefore, even if road faults or road repairs occur during navigation, the map can be updated promptly to avoid navigation errors, thus ensuring user safety and significantly improving the user experience.

[0185] For example, assuming the sample road image is acquired by the image acquisition device of the target vehicle, in the map data acquisition scenario, the image acquisition device of the target vehicle acquires the sample road image and its first position coordinate information in real time. By analyzing the sample road image and its first position coordinate information, the category and position coordinate information of traffic signs, the position coordinate information of road curves, the position coordinate information of road speed bumps, the number of lanes and the position coordinate information of lanes, etc., are obtained. Based on this information, the target road image in the map database is corrected. The map in the updated map database can be used for vehicle navigation, travel route planning, etc.

[0186] The target vehicle can be of various types and designs, such as cars (e.g., smart cars and autonomous vehicles), motorcycles, trucks, buses, semi-tractors, etc., and this embodiment does not limit them.

[0187] It is understood that although the steps in the above flowcharts are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated in this embodiment, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the above flowcharts may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps.

[0188] Reference Figure 9 , Figure 9 This is a schematic diagram of the structure of the map correction device provided in the embodiments of this application. The map correction device 900 includes:

[0189] The first acquisition module 901 is used to acquire real-time sample road images and the first position coordinate information of the sample road images;

[0190] The second acquisition module 902 is used to acquire the second location coordinate information of each candidate road image in the map database;

[0191] The target road image determination module 903 is used to determine a first similarity between first location information and second location information, and to determine a target road image from multiple candidate road images based on the first similarity.

[0192] The correction module 904 is used to determine the second similarity between the sample road image and the target road image, and to correct the target road image based on the second similarity.

[0193] Furthermore, the first location information is the first location coordinate information, and the second location information is the second location coordinate information. The aforementioned target road image determination module 903 is specifically used for:

[0194] Determine the difference in the horizontal coordinates between the first position coordinates and the second position coordinates, and the difference in the vertical coordinates between the first position coordinates and the second position coordinates;

[0195] The similarity between the difference in the horizontal coordinate and the difference in the vertical coordinate is used as the first similarity between the first position coordinate information and the second position coordinate information.

[0196] Furthermore, the first location coordinate information includes first longitude and first latitude information, and the second location coordinate information includes second longitude and second latitude information. The aforementioned target road image determination module 903 is specifically used for:

[0197] Determine the first difference information between the first longitude information and the second longitude information, and use the first difference information as the abscissa difference information between the first position coordinate information and the second position coordinate information;

[0198] Determine the second difference information between the first latitude information and the second latitude information, and use the second difference as the ordinate difference information between the first position coordinate information and the second position coordinate information.

[0199] Furthermore, the aforementioned target road image determination module 903 is specifically used for:

[0200] The first sum value is obtained by summing the product between the first longitude information and the first latitude information, and the product between the second longitude information and the second latitude information.

[0201] The product of the first longitude information and the second latitude information, and the product of the second longitude information and the first latitude information are summed to obtain the second sum value information;

[0202] The target difference information is obtained based on the difference between the first sum information and the second sum information, and the first modulus information of the first difference information and the second modulus information of the second difference information are determined.

[0203] Divide the target difference information by the first modulus information and the second modulus information to obtain the first similarity between the first difference information and the second difference information.

[0204] Furthermore, the sample road image is acquired by the image acquisition device of the target vehicle, and the second acquisition module 902 mentioned above is specifically used for:

[0205] Obtain the vehicle identification number of the target vehicle;

[0206] Obtain multiple candidate road images associated with vehicle identifiers from the map database, and obtain the second location information of each candidate road image.

[0207] Furthermore, the aforementioned correction module 904 is specifically used for:

[0208] The sample road image is input into a pre-trained image feature extraction model to extract image features from the sample road image, thereby obtaining the first image feature information of the sample road image;

[0209] The target road image is input into the image feature extraction model to extract image features and obtain the second image feature information of the target road image;

[0210] The second similarity between the sample road image and the target road image is determined based on the first image feature information and the second image feature information.

[0211] Furthermore, the aforementioned correction module 904 is specifically used for:

[0212] Enhanced road images are obtained by performing enhancement processing on sample road images. The enhancement processing includes at least one of magnification, reduction, rotation, flipping, and grayscale conversion.

[0213] The enhanced sample road image is input into a pre-trained image feature extraction model to extract image features and obtain the first image feature information of the enhanced sample road image.

[0214] Furthermore, the aforementioned correction module 904 is specifically used for:

[0215] The feature information of the first image and the feature information of the second image are multiplied to obtain the feature product information;

[0216] Determine the third modulus value of the first image feature information and the fourth modulus value of the second image feature information;

[0217] Dividing the feature product information by the third and fourth modulus information yields the second similarity between the sample road image and the target road image.

[0218] Furthermore, the aforementioned correction module 904 is specifically used for:

[0219] The target similarity is obtained by multiplying the first similarity and the second similarity.

[0220] The target road image is corrected based on the similarity of the target.

[0221] Furthermore, the aforementioned correction module 904 is specifically used for:

[0222] When the target similarity is greater than or equal to a preset similarity threshold, the target road image is retained in the map database;

[0223] Alternatively, when the target similarity is less than a similarity threshold, the target road image in the map database is replaced with the sample road image.

[0224] Furthermore, the sample road images are acquired by the image acquisition device of the target vehicle, and the aforementioned map correction device 900 is also used for:

[0225] During the navigation process for the target vehicle, a corrected image of the target road is obtained based on the target vehicle's current navigation route;

[0226] The corrected target road image is processed to identify road faults, and the road fault identification results are obtained.

[0227] Adjust the navigation route based on the road fault identification results.

[0228] The aforementioned map correction device 900 and map correction method are based on the same inventive concept. By acquiring real-time sample road images, they can promptly correct target road images, improving the efficiency of map correction. Furthermore, by introducing the difference in horizontal coordinates between the first and second location coordinates, and the difference in vertical coordinates between the first and second location coordinates, they can effectively determine the target road image that best matches the geographical location of the sample road image from multiple candidate road images based on the first similarity between the horizontal and vertical coordinate differences. Introducing the difference in horizontal and vertical coordinates to calculate the first similarity improves the accuracy of similarity calculation and reduces the impact of subtle positional deviations during real-time acquisition of sample road images, thereby effectively improving the positional matching degree between the sample road image and the target road image. By introducing the second similarity between the current road image and the target road image to correct the target road image, the deviation between the map image in the map database and the actual situation can be accurately corrected, thereby improving the accuracy of the map.

[0229] The electronic device provided in this application embodiment for performing the above-described map correction method can be a terminal, as shown below. Figure 10 , Figure 10 This is a partial structural block diagram of a terminal provided in an embodiment of this application. The terminal includes: a radio frequency (RF) circuit 1010, a memory 1020, an input unit 1030, a display unit 1040, a sensor 1050, an audio circuit 1060, a wireless fidelity (WiFi) module 1070, a processor 1080, and a power supply 1090, among other components. Those skilled in the art will understand that... Figure 10 The terminal structure shown does not constitute a limitation on the terminal and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0230] The RF circuit 1010 can be used to receive and transmit signals during information transmission or calls. In particular, it receives downlink information from the base station and processes it with the processor 1080; in addition, it transmits uplink data to the base station.

[0231] The memory 1020 can be used to store software programs and modules. The processor 1080 executes various functional applications and data processing of the terminal by running the software programs and modules stored in the memory 1020.

[0232] The input unit 1030 can be used to receive input numeric or character information, and to generate key signal inputs related to the terminal's settings and function control. Specifically, the input unit 1030 may include a touch panel 1031 and other input devices 1032.

[0233] The display unit 1040 can be used to display input or provided information, as well as various menus of the terminal. The display unit 1040 may include a display panel 1041.

[0234] Audio circuitry 1060, speaker 1061, and microphone 1062 provide an audio interface.

[0235] In this embodiment, the processor 1080 included in the terminal can execute the map correction method of the previous embodiment.

[0236] The electronic device provided in this application embodiment for performing the above-described map correction method can also be a server, see reference. Figure 11 , Figure 11 The diagram illustrates a partial structural block of a server provided in this application embodiment. The server 1100 can vary significantly due to different configurations or performance characteristics. It may include one or more Central Processing Units (CPUs) 1122 (e.g., one or more processors) and a memory 1132, and one or more storage media 1130 (e.g., one or more mass storage devices) for storing application programs 1142 or data 1144. The memory 1132 and storage media 1130 may be temporary or persistent storage. The program stored in the storage media 1130 may include one or more modules (not shown in the diagram), each module including a series of instruction operations on the server 1100. Furthermore, the CPU 1122 may be configured to communicate with the storage media 1130 and execute the series of instruction operations in the storage media 1130 on the server 1100.

[0237] Server 1100 may also include one or more power supplies 1126, one or more wired or wireless network interfaces 1150, one or more input / output interfaces 1158, and / or one or more operating systems 1141, such as Windows Server™, Mac OS X™, Unix™, Linux™, FreeBSD™, etc.

[0238] The processor in server 1100 can be used to execute map correction methods.

[0239] This application also provides a computer-readable storage medium for storing program code for executing the map correction methods of the foregoing embodiments.

[0240] This application also provides a computer program product, which includes a computer program stored in a computer-readable storage medium. A processor of a computer device reads the computer program from the computer-readable storage medium and executes the computer program, causing the computer device to perform the map correction method described above.

[0241] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented, for example, in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatuses.

[0242] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0243] It should be understood that in the description of the embodiments of this application, "multiple" means two or more, "greater than", "less than", "exceeding" etc. are understood to exclude the number itself, and "above", "below", "within" etc. are understood to include the number itself.

[0244] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between apparatuses or units, and may be electrical, mechanical, or other forms.

[0245] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0246] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0247] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0248] It should also be understood that the various implementation methods provided in this application can be combined arbitrarily to achieve different technical effects.

[0249] The above provides a detailed description of the preferred embodiments of this application. However, this application is not limited to the above-described embodiments. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of this application. All such equivalent modifications or substitutions are included within the scope defined by the claims of this application.

Claims

1. A map correction method, characterized in that, include: Acquire real-time sample road images and the first location information of the sample road images, the first location information including first longitude information and first latitude information; Obtain the second location information of each candidate road image in the map database, the second location information including second longitude information and second latitude information; A first difference information is determined between the first longitude information and the second longitude information, and a second difference information is determined between the first latitude information and the second latitude information. The first difference information includes a first difference matrix, and the second difference information includes a second difference matrix. The product between the first longitude information and the first latitude information, and the product between the second longitude information and the second latitude information are summed to obtain a first sum information. The product between the first longitude information and the second latitude information, and the product between the second longitude information and the first latitude information are summed to obtain a second sum information. The target difference information is obtained based on the difference between the first sum information and the second sum information. The first modulus information of the first difference matrix and the second modulus information of the second difference matrix are determined. Divide the target difference information by the first modulus information and the second modulus information to obtain the first similarity between the first difference information and the second difference information, and determine the target road image from the multiple candidate road images based on the first similarity. A second similarity is determined between the sample road image and the target road image, and the target road image is corrected based on the second similarity.

2. The map correction method according to claim 1, characterized in that, The sample road images are acquired by the image acquisition device of the target vehicle. The acquisition of the second location information of each candidate road image in the map database includes: Obtain the vehicle identifier of the target vehicle; Obtain multiple candidate road images associated with the vehicle identifier from the map database, and obtain the second location information of each candidate road image.

3. The map correction method according to claim 1, characterized in that, Determining the second similarity between the sample road image and the target road image includes: The sample road image is input into a pre-trained image feature extraction model to extract image features from the sample road image, thereby obtaining the first image feature information of the sample road image; The target road image is input into the image feature extraction model to extract image features from the target road image, thereby obtaining the second image feature information of the target road image; A second similarity between the sample road image and the target road image is determined based on the first image feature information and the second image feature information.

4. The map correction method according to claim 3, characterized in that, The step of inputting the sample road image into a pre-trained image feature extraction model to extract image features from the sample road image and obtain the first image feature information of the sample road image includes: The sample road image is enhanced to obtain an enhanced road image, wherein the enhancement process includes at least one of magnification, reduction, rotation, flipping, and grayscale conversion. The enhanced road image is input into a pre-trained image feature extraction model to extract image features from the enhanced road image, thereby obtaining the first image feature information of the enhanced road image.

5. The map correction method according to claim 3, characterized in that, The first image feature information includes a first feature matrix, and the second image feature information includes a second feature matrix. Determining the second similarity between the sample road image and the target road image based on the first image feature information and the second image feature information includes: The first image feature information and the second image feature information are multiplied to obtain feature product information; Determine the third modulus information of the first feature matrix and the fourth modulus information of the second feature matrix; Dividing the feature product information by the third modulus information and the fourth modulus information yields the second similarity between the sample road image and the target road image.

6. The map correction method according to claim 1, characterized in that, The step of correcting the target road image based on the second similarity includes: The target similarity is obtained by multiplying the first similarity and the second similarity. The target road image is corrected based on the target similarity.

7. The map correction method according to claim 6, characterized in that, The step of correcting the target road image based on the target similarity includes: When the target similarity is greater than or equal to a preset similarity threshold, the target road image is retained in the map database; Alternatively, when the target similarity is less than the similarity threshold, the target road image in the map database is replaced with the sample road image.

8. The map correction method according to claim 1, characterized in that, The sample road image is acquired by the image acquisition device of the target vehicle, and the map correction method further includes: During the navigation of the target vehicle, a corrected image of the target road is obtained based on the target vehicle's current navigation route; The corrected target road image is subjected to road fault identification processing to obtain road fault identification results; The navigation route is adjusted based on the road fault identification results.

9. A map correction device, characterized in that, include: The first acquisition module is used to acquire real-time sample road images and the first location information of the sample road images, wherein the first location information includes first longitude information and first latitude information; The second acquisition module is used to acquire the second location information of each candidate road image in the map database, the second location information including second longitude information and second latitude information; The target road image determination module is used to determine a first difference information between the first longitude information and the second longitude information, and to determine a second difference information between the first latitude information and the second latitude information. The first difference information includes a first difference matrix, and the second difference information includes a second difference matrix. The module sums the product of the first longitude information and the first latitude information, and the product of the second longitude information and the second latitude information, to obtain a first sum information. It also sums the product of the first longitude information and the second latitude information, and the product of the second longitude information and the first latitude information, to obtain a second sum information. Based on the difference between the first sum information and the second sum information, the module obtains target difference information and determines a first modulus information of the first difference matrix and a second modulus information of the second difference matrix. Divide the target difference information by the first modulus information and the second modulus information to obtain the first similarity between the first difference information and the second difference information, and determine the target road image from the multiple candidate road images based on the first similarity. The correction module is used to determine a second similarity between the sample road image and the target road image, and to correct the target road image based on the second similarity.

10. The map correction device according to claim 9, characterized in that, The sample road image is acquired by the image acquisition device of the target vehicle, and the second acquisition module is specifically used for: Obtain the vehicle identifier of the target vehicle; Obtain multiple candidate road images associated with the vehicle identifier from the map database, and obtain the second location information of each candidate road image.

11. The map correction device according to claim 9, characterized in that, The correction module is specifically used for: The sample road image is input into a pre-trained image feature extraction model to extract image features from the sample road image, thereby obtaining the first image feature information of the sample road image; The target road image is input into the image feature extraction model to extract image features from the target road image, thereby obtaining the second image feature information of the target road image; A second similarity between the sample road image and the target road image is determined based on the first image feature information and the second image feature information.

12. The map correction device according to claim 11, characterized in that, The correction module is specifically used for: The sample road image is enhanced to obtain an enhanced road image, wherein the enhancement process includes at least one of magnification, reduction, rotation, flipping, and grayscale conversion. The enhanced road image is input into a pre-trained image feature extraction model to extract image features from the enhanced road image, thereby obtaining the first image feature information of the enhanced road image.

13. The map correction device according to claim 11, characterized in that, The correction module is specifically used for: The first image feature information and the second image feature information are multiplied to obtain feature product information; Determine the third modulus information of the first image feature information and the fourth modulus information of the second image feature information; Dividing the feature product information by the third modulus information and the fourth modulus information yields the second similarity between the sample road image and the target road image.

14. The map correction device according to claim 9, characterized in that, The correction module is specifically used for: The target similarity is obtained by multiplying the first similarity and the second similarity. The target road image is corrected based on the target similarity.

15. The map correction device according to claim 14, characterized in that, The correction module is specifically used for: When the target similarity is greater than or equal to a preset similarity threshold, the target road image is retained in the map database; Alternatively, when the target similarity is less than the similarity threshold, the target road image in the map database is replaced with the sample road image.

16. The map correction device according to claim 9, characterized in that, The sample road image is acquired by the image acquisition device of the target vehicle, and the map correction device is also used for: During the navigation of the target vehicle, a corrected image of the target road is obtained based on the target vehicle's current navigation route; The corrected target road image is subjected to road fault identification processing to obtain road fault identification results; The navigation route is adjusted based on the road fault identification results.

17. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the map correction method according to any one of claims 1 to 8.

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

19. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the map correction method according to any one of claims 1 to 8.

Images