Image processing methods and apparatuses, electronic devices, and computer-readable storage media

Abstract

This application discloses an image processing method and apparatus, electronic device, and computer-readable storage medium, including: acquiring three-dimensional coordinate information and two-dimensional coordinate information of a target scene; converting each three-dimensional coordinate included in the three-dimensional coordinate information into first polar coordinates, and converting each first two-dimensional coordinate included in the two-dimensional coordinate information into second polar coordinates, wherein the reference coordinate systems corresponding to the first polar coordinates and the second polar coordinates are the same; adding identification information to the real-world object corresponding to the target first polar coordinate, wherein the identification information includes category information corresponding to the target second polar coordinate, the category information is used to describe the category to which the real-world object corresponding to the target second polar coordinate belongs, the polar angle of the target second polar coordinate matches the polar angle of the target first polar coordinate, and the target first polar coordinate is the first polar coordinate corresponding to the real-world object in the first polar coordinate system. Implementing this application embodiment can improve the calibration quality of the calibration information for each real-world object in a real-world scene.

Description

Technical Field

[0001] This application relates to the field of image processing technology, specifically to an image processing method and apparatus, electronic device, and computer-readable storage medium. Background Technology

[0002] With the rapid development of image processing technology, reversing cameras or autonomous driving systems now typically require the acquisition of image information about the vehicle's surroundings. Then, through corresponding image processing technology, real-world objects in the images are identified. The vehicle's controller can then control the vehicle's movement based on the identified real-world objects to avoid collisions.

[0003] However, in practice, it has been found that traditional image processing techniques usually use fusion algorithms to fuse point cloud information and image information of the acquired real scene, and then calibrate the real objects in the image based on the fusion result. However, since point cloud information and image information are difficult to synchronize during the fusion process, a large amount of original information will be lost, resulting in poor subsequent calibration quality. Summary of the Invention

[0004] This application discloses an image processing method and apparatus, an electronic device, and a computer-readable storage medium, which can improve the calibration quality of calibration information for various real-world objects in real-world scenes.

[0005] The first aspect of this application discloses an image processing method, the method comprising:

[0006] Obtain the three-dimensional coordinate information of the target scene, as well as the two-dimensional coordinate information of the target scene;

[0007] The three-dimensional coordinates included in the three-dimensional coordinate information are converted into first polar coordinates, and the first two-dimensional coordinates included in the two-dimensional coordinate information are converted into second polar coordinates. The reference coordinate system corresponding to the first polar coordinates and the reference coordinate system corresponding to the second polar coordinates are the same.

[0008] Add identification information to the real object corresponding to the first polar coordinate of the target. The identification information includes category information corresponding to the second polar coordinate of the target. The category information is used to describe the category to which the real object corresponding to the second polar coordinate of the target belongs. The polar angle of the second polar coordinate of the target matches the polar angle of the first polar coordinate of the target. The first polar coordinate of the target is the first polar coordinate of the first polar coordinate that corresponds to the real object. The second polar coordinate of the target is the second polar coordinate that corresponds to the category information.

[0009] As an optional implementation, in the first aspect of the embodiments of this application, the three-dimensional coordinate information is three-dimensional point cloud information, and the step of converting each three-dimensional coordinate included in the three-dimensional coordinate information into first polar coordinates includes:

[0010] The three-dimensional coordinates included in the three-dimensional coordinate information are projected into a rectangular coordinate system to obtain the second two-dimensional coordinates corresponding to each of the three-dimensional coordinates in the rectangular coordinate system;

[0011] Convert each of the second two-dimensional coordinates into the corresponding first polar coordinates.

[0012] As an optional implementation, in the first aspect of the embodiments of this application, the two-dimensional coordinate information is two-dimensional image information, and converting each of the first two-dimensional coordinates included in the two-dimensional coordinate information into second polar coordinates includes:

[0013] The two-dimensional coordinate information is converted into a bird's-eye view to obtain the third two-dimensional coordinates corresponding to each of the first two-dimensional coordinates included in the two-dimensional coordinate information in the bird's-eye view.

[0014] Convert each of the third two-dimensional coordinates into the corresponding second polar coordinates.

[0015] As an optional implementation, in the first aspect of the embodiments of this application, the step of converting the two-dimensional coordinate information into a bird's-eye view includes:

[0016] Based on the first two-dimensional coordinates, scale factor, and inverse perspective transformation matrix included in the two-dimensional coordinate information, the third two-dimensional coordinates corresponding to each of the first two-dimensional coordinates in the bird's-eye view are determined to obtain the bird's-eye view corresponding to the two-dimensional coordinate information.

[0017] As an optional implementation, in the first aspect of the embodiments of this application, before adding identification information to the object corresponding to the target first polar coordinates, the method further includes:

[0018] The category information corresponding to each real object in the bird's-eye view is determined, and the second polar coordinate corresponding to each real object is determined. The category information is used to describe the category to which the real object corresponding to the second polar coordinate belongs.

[0019] As an optional implementation, in the first aspect of the embodiments of this application, adding identification information to the object corresponding to the target first polar coordinates includes:

[0020] If there exists a target second polar coordinate that matches the target first polar coordinate, then the category information corresponding to the target second polar coordinate is used as the identification information of the real object corresponding to the target first polar coordinate;

[0021] If there is no target second polar coordinate that matches the target first polar coordinate, then the real object corresponding to the target first polar coordinate is determined to be an unknown object.

[0022] As an optional implementation method, in the first aspect of the embodiments of this application,

[0023] The three-dimensional coordinate information includes three-dimensional point cloud information of the target scene obtained by a three-dimensional lidar, and the two-dimensional coordinate information includes two-dimensional image information of the target scene obtained by a two-dimensional camera device. The information acquisition end of the three-dimensional lidar and the information acquisition end of the two-dimensional camera device are oriented in the same direction.

[0024] Alternatively, the three-dimensional coordinate information may include three-dimensional image information of the target scene acquired by a three-dimensional camera device, and the two-dimensional coordinate information may include two-dimensional point cloud information of the target scene acquired by a two-dimensional lidar, wherein the information acquisition end of the two-dimensional lidar is oriented in the same direction as the information acquisition end of the three-dimensional camera device.

[0025] A second aspect of this application discloses an image processing apparatus, the apparatus comprising:

[0026] The acquisition unit is used to acquire the three-dimensional coordinate information of the target scene and the two-dimensional coordinate information of the target scene;

[0027] The conversion unit is used to convert each of the three-dimensional coordinates included in the three-dimensional coordinate information into first polar coordinates, and to convert each of the first two-dimensional coordinates included in the two-dimensional coordinate information into second polar coordinates, wherein the reference coordinate system corresponding to the first polar coordinates and the reference coordinate system corresponding to the second polar coordinates are the same.

[0028] An adding unit is used to add identification information to the real object corresponding to the first polar coordinate of the target. The identification information includes category information corresponding to the second polar coordinate of the target. The category information is used to describe the category to which the real object corresponding to the second polar coordinate of the target belongs. The polar angle of the second polar coordinate of the target matches the polar angle of the first polar coordinate of the target. The first polar coordinate of the target is the first polar coordinate of the first polar coordinate that corresponds to the real object. The second polar coordinate of the target is the second polar coordinate that corresponds to the category information.

[0029] The third aspect of this application discloses an electronic device, including:

[0030] Memory containing executable program code;

[0031] A processor coupled to the memory;

[0032] The processor calls the executable program code stored in the memory to execute the image processing method disclosed in the first aspect of the embodiments of this application.

[0033] The fourth aspect of this application discloses a computer-readable storage medium storing a computer program, wherein the computer program causes a computer to execute the image processing method disclosed in the first aspect of this application.

[0034] The fifth aspect of this application discloses a computer program product that, when run on a computer, causes the computer to perform some or all of the steps of any one of the methods of the first aspect of this application.

[0035] The sixth aspect of this application discloses an application publishing platform for publishing computer program products, wherein when the computer program products are run on a computer, the computer performs some or all of the steps of any one of the methods of the first aspect of this application.

[0036] Compared with related technologies, the embodiments of this application have the following beneficial effects:

[0037] In this embodiment, three-dimensional coordinate information and two-dimensional coordinate information of the same target scene can be obtained. Then, the three-dimensional coordinates included in the three-dimensional coordinate information can be converted into first polar coordinates, and the first two-dimensional coordinates included in the two-dimensional coordinate information can be converted into second polar coordinates. Further, identification information can be added to the object corresponding to the target first polar coordinate, which corresponds to a real-world object. This identification information can include category information corresponding to the target second polar coordinate, whose polar angle matches the target first polar coordinate. This category information describes the category to which the real-world object corresponding to the target second polar coordinate belongs. The reference coordinate systems corresponding to the converted first and second polar coordinates are the same. Therefore, when adding identification information to the object corresponding to the target first polar coordinate, the three-dimensional and two-dimensional coordinate information can be fused in the same polar coordinate system, solving the problem of synchronization difficulties between three-dimensional and two-dimensional coordinate information. This avoids the loss of original information during the fusion process, thus allowing for more accurate determination of the position, category, and other calibration information of each real-world object in the real-world scene, improving the calibration quality of the calibration information for each object in the real-world scene. Attached Figure Description

[0038] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0039] Figure 1 This is a schematic flowchart of an image processing method disclosed in an embodiment of this application;

[0040] Figure 2A This is a schematic diagram of the setup of a data acquisition device disclosed in an embodiment of this application;

[0041] Figure 2B This is a schematic diagram of a two-dimensional coordinate to polar coordinate transformation disclosed in an embodiment of this application;

[0042] Figure 3 This is a schematic flowchart of another image processing method disclosed in an embodiment of this application;

[0043] Figure 4A This is a schematic diagram of a bird's-eye view of a two-dimensional coordinate information conversion disclosed in an embodiment of this application;

[0044] Figure 4B This is a schematic diagram of a semantic segmentation result disclosed in an embodiment of this application;

[0045] Figure 5 This is a schematic flowchart of another image processing method disclosed in the embodiments of this application;

[0046] Figure 6 This is a schematic diagram of the structure of an image processing apparatus disclosed in an embodiment of this application;

[0047] Figure 7 This is a schematic diagram of the structure of an electronic device disclosed in an embodiment of this application. Detailed Implementation

[0048] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0049] It should be noted that the terms "first," "second," "third," and "fourth," etc., used in the specification and claims of this application are used to distinguish different objects, not to describe a specific order. The terms "comprising" and "having," and any variations thereof, in the embodiments of this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes 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 these processes, methods, products, or devices.

[0050] This application discloses an image processing method and apparatus, an electronic device, and a computer-readable storage medium, which can improve the calibration quality of calibration information for various real-world objects in real-world scenes.

[0051] The technical solution of this application will be described in detail below with reference to specific embodiments.

[0052] To more clearly illustrate the image processing method, apparatus, electronic device, and computer-readable storage medium disclosed in the embodiments of this application, the application scenarios applicable to this method are first introduced. Optionally, this method can be applied to fields such as autonomous driving and reversing cameras, and is not limited thereto.

[0053] In related technologies, after the data acquisition device of an unmanned vehicle collects 3D and 2D coordinate information in front of the vehicle, the vehicle's processor directly performs fusion processing on the 3D and 2D coordinate information and calibrates the real-world objects in the image based on the fusion result. However, in practice, it has been found that because the 3D coordinate information (e.g., 3D point cloud information) and 2D coordinate information (e.g., 2D image information) collected by the data acquisition device correspond to different reference coordinate systems, it is difficult to synchronize the 3D and 2D coordinate information, resulting in the loss of a large amount of original information and consequently poor calibration quality.

[0054] Using the method provided in this application, after the data acquisition device of the unmanned vehicle acquires the three-dimensional coordinate information and two-dimensional coordinate information in front of the unmanned vehicle, the processor of the unmanned vehicle can first transform the natural coordinates included in the three-dimensional coordinate information and the two-dimensional coordinate information to the same polar coordinate system, and then fuse the three-dimensional coordinate information and the two-dimensional coordinate information in the same polar coordinate system. This solves the problem that multiple information with different reference coordinate systems are difficult to synchronize, and avoids the loss of original information during the fusion process. It can more accurately determine the position, category and other calibration information of each real object in the real scene. Further optionally, the processor can feed back the calibration information to the controller of the unmanned vehicle, so that the controller of the unmanned vehicle can control the unmanned vehicle to avoid obstacles and drive.

[0055] Based on this, the image processing method disclosed in the embodiments of this application will be described below.

[0056] Please see Figure 1 , Figure 1 This is a schematic flowchart illustrating an image processing method disclosed in an embodiment of this application. The method can be applied to a processor, such as an image processor. This embodiment uses a processor as an example for illustration and should not be construed as limiting the scope of this application. Optionally, the method may include the following steps:

[0057] 102. Obtain the three-dimensional coordinate information and the two-dimensional coordinate information of the target scene.

[0058] In this embodiment, the target scene may include various real-world scenarios, such as factories, warehouses, or highways, and is not limited thereto. The processor can acquire the three-dimensional coordinate information corresponding to the target scene, as well as the two-dimensional coordinate information of the same target scene, through a data acquisition device.

[0059] Optionally, the reference coordinate system corresponding to the three-dimensional coordinate information may be different from the reference coordinate system corresponding to the two-dimensional coordinate information; in other optional embodiments, the reference coordinate system corresponding to the three-dimensional coordinate information may be the same as the reference coordinate system corresponding to the two-dimensional coordinate information, which is not limited here.

[0060] As an optional implementation, the data acquisition device may include a lidar and a camera device; optionally, the lidar may include a three-dimensional lidar for acquiring three-dimensional point cloud information of the target scene, and the camera device may include a two-dimensional camera device for acquiring two-dimensional image information of the target scene.

[0061] LiDAR, or laser radar, is a device that emits laser light into a target scene and then images it based on the reflected light. LiDAR can collect three-dimensional point cloud information of the target scene. Three-dimensional point cloud information refers to a massive collection of point data representing the surface features of a target. In this embodiment, the three-dimensional point cloud information can refer to a collection of point data representing the surface features of various real-world objects in the target scene. The camera device can include optical cameras such as monocular cameras and binocular cameras, and is not limited thereto. Furthermore, the camera device can collect optical two-dimensional image information of the target scene.

[0062] In another alternative embodiment, the lidar may include a two-dimensional lidar for acquiring two-dimensional point cloud information of the target scene, and the camera device may include a three-dimensional camera device for acquiring three-dimensional image information of the target scene.

[0063] Please see Figure 2A , Figure 2A This is a schematic diagram of the setup of a data acquisition device disclosed in an embodiment of this application. Optionally, the information acquisition end 210 of the lidar 200 and the information acquisition end 230 of the camera device 220 can be oriented in the same direction, so that the same target scene can be acquired by the lidar 200 and the camera device 220 respectively.

[0064] It should be noted that, Figure 2A The illustrations are for illustrative purposes only and do not limit the relative positions of the lidar and the camera device. In some alternative embodiments, the lidar 200 and the camera device 220 may also be arranged side by side, which is not limited here.

[0065] Optionally, the 3D and 2D coordinate information of the target scene can be acquired simultaneously by the data acquisition device. This avoids inconsistencies between the 3D and 2D coordinate information, which could affect the subsequent calibration quality. For example, if the lidar acquires the 3D coordinate information of the target scene at the first moment, and a truck is present in the scene at that moment, the 3D coordinate information includes the truck. If the camera acquires the 2D coordinate information of the target scene at the second moment, but the truck has already left the scene, the 2D coordinate information does not include the truck. This inconsistency between the 3D and 2D coordinate information makes it difficult to determine whether a truck is present in the target scene during subsequent calibration, thus affecting the calibration quality.

[0066] By implementing the above method, the target scene in the three-dimensional coordinate information and the two-dimensional coordinate information can be kept consistent, thereby improving the subsequent calibration quality.

[0067] In some alternative implementations, the processor may also obtain the three-dimensional coordinate information of the target scene and the two-dimensional coordinate information of the target scene through other devices (e.g., other vehicles, user equipment, etc.), which are not limited here.

[0068] By implementing the above method, even if the data acquisition equipment malfunctions and cannot normally acquire 3D and 2D coordinate information, it is still possible to obtain 3D and 2D coordinate information through other devices, thereby improving the compatibility of the method.

[0069] 104. Convert each three-dimensional coordinate included in the three-dimensional coordinate information into first polar coordinates, and convert each first two-dimensional coordinate included in the two-dimensional coordinate information into second polar coordinates.

[0070] In this embodiment of the application, the processor can first convert each of the three-dimensional coordinates included in the three-dimensional coordinate information into a second two-dimensional coordinate, and then convert each of the second two-dimensional coordinates into a first polar coordinate.

[0071] In this system, the first polar coordinate is the coordinate point in the polar coordinate system. The polar coordinate system is formed by taking a fixed point O in the plane, called the pole, and then drawing a ray Ox, called the polar axis, and then choosing a unit of length and a positive direction for the angle. For any point M in the plane, the length of the line segment OM can be represented by r, called the polar radius; θ can be represented by the angle from Ox to OM, called the polar angle.

[0072] Please see Figure 2B , Figure 2B This is a schematic diagram of a two-dimensional coordinate to polar coordinate conversion disclosed in an embodiment of this application. The two-dimensional coordinates of the real object 'a' in the rectangular coordinate system are (x...).a y a The polar coordinates of the real object 'a' in the polar coordinate system are (r...). a θ a ).

[0073] Similarly, the processor can also convert the various first two-dimensional coordinates included in the two-dimensional coordinate information into second polar coordinates.

[0074] Optionally, the reference coordinate system corresponding to the first polar coordinate can be the same as the reference coordinate system corresponding to the second polar coordinate, which facilitates the subsequent fusion of three-dimensional coordinate information and two-dimensional coordinate information based on the first polar coordinate and the second polar coordinate, and avoids the loss of original information during the fusion process.

[0075] 106. Add identification information to the real-world object corresponding to the first polar coordinate of the target.

[0076] It is understood that the target scene includes various real-world objects, such as vehicles, pedestrians, or unknown obstacles, without limitation. Each of the various real-world objects in the target scene corresponds to a first polar coordinate in the polar coordinate system, and the position of each real-world object can be determined using these first polar coordinates.

[0077] Furthermore, the processor can use image recognition technology to determine the categories of various real-world objects included in the two-dimensional coordinate information, and determine the second polar coordinates corresponding to each real-world object, so as to associate the real-world object corresponding to each second polar coordinate with the category information corresponding to each real-world object.

[0078] Furthermore, the processor can match the target first polar coordinate and the target second polar coordinate that match the polar angle. If the target second polar coordinate has corresponding category information for describing the real object, then the category information corresponding to the target second polar coordinate is used as the identification information of the real object corresponding to the target first polar coordinate. If the target second polar coordinate that matches the target first polar coordinate does not have category information, then the real object corresponding to the target first polar coordinate is determined to be an unknown object.

[0079] Through the information fusion process described above, the location and category information of each real-world object in the target scene can be determined, thereby improving the calibration quality of the calibration information for each object in the real-world scene.

[0080] By implementing the methods disclosed in the above embodiments, three-dimensional coordinate information and two-dimensional coordinate information of the same target scene can be obtained. Then, the three-dimensional coordinates included in the three-dimensional coordinate information can be converted into first polar coordinates, and the first two-dimensional coordinates included in the two-dimensional coordinate information can be converted into second polar coordinates. Furthermore, identification information can be added to the object corresponding to the target first polar coordinate, which corresponds to a real-world object. This identification information may include category information corresponding to the target second polar coordinate, whose polar angle matches the target first polar coordinate. This category information describes the category to which the real-world object corresponding to the target second polar coordinate belongs. Therefore, by implementing the embodiments of this application, the coordinates included in the three-dimensional and two-dimensional coordinate information can be converted to a polar coordinate system. This allows for the fusion of the three-dimensional and two-dimensional coordinate information in the polar coordinate system, avoiding the loss of original information during the fusion process. Consequently, the location, category, and other calibration information of each real-world object in the real-world scene can be determined more accurately, improving the calibration quality of the calibration information for each object in the real-world scene.

[0081] Please see Figure 3 , Figure 3 This is a schematic flowchart of another image processing method disclosed in an embodiment of this application. This method can be applied to a processor, such as an image processor. This embodiment uses a processor as an example for illustration and should not be construed as limiting the scope of this application. Optionally, the method may include the following steps:

[0082] 302. Obtain the three-dimensional coordinate information and the two-dimensional coordinate information of the target scene.

[0083] 304. Convert each three-dimensional coordinate included in the three-dimensional coordinate information into first polar coordinates, and convert each first two-dimensional coordinate included in the two-dimensional coordinate information into second polar coordinates.

[0084] As an optional implementation, the processor can project each of the three-dimensional coordinates included in the three-dimensional coordinate information into a Cartesian coordinate system to obtain the second two-dimensional coordinates corresponding to each three-dimensional coordinate in the Cartesian coordinate system. Optionally, the Cartesian coordinate system can be a horizontal Cartesian coordinate system parallel to the horizontal ground, and the three-dimensional coordinate information is three-dimensional point cloud information.

[0085] Furthermore, the processor can determine the first polar coordinates corresponding to each second two-dimensional coordinate based on each second two-dimensional coordinate and the following formula 1, that is:

[0086]

[0087] θ1 = arctan(y2 / x2)

[0088] Where r1 represents the polar radius of the first polar coordinate system, θ1 represents the polar angle of the first polar coordinate system, x2 represents the abscissa of the second two-dimensional coordinate system, and y2 represents the ordinate of the second two-dimensional coordinate system.

[0089] In practice, it has been found that converting three-dimensional coordinates to polar coordinates is quite difficult. To address this, the above method can be implemented by first projecting the three-dimensional coordinates into a rectangular coordinate system to obtain the corresponding second two-dimensional coordinates of each three-dimensional coordinate in the rectangular coordinate system. Then, the second two-dimensional coordinates can be directly converted into first polar coordinates using the above formula 1, thereby reducing the difficulty of implementing the method.

[0090] As an optional implementation, the processor can first convert the acquired two-dimensional coordinate information into a bird's-eye view to obtain the corresponding third two-dimensional coordinates of each first two-dimensional coordinate in the bird's-eye view; then the processor can convert each third two-dimensional coordinate into the corresponding second polar coordinates, wherein the two-dimensional coordinate information is two-dimensional image information.

[0091] A bird's-eye view is a three-dimensional drawing created by looking down at the terrain from a high point using the principles of perspective and the high-viewpoint perspective method. In other words, it's an image seen from the air looking down at a particular area.

[0092] Please see Figure 4A , Figure 4A This is a schematic diagram of a two-dimensional coordinate information conversion into a bird's-eye view disclosed in an embodiment of this application. Since the shooting direction of the camera device is usually horizontal, the acquired two-dimensional coordinate information 410 is typically a front view of the target scene. However, it is difficult to determine the position of each real-world object in the target scene using only the front view, and consequently, it is difficult to determine the polar coordinates of each real-world object. Therefore, the processor can convert the two-dimensional coordinate information 410 into a bird's-eye view 420. Since the position of each real-world object in the bird's-eye view 420 is easily determined, it is convenient for the processor to determine the second polar coordinates and category of each real-world object included in the two-dimensional coordinate information.

[0093] It should be noted that, Figure 4A The examples provided are merely for illustrative purposes and should not be construed as limiting the scope of the embodiments described in this application.

[0094] Optionally, the processor may convert each third two-dimensional coordinate into its corresponding second polar coordinate in the same or similar way as the method described above for converting each second two-dimensional coordinate into its corresponding first polar coordinate, and will not be elaborated further here.

[0095] By implementing the above method, the processor can first convert the two-dimensional coordinate information into a bird's-eye view. Since the polar coordinates and categories of each real object in the bird's-eye view are easier to determine, it is more convenient to subsequently determine the polar coordinates and categories of each real object in the bird's-eye view.

[0096] As an optional implementation, the processor can convert the two-dimensional coordinate information into a bird's-eye view based on the principle of inverse perspective. Optionally, the processor can determine the corresponding third two-dimensional coordinates in the bird's-eye view based on the first two-dimensional coordinates, scale factor, and inverse perspective transformation matrix included in the two-dimensional coordinate information, so as to obtain the bird's-eye view corresponding to the two-dimensional coordinate information.

[0097] Optionally, the processor can determine the corresponding third two-dimensional coordinates in the bird's-eye view based on the first two-dimensional coordinates included in the two-dimensional coordinate information and the following formula 2, so as to obtain the bird's-eye view corresponding to the two-dimensional coordinate information, i.e.:

[0098]

[0099] Where (x3 / w′, y3 / w′) represents the third two-dimensional coordinates, (x1, y1) represents the first two-dimensional coordinates, and w and w′ represent the scale factors. This represents the inverse perspective transformation matrix.

[0100] Optional,

[0101] It represents a linear transformation of an image;

[0102] T2 = [a 13 a 23 It is used to generate image perspective transformation;

[0103] T3 = [a 31 a 32 ] indicates image translation.

[0104] By implementing the above method, the processor can convert two-dimensional coordinate information into a bird's-eye view through the principle of inverse perspective transformation. This conversion method is relatively easy to implement, thereby reducing the difficulty and cost of implementation.

[0105] 306. Determine the category information of each real-world object in the bird's-eye view, and determine the second polar coordinates of each real-world object.

[0106] In this embodiment of the application, the processor can perform deep learning-based semantic segmentation processing on the bird's-eye view to determine the category information corresponding to each real-world object in the bird's-eye view.

[0107] Optionally, the processor can train a classification model using machine learning, and then use the classification model to determine the category information corresponding to each real-world object in the bird's-eye view. This category information is used to describe the category to which each real-world object in the bird's-eye view belongs, such as: vehicle, pedestrian, or unknown obstacle, etc., without limitation. The classification module can be trained based on a large number of bird's-eye view samples, without limitation.

[0108] Please see Figure 4B , Figure 4B This is a schematic diagram of a semantic segmentation result disclosed in an embodiment of this application. The processor first determines a bird's-eye view D based on two-dimensional coordinate information A, B, and C. Then, the classification model can label each real-world object in the bird's-eye view D with different color labels, each corresponding to a different category. For example, gray and white in the bird's-eye view D can represent different category information.

[0109] It needs to be further explained that, Figure 4B The illustrations are for illustrative purposes only and should not be construed as limiting the embodiments of this application.

[0110] As an optional implementation, the processor can be mounted on an unmanned vehicle, including an unmanned forklift used for transporting goods. Optionally, if a target object in the bird's-eye view cannot be identified by category information, a communication connection is established with the unmanned vehicle's control system. The control system then obtains the target task assigned to the unmanned vehicle within the current time period, determines the unmanned vehicle's current working environment and work content based on the target task, and further determines the category information corresponding to the target object based on the working environment and work content.

[0111] Optionally, if the unmanned vehicle's working environment is a furniture warehouse and its task is to transport furniture, the processor can determine whether the target object is furniture based on the furniture warehouse and the transported furniture. Optionally, the processor can identify the two-dimensional coordinate information of the target object and then match it with furniture to determine whether the target object is furniture. If it is furniture, the processor can further determine the type of furniture it belongs to.

[0112] By implementing the above method, the processor can further determine the category information of the target real object to which the category information could not be identified based on the working environment and work content of the unmanned vehicle, thereby improving the flexibility of the method.

[0113] Furthermore, the processor can determine the second polar coordinates corresponding to each real-world object whose category information has been determined; this facilitates subsequent information fusion based on the polar coordinates.

[0114] 308. Add identification information to the real-world object corresponding to the first polar coordinate of the target.

[0115] The methods disclosed in the above embodiments can transform the coordinates included in the three-dimensional coordinate information and the two-dimensional coordinate information into a polar coordinate system. This allows for the fusion of the three-dimensional and two-dimensional coordinate information in the polar coordinate system, enabling more accurate determination of the location, category, and other calibration information of each real-world object in the real-world scene. This improves the calibration quality of the calibration information for each object in the real-world scene. Furthermore, the methods can first transform the three-dimensional coordinates included in the three-dimensional coordinate information into two-dimensional coordinates, facilitating the transformation of these coordinates into a polar coordinate system to obtain the polar coordinates corresponding to each three-dimensional coordinate, thus improving the flexibility of the method. Additionally, the two-dimensional coordinate information can be first converted into a bird's-eye view. Since the polar coordinates and categories of each real-world object are easier to determine in the bird's-eye view, it facilitates the subsequent determination of the polar coordinates and categories of each real-world object in the bird's-eye view. Finally, the conversion of two-dimensional coordinate information into a bird's-eye view can be achieved through the principle of inverse perspective transformation. This conversion method is relatively easy to implement, thereby reducing the implementation difficulty and cost of the method.

[0116] Please see Figure 5 , Figure 5 This is a schematic flowchart illustrating another image processing method disclosed in this application. This method can be applied to a processor, such as an image processor. This application uses a processor as an example for illustration and should not be construed as limiting the scope of this application. Optionally, the method may include the following steps:

[0117] 502. Obtain the three-dimensional coordinate information of the target scene through LiDAR, and obtain the two-dimensional coordinate information of the target scene through a camera device.

[0118] As an optional implementation, the processor can calibrate the intrinsic parameters of the camera device using an intrinsic parameter calibration method to optimize the shooting effect of the camera device. Optionally, the intrinsic parameter calibration method may include a checkerboard calibration method, etc., and is not limited thereto.

[0119] As another optional implementation, after the processor acquires the two-dimensional coordinate information of the target scene through the camera device, it can perform distortion correction processing on the two-dimensional coordinate information to facilitate subsequent information fusion based on the two-dimensional coordinate information.

[0120] 504. Convert each three-dimensional coordinate included in the three-dimensional coordinate information into first polar coordinates, and convert each first two-dimensional coordinate included in the two-dimensional coordinate information into second polar coordinates.

[0121] 506. Add identification information to the real-world object corresponding to the first polar coordinate of the target.

[0122] As an optional implementation, the processor can form a first sequence of multiple target first polar coordinates that correspond to real objects in the first polar coordinates, such as (target first polar coordinate a, target first polar coordinate b... target first polar coordinate n); and form a second sequence of multiple target second polar coordinates that correspond to category information in the second polar coordinates, such as (target second polar coordinate a, target second polar coordinate b... target second polar coordinate n).

[0123] The processor can then match the polar coordinates in the first and second sequences. If there is a target second polar coordinate in the second sequence that matches the target first polar coordinate, the category information corresponding to the target second polar coordinate is used as the identification information of the real object corresponding to the target first polar coordinate.

[0124] If there is no target second polar coordinate in the second sequence that matches the target first polar coordinate, then the real object corresponding to the target first polar coordinate is determined to be an unknown object.

[0125] Optionally, the processor can match the polar coordinates in the first and second sequences according to the principle of minimizing angular deviation. That is, if the difference between the polar angles of the first and second target polar coordinates is less than or equal to an angle threshold (which can be set by the developers based on extensive development experience), then the first and second target polar coordinates are determined to be a match; if the difference is greater than the angle threshold, then the first and second target polar coordinates are determined to be a mismatch. Optionally, if the polar angles of the first and second target polar coordinates are equal, then the first and second target polar coordinates are determined to be a match; if the polar angles of the first and second target polar coordinates are not equal, then the first and second target polar coordinates are determined to be a mismatch, without further limitation.

[0126] Understandably, the processor may not be able to fully identify the category information of all real-world objects in the bird's-eye view. Therefore, the second sequence may only include the second polar coordinates corresponding to some real-world objects whose categories can be identified. However, the first sequence includes the first polar coordinates of all real-world objects in the bird's-eye view, so the number of target first polar coordinates in the first sequence may be greater than or equal to the number of second polar coordinates included in the second sequence. Therefore, if a target first polar coordinate can be matched with a target second polar coordinate, it means that the processor can identify the category information of the real-world object corresponding to that target first polar coordinate; conversely, if a target first polar coordinate cannot be matched with a target second polar coordinate, it means that the processor cannot identify the category information of the real-world object corresponding to that target first polar coordinate.

[0127] As an optional implementation, the processor can acquire the target two-dimensional coordinate information of the real object corresponding to the target first polar coordinate that cannot be matched with the target second polar coordinate, and send the target two-dimensional coordinate information to the cloud server. The cloud server can then identify the target outline of the real object included in the target two-dimensional coordinate information; and then match the target outline with a standard outline template to determine the target standard outline template that matches the target outline, and feed back the category information corresponding to the target standard outline template to the processor. The processor then uses the category information corresponding to the target standard outline template as the identification information of the real object corresponding to the target first polar coordinate that cannot be matched with the target second polar coordinate.

[0128] For example, if the target outline of the real object corresponding to the first polar coordinate of the target cannot be matched with the second polar coordinate, but the category information corresponding to the target standard outline template is electric vehicle, then the electric vehicle can be used as the identification information of the real object corresponding to the first polar coordinate of the target.

[0129] By implementing the above method, the processor can also send the target two-dimensional coordinate information corresponding to a real object in the target scene to the cloud server when it is unable to determine the category information of a real object, so that the cloud server can determine the category information corresponding to the real object, thereby improving the flexibility of the method.

[0130] As an optional implementation, the processor can also acquire thermal infrared images of the target scene through a data acquisition device, and then determine the temperature of each real object in the target scene based on the thermal infrared images; further, if there is no target second polar coordinate in the second sequence that matches the target first polar coordinate, then the target temperature information of the real object corresponding to the target first polar coordinate is determined based on the thermal infrared image; if the target temperature information of the real object is greater than or equal to a first temperature threshold (this temperature threshold can be set by developers based on extensive development experience, or by users based on usage requirements, and is not limited here), then the real object corresponding to the target first polar coordinate is determined to be a dangerous object; if the target temperature information of the real object is less than or equal to a second temperature threshold, then the real object corresponding to the target first polar coordinate is determined to be a safe object, and the second temperature threshold is less than the first temperature threshold.

[0131] By implementing the above method, it is possible to further determine whether unknown objects in the target scene are dangerous objects based on thermal infrared images, thereby improving the calibration quality of calibration information for each object in the real scene.

[0132] By implementing the methods disclosed in the above embodiments, the coordinates included in the three-dimensional coordinate information and the two-dimensional coordinate information can be transformed into a polar coordinate system. Then, the three-dimensional coordinate information and the two-dimensional coordinate information can be fused in the polar coordinate system to more accurately determine the position, category, and other calibration information of each real object in the real scene, thereby improving the calibration quality of the calibration information of each object in the real scene. In addition, when the category information of a real object in the target scene cannot be determined, the target two-dimensional coordinate information corresponding to the real object can be sent to the cloud server so that the cloud server can determine the category information corresponding to the real object, thereby improving the flexibility of the method.

[0133] Please see Figure 6 , Figure 6 This is a schematic diagram of the structure of an image processing apparatus disclosed in an embodiment of this application. This apparatus can be applied to a processor, such as an image processor. This application embodiment uses a processor as an example for illustration and should not be construed as limiting the scope of this application embodiment. Optionally, the apparatus may include an acquisition unit 602, a conversion unit 604, and an addition unit 606, wherein:

[0134] The acquisition unit 602 is used to acquire the three-dimensional coordinate information and the two-dimensional coordinate information of the target scene.

[0135] The conversion unit 604 is used to convert each three-dimensional coordinate included in the three-dimensional coordinate information into first polar coordinates, and to convert each first two-dimensional coordinate included in the two-dimensional coordinate information into second polar coordinates. The reference coordinate system corresponding to the first polar coordinates and the reference coordinate system corresponding to the second polar coordinates are the same.

[0136] Adding unit 606 is used to add identification information to the real object corresponding to the first polar coordinate of the target. The identification information includes category information corresponding to the second polar coordinate of the target. The category information is used to describe the category to which the real object corresponding to the second polar coordinate of the target belongs. The polar angle of the second polar coordinate of the target matches the polar angle of the first polar coordinate of the target. The first polar coordinate of the target is the first polar coordinate of the first polar coordinate that corresponds to the real object. The second polar coordinate of the target is the second polar coordinate that corresponds to the category information.

[0137] By implementing the above-described apparatus, three-dimensional coordinate information and two-dimensional coordinate information of the same target scene can be acquired. Then, the three-dimensional coordinates included in the three-dimensional coordinate information can be converted into first polar coordinates, and the first two-dimensional coordinates included in the two-dimensional coordinate information can be converted into second polar coordinates. Furthermore, identification information can be added to the object corresponding to the target first polar coordinate, which may include category information corresponding to the target second polar coordinate whose polar angle matches the target first polar coordinate. This category information describes the category to which the real-world object corresponding to the target second polar coordinate belongs. Therefore, by implementing the embodiments of this application, the coordinates included in the three-dimensional and two-dimensional coordinate information can be converted to a polar coordinate system. This allows for the fusion of the three-dimensional and two-dimensional coordinate information in the polar coordinate system, avoiding the loss of original information during the fusion process. Consequently, the location, category, and other calibration information of each real-world object in the real-world scene can be determined more accurately, improving the calibration quality of the calibration information for each object in the real-world scene.

[0138] As an optional implementation, the three-dimensional coordinate information is three-dimensional point cloud information. The conversion unit 604 is also used to project each of the three-dimensional coordinates included in the three-dimensional coordinate information into a rectangular coordinate system to obtain the second two-dimensional coordinates corresponding to each three-dimensional coordinate; and to convert each of the second two-dimensional coordinates into the corresponding first polar coordinates.

[0139] By implementing the above-mentioned device, the three-dimensional coordinates included in the three-dimensional coordinate information can be converted into two-dimensional coordinates, which in turn facilitates the conversion of the three-dimensional coordinates included in the three-dimensional coordinate information into the polar coordinate system, so as to obtain the polar coordinates corresponding to the three-dimensional coordinates included in the three-dimensional coordinate information, thereby improving the flexibility of the method.

[0140] As an optional implementation, the two-dimensional coordinate information is a two-dimensional image information conversion unit 604, which is further used to convert the two-dimensional coordinate information into a bird's-eye view to obtain the third two-dimensional coordinates corresponding to each of the first two-dimensional coordinates included in the two-dimensional coordinate information in the bird's-eye view; and to convert each of the third two-dimensional coordinates into the corresponding second polar coordinates.

[0141] By implementing the above device, the two-dimensional coordinate information can be converted into a bird's-eye view. Since the polar coordinates and categories of each real object in the bird's-eye view are easier to determine, it is more convenient to determine the polar coordinates and categories of each real object in the bird's-eye view in the subsequent process.

[0142] As an optional implementation, the conversion unit 604 is also used to determine the third two-dimensional coordinates corresponding to each of the first two-dimensional coordinates in the bird's-eye view based on the first two-dimensional coordinates, scale factor and inverse perspective transformation matrix included in the two-dimensional coordinate information, so as to obtain the bird's-eye view corresponding to the two-dimensional coordinate information.

[0143] By implementing the above device, two-dimensional coordinate information can be converted into a bird's-eye view through the principle of inverse perspective transformation. This conversion method is relatively easy to implement, thereby reducing the difficulty and cost of implementation.

[0144] As an optional implementation method, Figure 6 The apparatus shown may also include a determining unit (not shown), wherein:

[0145] The determination unit is used to determine the category information of each real object in the bird's-eye view before adding identification information to the object corresponding to the first polar coordinate of the target, and to determine the second polar coordinate of each real object. The category information is used to describe the category to which the real object corresponding to the second polar coordinate belongs.

[0146] By implementing the above-mentioned device, the category information of each real object in the real scene can be identified in advance through a bird's-eye view, which facilitates subsequent information fusion based on polar coordinates, thereby improving the calibration quality of the calibration information of each object in the real scene.

[0147] As an optional implementation, the adding unit 606 is further configured to use the category information as the identification information of the real object corresponding to the first polar coordinate when there is corresponding category information for the second polar coordinate that matches the first polar coordinate; and to determine that the real object corresponding to the first polar coordinate is an unknown object when there is no category information for the second polar coordinate that matches the first polar coordinate.

[0148] By implementing the above device, corresponding category information can be added to real objects whose category information can be identified, while real objects whose category information cannot be identified are determined to be unknown objects, thereby improving the calibration quality of the calibration information for each object in the real scene.

[0149] As an optional implementation, the acquisition unit 602 is also used to acquire three-dimensional point cloud information of the target scene through a three-dimensional lidar, and to acquire two-dimensional image information of the target scene through a two-dimensional camera device, wherein the information acquisition end of the three-dimensional lidar and the information acquisition end of the two-dimensional camera device are oriented in the same direction.

[0150] Alternatively, two-dimensional point cloud information of the target scene can be obtained through a two-dimensional LiDAR, and three-dimensional image information of the target scene can be obtained through a three-dimensional camera device, with the information acquisition end of the two-dimensional LiDAR and the information acquisition end of the three-dimensional camera device facing the same direction.

[0151] Implementing the above-mentioned device allows the lidar and camera to collect the three-dimensional coordinate information and two-dimensional coordinate information of the same target scene, respectively, which facilitates subsequent information fusion based on polar coordinates and improves the calibration quality of calibration information for various objects in the real scene.

[0152] Please see Figure 7 , Figure 7 This is a schematic diagram of the structure of an electronic device disclosed in an embodiment of this application. For example... Figure 7 As shown, the electronic device may include:

[0153] Memory 701 storing executable program code;

[0154] Processor 702 coupled to memory 701;

[0155] The processor 702 calls the executable program code stored in the memory 701 to execute the image processing methods disclosed in the above embodiments.

[0156] This application discloses a computer-readable storage medium storing a computer program that causes a computer to execute the image processing methods disclosed in the above embodiments.

[0157] This application also discloses an application publishing platform, which is used to publish computer program products. When the computer program products are run on a computer, the computer performs some or all of the steps of the methods described in the above method embodiments.

[0158] It should be understood that the phrase "one embodiment" or "an embodiment" throughout the specification means that a specific feature, structure, or characteristic related to the embodiment is included in at least one embodiment of this application. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. Furthermore, these specific features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. Those skilled in the art should also recognize that the embodiments described in the specification are optional embodiments, and the actions and modules involved are not necessarily essential to this application.

[0159] In the various embodiments of this application, it should be understood that the sequence number of each process does not necessarily imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0160] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they can 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.

[0161] 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.

[0162] If the aforementioned integrated units are implemented as software functional units and sold or used as independent products, they can be stored in a computer-accessible memory. 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 memory and includes several requests to cause a computer device (which can be a personal computer, server, or network device, specifically a processor in the computer device) to execute some or all of the steps of the methods described in the various embodiments of this application.

[0163] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, including read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electrically-Erasable Programmable Read-Only Memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, disk storage, magnetic tape storage, or any other computer-readable medium capable of carrying or storing data.

[0164] The foregoing has provided a detailed description of an image processing method, apparatus, electronic device, and computer-readable storage medium disclosed in the embodiments of this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the embodiments above are only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. An image processing method, characterized by, The method includes: Obtain the three-dimensional coordinate information of the target scene, as well as the two-dimensional coordinate information of the target scene; The three-dimensional coordinates included in the three-dimensional coordinate information are converted into first polar coordinates, and the first two-dimensional coordinates included in the two-dimensional coordinate information are converted into second polar coordinates. The reference coordinate system corresponding to the first polar coordinates and the reference coordinate system corresponding to the second polar coordinates are the same. If a target second polar coordinate exists that matches the target first polar coordinate, then the category information corresponding to the target second polar coordinate is used as the identification information of the real object corresponding to the target first polar coordinate; if no target second polar coordinate exists that matches the target first polar coordinate, then the real object corresponding to the target first polar coordinate is determined to be an unknown object. The identification information includes the category information corresponding to the target second polar coordinate. The category information is used to describe the category to which the real object corresponding to the target second polar coordinate belongs. The polar angle of the target second polar coordinate matches the polar angle of the target first polar coordinate. The target first polar coordinate is the first polar coordinate that corresponds to the real object in the first polar coordinate system, and the target second polar coordinate is the second polar coordinate that corresponds to the category information. The method further includes: determining the category information corresponding to each real object in the bird's-eye view corresponding to the two-dimensional coordinate information, and determining the second polar coordinates corresponding to each real object, wherein the category information is used to describe the category to which the real object corresponding to the second polar coordinate belongs.

2. The method of claim 1, wherein, The three-dimensional coordinate information is three-dimensional point cloud information, and the conversion of each three-dimensional coordinate included in the three-dimensional coordinate information into first polar coordinates includes: Project each of the three-dimensional coordinates included in the three-dimensional coordinate information into a rectangular coordinate system to obtain the second two-dimensional coordinates corresponding to each of the three-dimensional coordinates in the rectangular coordinate system; Convert each of the second two-dimensional coordinates into the corresponding first polar coordinates.

3. The method of claim 1, wherein, The two-dimensional coordinate information is two-dimensional image information. Converting each of the first two-dimensional coordinates included in the two-dimensional coordinate information into second polar coordinates includes: The two-dimensional coordinate information is converted into a bird's-eye view to obtain the third two-dimensional coordinates corresponding to each of the first two-dimensional coordinates included in the two-dimensional coordinate information in the bird's-eye view. Convert each of the third two-dimensional coordinates into the corresponding second polar coordinates.

4. The method of claim 3, wherein, The process of converting the two-dimensional coordinate information into a bird's-eye view includes: Based on the first two-dimensional coordinates, scale factor, and inverse perspective transformation matrix included in the two-dimensional coordinate information, the third two-dimensional coordinates corresponding to each of the first two-dimensional coordinates in the bird's-eye view are determined to obtain the bird's-eye view corresponding to the two-dimensional coordinate information.

5. The method of claim 1, wherein, The three-dimensional coordinate information includes three-dimensional point cloud information of the target scene obtained by a three-dimensional lidar, and the two-dimensional coordinate information includes two-dimensional image information of the target scene obtained by a two-dimensional camera device. The information acquisition end of the three-dimensional lidar and the information acquisition end of the two-dimensional camera device are oriented in the same direction. Alternatively, the three-dimensional coordinate information includes three-dimensional image information of the target scene acquired by a three-dimensional camera device, and the two-dimensional coordinate information includes two-dimensional point cloud information of the target scene acquired by a two-dimensional lidar, wherein the information acquisition end of the two-dimensional lidar and the information acquisition end of the three-dimensional camera device are oriented in the same direction.

6. An image processing apparatus characterized by comprising: The device includes: The acquisition unit is used to acquire the three-dimensional coordinate information of the target scene and the two-dimensional coordinate information of the target scene; The conversion unit is used to convert each of the three-dimensional coordinates included in the three-dimensional coordinate information into first polar coordinates, and to convert each of the first two-dimensional coordinates included in the two-dimensional coordinate information into second polar coordinates, wherein the reference coordinate system corresponding to the first polar coordinates and the reference coordinate system corresponding to the second polar coordinates are the same. An adding unit is configured to, when a target second polar coordinate that matches the target first polar coordinate exists, use the category information corresponding to the target second polar coordinate as the identification information of the real object corresponding to the target first polar coordinate; and, when no target second polar coordinate matches the target first polar coordinate, determine that the real object corresponding to the target first polar coordinate is an unknown object, wherein the identification information includes category information corresponding to the target second polar coordinate, the category information being used to describe the category to which the real object corresponding to the target second polar coordinate belongs, the polar angle of the target second polar coordinate matches the polar angle of the target first polar coordinate, the target first polar coordinate is the first polar coordinate in the first polar coordinate that corresponds to a real object, and the target second polar coordinate is the second polar coordinate that corresponds to category information; The determining unit is used to determine the category information corresponding to each real object in the bird's-eye view corresponding to the two-dimensional coordinate information before adding identification information to the object corresponding to the first polar coordinate of the target, and to determine the second polar coordinate corresponding to each real object. The category information is used to describe the category to which the real object corresponding to the second polar coordinate belongs.

7. An electronic device, comprising: The method includes a memory storing executable program code and a processor coupled to the memory; wherein the processor invokes the executable program code stored in the memory to perform the method as described in any one of claims 1 to 5.

8. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 5.

Images