Image recognition methods, devices, electronic equipment and media

Abstract

This application provides an image recognition method, apparatus, electronic device, and medium, aiming to reduce the operational difficulty of adjusting the object to be re-identified when it is necessary to adjust the initial recognition result of an object in an image. The method includes: displaying a target image and initial recognition information for recognizing a first object in the target image on a display interface; when a preset trigger event on the display interface is triggered, in response to a swipe operation performed on the target image based on the trigger event, determining a second object to be re-identified indicated by the swipe operation; wherein the trigger event is used to characterize re-indicating the object to be identified; recognizing the second object, and updating the initial recognition information in the display interface to the recognition information for recognizing the second object.

Description

Technical Field

[0001] This application relates to the field of image recognition, and in particular to an image recognition method, apparatus, electronic device and medium. Background Technology

[0002] Many current client applications have pre-installed image recognition functionality. This function is used to identify images, obtain recognition results, and display them. For example, some apps offer a scan function, which allows users to scan QR codes to identify information linked to them, such as item information or store information. Alternatively, users can scan an item to take a picture and then identify the object in the image to obtain its information, such as the object's name or store discount information.

[0003] However, in scenarios involving object recognition within captured images, related technologies typically involve a simple distinction between the foreground and background before identifying objects in the foreground. This approach often results in recognitions that are not what the user expects. For example, due to camera limitations, the object to be recognized might be located in the background, leading to an undesirable result. In such cases, the user often needs to repeatedly recalibrate the camera and retake images until the object is correctly identified in the foreground.

[0004] Therefore, object recognition in related technologies suffers from the dilemma of insufficient recognition flexibility and highly complex user operations for adjusting the object to be recognized. This dilemma directly leads to the problems of low efficiency and low accuracy in object recognition in related technologies. Summary of the Invention

[0005] To address the aforementioned issues, this application provides an image recognition method, apparatus, electronic device, and medium, aiming to improve the flexibility of object recognition and reduce the operational complexity for users who need to re-perform object recognition.

[0006] A first aspect of this disclosure provides an image recognition method, the method comprising:

[0007] The target image and initial identification information for identifying the first object in the target image are displayed on the display interface;

[0008] When a preset trigger event is triggered on the display interface, in response to a swipe operation performed on the target image based on the trigger event, a second object to be re-identified indicated by the swipe operation is determined; wherein, the trigger event is used to characterize the re-indication of the object to be identified;

[0009] The second object is identified, and the initial identification information in the display interface is updated to the identification information for identifying the second object.

[0010] Optionally, when a preset trigger event is triggered on the display interface, in response to a swipe operation performed on the target image based on the trigger event, determining the second object to be re-identified indicated by the swipe operation includes:

[0011] When a preset location icon on the display interface is triggered, in response to a swipe operation on the location icon, the movement trace of the location icon on the target image is determined; wherein, the location icon is used to indicate the location of the object to be identified;

[0012] Identify the image region where the end point of the movement trace is located in the target image;

[0013] The object located in the identified image region is identified as the second object.

[0014] Optionally, when a preset trigger event is triggered on the display interface, in response to a swipe operation performed on the target image based on the trigger event, determining the second object to be re-identified indicated by the swipe operation includes:

[0015] When a preset area icon on the display interface is triggered, in response to a swipe operation performed on the target image, the image area selected by the swipe trajectory of the swipe operation is determined in the target image; wherein, the area icon is used to indicate the object to be identified.

[0016] The object located in the image region is identified as the second object.

[0017] Optionally, determining the image region selected by the swipe trajectory of the swipe operation in the target image includes:

[0018] When the swipe trajectory is a closed trajectory, the area enclosed by the swipe trajectory is defined as the image area;

[0019] When the wiping trajectory is not closed, the starting point and the ending point of the wiping trajectory are connected, and the area enclosed by the wiping trajectory after connecting the starting point and the ending point is determined as the image area.

[0020] Optionally, the method further includes:

[0021] The trajectory graphic of the swipe path is displayed on the display interface;

[0022] In response to a preset operation triggered on the image region where the trajectory graphic is located, the positional relationship between the trigger point that triggered the preset operation and the trajectory graphic is determined;

[0023] In response to a movement operation performed on the trajectory graphic, the position of the trajectory graphic or the shape of the trajectory graphic is updated based on the positional relationship;

[0024] Determining an object located in the image region as the second object includes:

[0025] The object located in the image region framed by the updated trajectory graph is identified as the second object.

[0026] Optionally, in response to a movement operation on the trajectory graphic, updating the position of the trajectory graphic or the shape of the trajectory graphic based on the positional relationship includes:

[0027] If the location point that triggers the preset operation is located on the trajectory line of the trajectory graphic, the size of the trajectory graphic is updated in response to the movement operation performed on the trajectory graphic.

[0028] If the location point that triggers the preset operation, as indicated by the positional relationship, is located within the image area bounded by the trajectory graphic, the position of the trajectory graphic is updated in response to a movement operation performed on the trajectory graphic.

[0029] Optionally, the identification of the second object includes:

[0030] Identify the object name of the second object;

[0031] Based on the identified object name, at least one service information including the second object is determined from a preset set of services;

[0032] The business popularity corresponding to each of the at least one business information is determined, and based on the business popularity, the target business information corresponding to the second object is obtained, and the target business information is used as the identification information for identifying the second object.

[0033] Optionally, the method further includes at least one of the following:

[0034] When the initial recognition information in the display interface is updated to recognition information for recognizing the second object, a prompt message is output, which includes any one of sound prompt information, vibration prompt information, or sound and light prompt information;

[0035] If no identification information for recognizing the second object is obtained, guidance information is displayed, and a viewfinder page is displayed in response to a triggering operation of the guidance information; the viewfinder page is used to instruct the re-acquisition of the target image.

[0036] Optionally, the method further includes:

[0037] Perform object recognition on the target image;

[0038] The initial identification information for identifying the first object in the target image is displayed on the display interface, including:

[0039] When multiple objects are detected, in response to a selection operation of any one of the multiple objects, the initial identification information of the first object corresponding to the selection operation and the position information of the other objects besides the first object are displayed in the display interface.

[0040] When a single object is identified, the initial identification information of the single first object is displayed on the display interface.

[0041] Optionally, the initial identification information for identifying the first object in the target image is displayed in the display interface, including:

[0042] Add a location icon to the location of the identified object, and add an information display layer to the display interface;

[0043] The intermediate identification information during the identification process of the identified object is iteratively displayed at the location icon; the intermediate identification information includes category information, name information, and identification image;

[0044] Based on the user's selection of the first object among the identified objects, the recognition result information at the end of the recognition of the first object is displayed on the information display layer; the recognition result information includes the store information and business information to which the first object belongs.

[0045] Optionally, the method further includes:

[0046] When a preset trigger event on the display interface is triggered, the intermediate identification information is hidden;

[0047] The step of updating the initial recognition information in the display interface to recognition information for recognizing the second object includes:

[0048] The recognition result information in the information display layer is updated to the recognition information for recognizing the second object.

[0049] A second aspect of this application discloses an image recognition device, the device comprising:

[0050] The first display module is used to display the target image and initial identification information for identifying the first object in the target image in the display interface;

[0051] A response module is configured to, in response to a swipe operation performed on the target image based on the trigger event when a preset trigger event is triggered on the display interface, determine a second object to be re-identified indicated by the swipe operation; wherein the trigger event is used to characterize the re-indication of the object to be identified;

[0052] The re-identification module is used to identify the second object and update the initial identification information in the display interface with the identification information for identifying the second object.

[0053] A third aspect of this disclosure provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executed, implements the image recognition method as described in one aspect.

[0054] Furthermore, embodiments of this application also provide a computer-readable storage medium storing a computer program that causes a processor to execute the image recognition method as described in the first aspect.

[0055] In the technical solution of this application embodiment, a target image and initial identification information for identifying a first object in the target image can be displayed on the display interface; when a preset trigger event on the display interface is triggered, in response to a swipe operation performed on the target image, a second object to be re-identified indicated by the swipe operation is determined; then, the second object is identified, and the initial identification information in the display interface is updated to the identification information for identifying the second object.

[0056] The technical solution adopted in this application has at least the following advantages:

[0057] On the one hand, since the second object to be re-identified can be specified directly based on a trigger event by swiping on the target image, the background can then be instructed to identify the second object. This allows users to simply swipe on the target image displayed on the screen to indicate the second object without needing to exit the screen, readjust the camera, and retake the image, thus reducing user complexity and improving object recognition efficiency. On the other hand, since the foreground and background do not need to be considered when adjusting or changing the object to be identified, the second object is determined directly based on the user's swipe operation. This makes the object recognition method more flexible, and the clear definition of the second object based on the user's operation further improves the accuracy of object recognition. Attached Figure Description

[0058] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the description of the embodiments or related technologies 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.

[0059] Figure 1 This is a schematic diagram of the display interface of an image recognition method according to an embodiment of this application;

[0060] Figure 2 This is a schematic flowchart illustrating the steps of an image recognition method according to an embodiment of this application;

[0061] Figure 3 This is a schematic diagram illustrating the interface changes of an assistance measure triggered by the location icon, as shown in one embodiment of this application.

[0062] Figure 4 This is a schematic diagram illustrating the interface changes of an assistance measure triggered by the location icon, as shown in one embodiment of this application.

[0063] Figure 5 This is a schematic diagram of an interface for modifying the size of a trajectory graphic, as shown in one embodiment of this application;

[0064] Figure 6 This is a schematic diagram of an interface for correcting the position of a trajectory graphic, as shown in an embodiment of this application.

[0065] Figure 7 This is a flowchart illustrating the steps for identifying a second object according to an embodiment of this application;

[0066] Figure 8 This is a schematic diagram of the frame of an image recognition device according to an embodiment of this application. Detailed Implementation

[0067] 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, 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.

[0068] It should be noted that the terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this disclosure 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 disclosure described herein can be implemented in orders other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this disclosure. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this disclosure as detailed in the appended claims.

[0069] In related technologies, when recognizing objects in an image, if the identified object information is inaccurate—for example, not the information the user needs—the object to be recognized is usually readjusted. For instance, the system might identify object A, but the user actually wants to identify object B. In this case, the user will adjust the recognition target to object B. Typically, the user will recalibrate the camera multiple times and retake images until object B is located in the foreground area. However, this method requires frequent camera calibration, making the process cumbersome.

[0070] In related technologies, a method for adjusting and replacing the object to be identified is also provided. Specifically, this involves offering users assistance in adjusting the object. For example, the identified object is marked with a regular rectangle. The user can press and drag the four corners of the rectangle to adjust its length and width, or drag the rectangle to adjust its position, thereby reselecting the object. However, this method requires the user to press and drag the four corners of the rectangle, which is inconvenient. Furthermore, since the identified object needs to be marked with a regular rectangle, additional object position recognition and selection are required, increasing the workload. Even if the accuracy of the selected object's position is not high, the user still needs to make corrective adjustments to the rectangle, which is no simpler than recalibrating a camera lens.

[0071] In view of this, this application proposes the following technical concept: a trigger event is preset on the display interface to indicate the object to be replaced. When the trigger event is triggered, the second object to be re-identified can be determined according to the swipe operation performed by the user on the target image. Then, the initial recognition information in the display interface can be updated to the recognition information for recognizing the second object. In this way, the replacement of the object to be recognized is realized.

[0072] Because the second object to be re-identified can be specified directly based on a trigger event by swiping on the target image, the background process is instructed to identify the second object. This allows users to simply swipe on the target image displayed on the screen to indicate the second object. Compared to related technologies, this application eliminates the need to exit the display interface, readjust the camera, and retake images, as well as the need to pre-select already identified objects using a rectangle. Therefore, no corrective adjustments to the rectangle are required, significantly simplifying the user's operation and improving overall object recognition efficiency.

[0073] Reference Figure 1 As shown, a schematic diagram of an image recognition display interface is illustrated. Figure 1 As shown, the display interface displays a target image, which may include an image area of ​​at least one object. The display interface can also display recognition information for identifying the object in the target image.

[0074] Below, in conjunction with Figure 1 As shown, the image recognition method of this application will be described in detail. It should be noted that the image recognition method of this application can be applied to a client or to a system consisting of a client and a server. When applied to a client, the image recognition method of this application can be implemented based on the client. When applied to a system consisting of a client and a server, the client can be used to support the display of the target image and recognition information, as well as the readjustment of the object to be recognized, while the server can undertake the recognition task of recognizing the specified object in the target image.

[0075] Reference Figure 2 As shown, a flowchart illustrating the steps of the image recognition method of this application is presented, as follows: Figure 2 As shown, the specific steps may include:

[0076] Step S201: Display the target image and initial identification information for identifying the first object in the target image in the display interface.

[0077] In this embodiment of the application, the display interface may be as follows: Figure 1The display interface shown can simultaneously display the target image and initial identification information for recognizing the first object in the target image. The target image can be a real-time image captured by a camera, or an image captured during a historical process and stored in a database.

[0078] The target image can be a photograph taken of an object. Since it is a photograph taken of an object, other objects in the environment in which the object is located will also be captured. Therefore, the target image can include images of one or more objects, such as... Figure 1 As shown, the target image may include two Coca-Cola bottles, as well as objects in the environment such as chairs and windows.

[0079] In this configuration, the initial recognition information displayed on the screen can be separate from the target image; that is, the target image occupies a separate display area for complete display, while the initial recognition information occupies another separate display area. In this case, the target image can be fully displayed on the screen.

[0080] Alternatively, the initial recognition information displayed in the interface may cover a portion of the target image area; for example, the initial recognition information may cover a portion of the left side, a portion of the right side, a portion of the top side, or a portion of the bottom side of the target image. Figure 1 As shown, the initial recognition information covers the lower half of the target image. In this case, the target image is covered by the initial recognition information in a portion of the displayed interface.

[0081] Of course, this application does not restrict the display method between the initial identification information and the target image.

[0082] Step S202: When a preset trigger event is triggered on the display interface, in response to a swipe operation performed on the target image based on the trigger event, determine the second object to be re-identified indicated by the swipe operation.

[0083] The triggering event is used to characterize the re-indication of the object to be identified.

[0084] In this embodiment, the trigger event can be preset in the display interface. When the trigger event is triggered, it indicates that the user wants to re-specify the object to be identified. In this case, the user will perform a swipe operation based on the trigger event, and then, in conjunction with the trigger event, the second object to be re-identified indicated by the swipe operation can be determined. After the second object is determined, the area where the second object is located can be highlighted, or the brightness of other areas in the target image other than the second object can be reduced.

[0085] Here, the second object to be re-identified can be understood as an object that needs to be re-identified, and this second object can be different from the first object. Alternatively, when the initial identification information of the first object is inaccurate, for example, when the identified first object contains interfering information from other objects, it can help to more accurately define the first object. In this case, the second object and the first object can be substantially the same object, but the second object contains less interfering information.

[0086] Different triggering events can lead to different ways of identifying the second object through the swipe operation. Specifically, depending on the triggering event, the swipe operation can identify the second object by either locating its position or by defining the image region containing it. For example, such as... Figure 1 As shown, assuming the first object is the cola bottle on the left and the second object is the cola bottle on the right, the swipe operation can be to circle the cola bottle on the right or to point out the position of the cola bottle on the right.

[0087] In this way, different swipe operation response logics can be activated based on different triggering events, thereby determining the second object indicated by the swipe operation based on the activated response logic.

[0088] In this embodiment, from the perspective of the backend response logic of the display interface, the swipe operation specifically refers to only a single location point on the display interface being triggered at the same time, rather than requiring multiple location points to be triggered simultaneously. Therefore, the backend only needs to track the single triggered location point at a time, reducing the monitoring load and the probability of monitoring errors. From the user's perspective, the swipe operation specifically refers to the user swiping on the display interface with a single finger, eliminating the need for the user to swipe with two or more fingers simultaneously as described in the related technologies of the above embodiments. This reduces the difficulty of user operation. Therefore, the swipe operation reduces the difficulty of operation and the response pressure on the backend, both from the perspective of backend response and frontend user operation.

[0089] Step S203: Identify the second object and update the initial identification information in the display interface to the identification information for identifying the second object.

[0090] In this embodiment, the second object can be re-identified. In practice, identifying the second object may include identifying the name, category, trademark, and business information of the second object. Thus, the identification information may include one or more of the name, category, trademark, and business information of the second object.

[0091] Specifically, upon obtaining the recognition information of the second object, the initial recognition information in the display interface can be updated to reflect the second object's recognition information, thus allowing the second object's recognition information to be displayed on the interface. This achieves a readjustment of the target object for object recognition in the target image.

[0092] It should be noted that in step S201 above, the initial recognition information is indicated to cover part of the target image area. In this case, when the trigger event on the display interface is triggered, the initial recognition information can be hidden to display the complete target image. When the recognition information of the second object is obtained, the initial recognition information can be replaced with the recognition information of the second object, and then the recognition information of the second object can be covered on part of the target image area.

[0093] The technical solution adopted in this application has several advantages. First, because a second object to be re-identified can be specified directly based on a trigger event by swiping on the target image, the user can simply indicate the second object on the target image displayed on the screen without needing to exit the screen, readjust the camera, and retake the image. This reduces the complexity of user operations and improves object recognition efficiency. Second, because the object to be identified does not need to be considered in terms of foreground and background when adjusting or changing it, the second object can be determined directly based on the user's swipe operation. This makes the object recognition method more flexible, and the accuracy of object recognition is further improved by clearly defining the second object to be identified based on the user's operation. Third, because it is a detected swipe operation, the difficulty of operation and the response pressure on the backend are reduced, both from the perspective of backend response and frontend user operation.

[0094] In this application, different triggering events can lead to different methods of identifying the second object indicated by the swipe operation. To provide users with more flexible ways to specify the second object to be re-identified, two assistance measures are provided to help users readjust the second object to be identified. Specifically, the triggering event can be the selection of a preset icon, which may include a positioning icon and a zone icon. When different icons are triggered, different response logic can be executed in response to the swipe operation, thereby providing different assistance measures to determine the indicated second object.

[0095] In practice, when a preset trigger event is detected, the corresponding response logic can be determined based on the type of the triggered event, and the second object to be re-identified indicated by the swipe operation can be determined according to the corresponding response logic.

[0096] The two assistance measures will be described separately below:

[0097] The first assistance measure is one triggered by the location icon, where the location icon indicates the location of the object to be identified. Specifically:

[0098] First, when a preset location icon on the display interface is triggered, in response to a swipe operation on the location icon, the movement trace of the location icon on the target image is determined.

[0099] Next, the image region where the end point of the movement trace is located in the target image is identified.

[0100] Then, the object located in the identified image area is identified as the second object.

[0101] In this embodiment, a preset positioning icon can be displayed at the location of the first object to indicate its position in the target image. Of course, the preset positioning icon can also be displayed anywhere on the display interface, as long as it is easily accessible to the user.

[0102] The system can be triggered by clicking a location icon. When the location icon is triggered, the identification method for the second object indicated by the swipe operation is to locate the position of the second object to be identified. The corresponding response logic can be to determine the movement trace of the swipe operation on the location icon, and then determine the ending position of the location icon based on the movement trace. Then, the image area where the ending position of the location icon is located is identified, and the object in the identified image area is identified as the second object.

[0103] The end point of the movement trace is the end point of the bitmap. When identifying the image region where the end point is located, the texture features of the region can be obtained. Since texture features can be used to distinguish between objects and background regions that do not belong to objects, the region formed by the texture features of the same object can be identified as the image region based on the texture features.

[0104] For example, refer to Figure 3 As shown, this diagram illustrates the interface changes of the assistance measures in this application, where the location icon is triggered as a result. Figure 3 As shown, including Figure 3 -A、 Figure 3 -B and Figure 3 -C, such as Figure 3 As shown in -A, the location icon is displayed at the position of the first object (the Coke bottle on the left). When the location icon is selected, it becomes draggable. You can drag the location icon to the position of the Coke bottle on the right with a single finger, as shown below. Figure 3As shown in -B, the positioning icon stays at the location of the right cola bottle, thus recognizing the image area of ​​the right cola bottle. Then, the recognition information of the right cola bottle is displayed, such as... Figure 3 -C is shown.

[0105] In another implementation, the target image displaying the end position of the positioning icon can be included in a pre-trained object recognition model. The object recognition model then determines the second object in the image region corresponding to the end position of the positioning icon. The training samples for the object recognition model can include multiple target image samples, each containing a positioning icon and a bounding box labeled with a rectangle used to select objects within the image region containing the positioning icon. During training, a loss function can be constructed based on the bounding boxes of the objects recognized by the model and the positional differences between the labeled bounding boxes. The model parameters are then updated based on the loss value of the loss function, thus obtaining the object recognition model.

[0106] In this implementation, the second object to be re-identified can be located by moving the positioning icon. This makes the operation more engaging and improves overall efficiency due to its simplicity. Furthermore, since the positioning icon can be displayed at the location of the first object, it indicates that the initially identified object has been identified, thus enriching the information displayed on the interface.

[0107] The second assistance measure is one that uses the triggering of a zone icon as the trigger event. The zone icon is used to indicate the object to be identified, as detailed below:

[0108] First, when a preset area icon on the display interface is triggered, in response to a swipe operation performed on the target image, the image area selected by the swipe trajectory of the swipe operation is determined in the target image.

[0109] Next, the object located in the image region is identified as the second object.

[0110] In this embodiment, a pre-set area icon can be set on the display interface. When the area icon is triggered, the swiping operation performed by the user on the target image can be detected. In practice, since the area icon indicates the selection of the object to be identified, the swiping operation performed by the user on the target image can be the operation of circling the second object to be re-identified. In this way, the object in the area selected by the swiping operation trajectory can be identified as the second object to be re-identified. At this time, the swiping trajectory can be understood as a selection box for selecting objects.

[0111] For example, refer to Figure 4The diagram illustrates the interface changes of the assistance measures described in this application, where the location icon is triggered as a specific event. Figure 4 The three figures -A, 4-B, and 4-C, as follows: Figure 4 As shown in -A, the area icon is located in the upper right corner of the initial recognition information. Clicking the area icon will take you to... Figure 4 middle Figure 4 The -B display interface allows users to swipe on the target image, with swipe trajectories such as... Figure 4 As shown in -B, the cola bottle selected by the swipe trajectory can then be identified as the second object, thus... Figure 4 The display interface shown in -D identifies the information of the selected Coca-Cola bottles.

[0112] When a user triggers the area icon and selects the second object to be re-identified by swiping, the swiping trajectory may be closed or open. For example... Figure 4 -B shows an unclosed sliding trajectory.

[0113] In practice, to ensure the accuracy of determining the image region based on the swipe trajectory formed by the swipe operation, adaptive processing can be performed on both closed and open swipe trajectories.

[0114] In specific implementation, when the wiping trajectory is a closed trajectory, the area enclosed by the wiping trajectory is defined as the image area.

[0115] When the wiping trajectory is not closed, the starting point and the ending point of the wiping trajectory are connected, and the area enclosed by the wiping trajectory after connecting the starting point and the ending point is determined as the image area.

[0116] In this embodiment, if the swiping trajectory is a closed trajectory, the swiping trajectory can enclose a closed image area. In this way, the area enclosed by the swiping trajectory can be directly determined as the image area, and the object in the image area can be determined as the second object to be re-identified.

[0117] In this embodiment, if the trajectories are not closed, such as... Figure 4 -B shows a non-closed sliding trajectory. In this case, the sliding trajectory can be processed; specifically, the starting and ending points of the sliding trajectory can be automatically connected to form a closed sliding trajectory. For example... Figure 4 -C indicates that... Figure 4 Connecting the two endpoints of the -B swipe trajectory results in a closed trajectory. In this case, the area enclosed by the swipe trajectory can be defined as the image region, and the object in this image region can be identified as the second object to be re-identified.

[0118] This implementation method can improve the fault tolerance of the user's swipe operation when the second object is circled, so that the second object can be accurately located even if the user draws a non-closed trajectory. In this way, it is not required for the user to draw a closed trajectory, thereby reducing the operational requirements for the user.

[0119] Of course, in some embodiments, the swipe trajectory formed by the swipe operation can also be directly modified and edited to readjust the object to be re-identified.

[0120] Specifically, the trajectory graphic of the swipe path can be displayed on the display interface, and in response to a preset operation triggered on the image area where the trajectory graphic is located, the positional relationship between the trigger point that triggers the preset operation and the trajectory graphic can be determined; then, in response to a movement operation performed on the trajectory graphic, the position of the trajectory graphic or the shape of the trajectory graphic can be updated based on the positional relationship.

[0121] Accordingly, objects located in the image region framed by the updated trajectory graph can be identified as the second object.

[0122] In this embodiment, the trajectory graph of the wiping path can be as follows: Figure 4 The graphic shown in -B displays the user's swipe trajectory. When it's necessary to modify or edit the trajectory to more accurately define a second object, the shape and size of the trajectory can be changed, as can its position. Specifically, the shape, size, and position of the trajectory can be modified by triggering different locations within the image area containing the trajectory graphic.

[0123] In practice, the preset operation triggered on the image area where the trajectory graphic is located can be a long press operation. When the user long presses a certain position in the image area where the trajectory graphic is located, the shape, size and position of the trajectory graphic can be modified. In practice, the user can perform a movement operation on the trajectory graphic. Specifically, the movement operation can also refer to a single-finger swipe operation.

[0124] The backend determines the response logic for the movement operation based on the positional relationship between the triggered preset operation location and the trajectory graphic. This positional relationship can refer to whether the location triggering the preset operation lies on the trajectory line of the trajectory graphic. Based on this result, the response logic for the movement operation is determined. This response logic can include logic regarding shape and size, as well as logic regarding position changes, as described below:

[0125] If the location point that triggers the preset operation is located on the trajectory line of the trajectory graphic, the size of the trajectory graphic is updated in response to the movement operation performed on the trajectory graphic.

[0126] If the location point that triggers the preset operation, as indicated by the positional relationship, is located within the image area bounded by the trajectory graphic, the position of the trajectory graphic is updated in response to a movement operation performed on the trajectory graphic.

[0127] In this embodiment, when the trigger point for the preset operation is located on the trajectory line of the trajectory graphic, it indicates that the user has triggered the preset operation on the trajectory line. For example, the user long-presses the trajectory line, thereby triggering a modification of the size of the trajectory graphic. In this case, the backend response logic controls the trajectory graphic to enlarge or shrink as the user moves. For example, if the movement is along a first direction, the trajectory graphic will enlarge, and if the movement is along a second direction, the trajectory graphic will shrink. The first direction is the opposite of the second direction.

[0128] Reference Figure 5 As shown, this is a schematic diagram of the interface for modifying the size of the trajectory graph in this case. Figure 5 As shown in -A, users can long-press the trajectory line of the graphic and then drag it. The trajectory line will change according to the direction of the drag operation, such as... Figure 5 middle Figure 5 As shown in -B, dragging to the right enlarges the trajectory graphic, while dragging to the left shrinks it. When the user deems the trajectory graphic to be of a suitable size, they can lift their finger to stop resizing it. The system then identifies the object selected by the changed trajectory graphic and obtains the desired result. Figure 5 The result shown in -C.

[0129] In this embodiment, when the trigger point for the preset operation is located within the image area enclosed by the trajectory graphic, it indicates that the user triggered the preset operation within the area enclosed by the trajectory line, rather than on the trajectory line itself. For example, the user long-presses an area within the trajectory line, thereby triggering a correction to the position of the trajectory graphic. In this case, the backend response logic controls the trajectory graphic to move in accordance with the user's movement.

[0130] Reference Figure 6 As shown, a schematic diagram of the interface for correcting the position of the trajectory graph in this case is illustrated. Figure 6 As shown in -A, users can long-press on an area within the trajectory line and then perform a movement operation. The trajectory graphic will then move along with the movement operation. Figure 6 middle Figure 6As shown in -B, for example, dragging to the right moves the trajectory graphic to the cola bottle on the right, thus changing the second object and ultimately obtaining... Figure 6 The result shown in -C.

[0131] In this embodiment, after changing the size and position of the trajectory graphic, the object in the image area selected by the new trajectory graphic can be identified as the second object.

[0132] In this implementation, because the positional relationship between the trigger point and the trajectory line of the trajectory graphic can be used to respond to the user's movement operation using different response logics, the size and position of the trajectory graphic can be changed. This allows the user to modify the size and position of the graphic with simple operations. For example, the user can modify the graphic size by long-pressing and dragging the trajectory line, and correct the graphic position by long-pressing and dragging the area enclosed by the trajectory line. Compared to related technologies where modifying the graphic size requires moving all endpoints of the graphic simultaneously, this application reduces the difficulty of user operation.

[0133] In some embodiments, the object recognition of this application is described in detail.

[0134] In this embodiment, the initial identification information of the first object can be obtained by automatically identifying the object in the target image when the target image is captured. Specifically, automatic identification includes the identification of the first object in the target image and the identification information of the first object. That is, when performing object identification on the target image, it is necessary to identify both the object included in the target image and the identification information of the object.

[0135] Of course, a target image can generally include multiple objects, or it can include only a single object. When displaying the initial recognition information for identifying the first object in the target image on the display interface, the following cases can be handled:

[0136] Scenario 1: When multiple objects are detected, in response to the selection operation of any one of the multiple objects, the initial identification information of the first object corresponding to the selection operation and the position information of the other objects besides the first object are displayed on the display interface.

[0137] Scenario 2: When a single object is identified, the initial identification information of the single first object is displayed on the display interface.

[0138] In this embodiment, when there are multiple objects in the target image, during initial recognition, some or all of the objects can be identified. For example, Figure 1As shown, the target image contains five objects: two cola bottles, a window, a chair, and a windowsill. During initial recognition, three objects were identified, such as the two cola bottles and the chair. In this case, since multiple objects were identified, the initial recognition information for one specific object can be displayed based on the user's selection.

[0139] For example, such as Figure 1 As shown, if the user selects the cola bottle on the left, the initial identification information of the cola bottle will be displayed on the screen. Among these,

[0140] The timing of the selection operation can be before object recognition is performed on the target image. For example, when the target image is acquired, the user can click on one of the objects for recognition, and then the object recognition is performed. In this case, if multiple objects are still recognized, the initial recognition information of the object clicked by the user will be displayed first.

[0141] Of course, the timing of the operation can also be after object recognition is performed on the target image. If multiple objects are recognized, the recognition information of multiple objects may be displayed on the display interface. In this case, you can click on one of the objects, which will hide the recognition information of other objects on the display interface and display the initial recognition information of the first selected object.

[0142] Of course, if only a single object is identified during object recognition, and that single object is the first object mentioned above, then the initial recognition information of the single first object can be displayed on the display interface.

[0143] In order to enable users to obtain the progress of the recognition process while waiting for the initial recognition information, some implementations may display the recognition process information of the target image on the display interface, including displaying intermediate recognition information and recognition result information during the object recognition process in the target image.

[0144] In practice, a location icon can be added to the location of the identified object, and an information display layer can be added to the display interface. Then, intermediate identification information during the object identification process can be iteratively displayed at the location icon; this intermediate identification information includes category information, name information, and an identifier image.

[0145] Subsequently, based on the user's selection of the first object among the identified objects, the recognition result information at the end of the recognition of the first object is displayed on the information display layer; the recognition result information includes the store information and business information to which the first object belongs.

[0146] In this embodiment, when performing initial object recognition on the target image, the objects included in the target image can be identified first, then the location of the objects can be identified, and a positioning icon can be added to the location of each identified object, as well as an information display layer can be added to the display interface.

[0147] The location icon can be the same as the location icon described in the above embodiments, used to indicate the identified object. The intermediate identification information in the process of identifying the identified object can be iteratively displayed near the location icon. Iterative display means that when an intermediate identification information is identified, the displayed intermediate identification information is updated to the currently identified intermediate identification information.

[0148] In practice, the identification process can proceed in the following order: first, identify the category to which the object belongs; then, identify the name of the object; and finally, identify the object's logo image. The logo image can refer to the brand logo image to which the object belongs.

[0149] For example, such as Figure 1 As shown, if three objects are identified, namely two Coke bottles and a chair, then location icons can be displayed on the two Coke bottles and the chair respectively, and the intermediate identification information can be displayed near the location icons.

[0150] Upon completion of object recognition, recognition result information is obtained. In this case, as described in the above embodiment, based on the user's selection operation of the first object among the recognized objects, the recognition result information at the end of the recognition of the first object can be displayed on the information display layer. As mentioned above, the recognition result information includes the store information and business information to which the object belongs. The selection operation can be performed before the initial object recognition begins.

[0151] In this scenario, the recognition process can be displayed on the interface, making the backend recognition process visual. For example, the recognition process, which displays information such as the object's location, category, and name, can be scrolled out so that users can track the recognition progress. When displaying the recognition results, only the recognition information of the objects that the user is interested in can be shown. This satisfies both the flexibility of the recognition information display and improves the alignment between the displayed recognition information and the user's expectations.

[0152] It should be noted that this application displays the recognition result information through an information display layer. This information display layer may cover part of the target image or may not overlap with the target image.

[0153] After displaying the initial recognition information, if the user points to a second object for re-recognition, the second object will be recognized. In one implementation, when outputting the recognition information for the second object, the user can be prompted accordingly, specifically, to prompt the user that the recognition information for the second object has been recognized.

[0154] In specific implementation, when the initial recognition information in the display interface is updated to the recognition information for recognizing the second object, a prompt message can be output. The prompt message includes any one of sound prompt message, vibration prompt message, or sound and light prompt message.

[0155] The audio prompt can be a ringtone, which can be used to alert the user when the initial recognition information is updated to recognize a second object. Alternatively, the vibration component on the terminal can be activated to provide a vibrational alert. Of course, a combination of both can be used, meaning both audio and vibrational prompts can be employed.

[0156] The audio-visual prompts can be selected appropriately based on the hardware configuration of the terminal. For example, if the terminal's hardware configuration includes an information prompt light source, the light source and prompt tone can be activated to provide audio-visual prompts. The information prompt light source could be, for example, a flashing light.

[0157] In this embodiment, if identification information for recognizing the second object is not obtained, i.e., if the recognition of the second object fails, the user can return to the initial display interface to guide them to retake the image. Specifically, if identification information for recognizing the second object is not obtained, guidance information can be displayed, and a viewfinder page can be displayed in response to a triggering operation of the guidance information; the viewfinder page is used to instruct the user to re-acquire the target image.

[0158] In this application, the guidance information may carry a page jump control, which is used to jump from the current display interface to the initial display interface. The initial display interface may be a viewfinder interface depending on the object recognition scene. When the user triggers the page jump control in the guidance information, they can jump to the viewfinder page, which displays the viewfinder frame. At the same time, when the viewfinder page is displayed, the camera device of the terminal can be automatically turned on, and the scene pointed at by the camera device can be displayed in the viewfinder frame, thereby instructing the user to retake the target image using the camera device.

[0159] Of course, in some implementations, prompting information can also be output simultaneously with the guidance information. This prompting information may differ from the prompting information output upon obtaining the identification information of the second object, as described above. For example, if the prompting information output for the identification information of the second object is a vibration prompt, then the prompting information output for the guidance information may be a ringtone prompt. Alternatively, the vibration frequency of the vibration prompt output for the identification information of the second object may differ from the vibration frequency of the vibration prompt output for the guidance information.

[0160] When this implementation method is adopted, the identification information of the second object can be effectively prompted, and if the identification information of the second object is not identified, the user can automatically return to the viewfinder page to retake the image, thereby optimizing the user experience.

[0161] The process of identifying the second object will be explained below, referring to... Figure 7 As shown, a flowchart illustrating the steps for recognizing a second object is presented, such as... Figure 7 As shown, it includes the following steps:

[0162] Step S701: Identify the object name of the second object.

[0163] Step S702: Based on the identified object name, determine at least one service information that includes the second object from a set of preset services.

[0164] Step S703: Determine the business popularity corresponding to the at least one business information, obtain the target business information corresponding to the second object based on the business popularity, and use the target business information as the identification information for identifying the second object.

[0165] In this embodiment, the object name of the second object can be identified first. This object name can be determined based on the category to which the second object belongs. The object name of the second object can be a generic name of its category or a brand name of its brand, such as... Figure 1 As shown, it can be the generic name of cola, "Cola," or it can be "Coca-Cola."

[0166] Next, based on the object name, business information corresponding to the object name can be searched in the current database. This business information can refer to business information related to the object name, such as the sales information of the object published by the merchant. For example, in a product search scenario, it could be the discount information of the second object published by the store that sells the second object, or it could be the sales details information of the second object published by the store that sells the second object.

[0167] In practice, there are usually multiple business information items, including the second object. In this case, the business popularity corresponding to each business information item can be determined. The business popularity can be determined based on the number of clicks, orders, etc. of each business information item. For example, the higher the number of clicks and orders of a business information item, the higher the corresponding business popularity.

[0168] In this embodiment, business information with a popularity higher than a preset popularity level can be selected from at least one type of business information as target business information, and this target business information can be used as the identification information of a second object. There can be one or more target business information items. When there are multiple target business information items, they can be arranged sequentially on the display interface according to their popularity level from highest to lowest.

[0169] For example, if the second object is a flower named "chrysanthemum," then business information related to chrysanthemums will be retrieved. This information includes chrysanthemum exhibitions held at AA Park, chrysanthemum sales information from a flower shop, and chrysanthemum tea sales information from other tea shops. In practice, the information about the chrysanthemum exhibitions held at AA Park receives the most attention, so this information can be used as the identification information for chrysanthemums.

[0170] When the technical solution of this implementation method is adopted, it can be used in a wider range of scenarios, including the identification of objects such as goods, plants, exhibits, food, signs, and posters, as well as indoor environments, parks, landmarks, and other scenarios. It can be effectively integrated with various online businesses, such as food delivery, store search, exhibition booking, and order placement.

[0171] Accordingly, since intermediate recognition information and recognition result information are displayed on the display interface during the initial recognition process, the intermediate recognition information can be hidden when a preset trigger event on the display interface is triggered. This method ensures that the target image is not obscured by the intermediate recognition information at the first object, allowing the complete target image to be displayed on the display interface, thus facilitating the user to re-indicate the object to be recognized.

[0172] Next, when the recognition information of the second object is obtained, the recognition result information in the information display layer can be updated to the recognition information for recognizing the second object.

[0173] The image recognition method of this application will be introduced below using a specific example. Figure 1 The following explanation uses the displayed interface as an example:

[0174] S1, the user clicks the "Scan" control in the client to enter the viewfinder interface. The viewfinder displays the current environmental scene that the camera can capture. The user can determine the shooting quality of the target object by looking at the displayed environmental scene.

[0175] S2, the user clicks the shooting control, places the object to be identified in the viewfinder, and takes a picture as shown. Figure 1 The target image shown includes two Coca-Cola bottles, as well as a chair, window, and windowsill in the environment.

[0176] S3: When the user clicks on the Coca-Cola bottle on the left, it triggers initial object recognition of the target image and displays a location icon on the Coca-Cola bottle at the clicked location.

[0177] S4, the client identifies the Coca-Cola bottle, chair, and window on the left side of the target image. The positioning icons of the chair and window are displayed sequentially on the display interface. Near the positioning icons of each identified object, the category and object brand name are displayed sequentially according to the recognition process.

[0178] The recognition result of the Coca-Cola bottle on the left is a Coca-Cola promotional information launched by a store. This promotional information is displayed in a pop-up window (information display layer) below the target image and covers part of the target image.

[0179] S5, the user clicks the area icon to start re-designating the object to be identified. At this time, the pop-up window displaying the discount information of the Coke bottle on the left slides down to hide, and the positioning icon and various intermediate recognition information on the target image are hidden to reveal the complete target image.

[0180] S6, begin region division. Detect user swipe actions on the target image and obtain the swipe trajectory, such as... Figure 4 middle Figure 4 The drawing trajectory shown on the -B display interface is illustrated. When the drawn outline is not closed, the system will automatically connect the two ends of the line to close the outline when the user ends the drawing operation. After the outline is closed, a selection box is formed. The brightness of the unselected area in the selection box can be reduced, and the client begins to identify the Coca-Cola bottle in the selection box.

[0181] S7 displays the recognition result of the Coca-Cola bottle on the right. When the recognition information of the Coca-Cola bottle on the right is obtained, a pop-up window is swiped up, displaying the recognition information of the Coca-Cola bottle on the right. At the same time as the pop-up window is swiped up, the smartphone on the client side vibrates; if the recognition information of the Coca-Cola bottle on the right is not recognized, guidance information is displayed, and a vibration effect is generated, guiding the user back to the viewfinder interface in step S1.

[0182] When the "District Recognition" icon changes to a "Close" button, you can click the "Close" button to turn off the zone recognition function.

[0183] Of course, if the user thinks that the recognition information of the Coca-Cola bottle on the right is not accurate enough after receiving the recognition information, the reason may be that the position of the drawn area does not meet expectations. In this case, the user can press and hold the area within the drawing trajectory with one finger and drag it. The entire selection box will be moved. After moving the drawing trajectory to the appropriate position, lift the finger and immediately re-recognize the selected area.

[0184] Alternatively, if the user thinks the selected area is too large or too small, they can press and hold the swipe path with one finger and drag it. The swipe path will change with the direction of the finger movement. After moving the swipe path to the appropriate position, lift the finger, and the selection box will form a new shape. The system will then immediately re-identify the selected area.

[0185] Based on the same inventive concept as the above embodiments, a second aspect of the present disclosure provides an image recognition device, referring to... Figure 8 As shown, a schematic diagram of the image recognition device is presented, such as... Figure 7 As shown, the device may specifically include the following modules:

[0186] The first display module 801 is used to display a target image and initial identification information for identifying a first object in the target image in a display interface.

[0187] The response module 802 is configured to, in response to a swipe operation performed on the target image based on the trigger event when a preset trigger event is triggered on the display interface, determine the second object to be re-identified indicated by the swipe operation; wherein the trigger event is used to characterize the re-indication of the object to be identified;

[0188] The re-identification module 803 is used to identify the second object and update the initial identification information in the display interface to the identification information for identifying the second object.

[0189] Optionally, the response module 802 includes:

[0190] A motion trace monitoring unit is used to determine the motion trace of the positioning icon on the target image in response to a swipe operation on the positioning icon when a preset positioning icon on the display interface is triggered; wherein the positioning icon is used to indicate the location of the object to be identified.

[0191] The region recognition unit is used to recognize the image region in the target image where the end point of the movement trace is located;

[0192] The first object determination unit is used to determine the object located in the identified image region as the second object.

[0193] Optionally, the response module 802 includes:

[0194] The region determination unit is used to determine the image region selected by the swiping trajectory of the swiping operation in the target image when the preset area icon on the display interface is triggered, in response to the swiping operation performed on the target image; wherein the area icon is used to indicate the object to be identified.

[0195] The second object determination unit is used to determine the object located in the image region as the second object.

[0196] Optionally, the region determination unit includes:

[0197] The first determining subunit is used to determine the area enclosed by the wiping trajectory as the image area when the wiping trajectory is a closed trajectory;

[0198] The second determining subunit is used to connect the starting point and the ending point of the wiping trajectory when the wiping trajectory is not closed, and to determine the area enclosed by the wiping trajectory after connecting the starting point and the ending point as the image area.

[0199] Optionally, the device further includes:

[0200] A trajectory graphic display module is used to display the trajectory graphic of the wiping trajectory on the display interface;

[0201] The positional relationship determination module is used to determine the positional relationship between the trigger point that triggers the preset operation and the trajectory graphic in response to a preset operation triggered on the image area where the trajectory graphic is located.

[0202] A trajectory graphic update module is used to update the position of the trajectory graphic or the shape of the trajectory graphic based on the positional relationship in response to a movement operation performed on the trajectory graphic.

[0203] The second object determination unit is specifically used to determine the object located in the image area selected by the updated trajectory graphic as the second object.

[0204] Optionally, the trajectory graphics update module is specifically used to perform the following steps:

[0205] If the location point that triggers the preset operation is located on the trajectory line of the trajectory graphic, the size of the trajectory graphic is updated in response to the movement operation performed on the trajectory graphic.

[0206] If the location point that triggers the preset operation, as indicated by the positional relationship, is located within the image area bounded by the trajectory graphic, the position of the trajectory graphic is updated in response to a movement operation performed on the trajectory graphic.

[0207] Optionally, the device further includes at least one of the following modules:

[0208] The prompting module is used to output prompting information when the initial recognition information in the display interface is updated to recognition information for recognizing the second object. The prompting information includes any one of sound prompting information, vibration prompting information, or sound and light prompting information.

[0209] An initialization module is used to display guidance information when no identification information for recognizing the second object is obtained, and to display a framing page in response to a triggering operation of the guidance information; the framing page is used to instruct the re-acquisition of the target image.

[0210] Optionally, the device further includes:

[0211] An initial recognition module is used to perform object recognition on the target image;

[0212] The first display module 801 is specifically used to perform the following steps:

[0213] When multiple objects are detected, in response to a selection operation of any one of the multiple objects, the initial identification information of the first object corresponding to the selection operation and the position information of the other objects besides the first object are displayed in the display interface.

[0214] When a single object is identified, the initial identification information of the single first object is displayed on the display interface.

[0215] Optionally, the first display module 801 is also specifically used to perform the following steps:

[0216] Add a location icon to the location of the identified object, and add an information display layer to the display interface;

[0217] The intermediate identification information during the identification process of the identified object is iteratively displayed at the location icon; the intermediate identification information includes category information, name information, and identification image;

[0218] Based on the user's selection of the first object among the identified objects, the recognition result information at the end of the recognition of the first object is displayed on the information display layer; the recognition result information includes the store information and business information to which the first object belongs.

[0219] Optionally, the device further includes:

[0220] An information hiding module is used to hide the intermediate identification information when a preset trigger event on the display interface is triggered.

[0221] The re-identification module 803 is specifically used to update the identification result information in the information display layer to the identification information for identifying the second object.

[0222] Optionally, the step of identifying the second object specifically includes:

[0223] Identify the object name of the second object;

[0224] Based on the identified object name, at least one service information including the second object is determined from a preset set of services;

[0225] The business popularity corresponding to each of the at least one business information is determined, and based on the business popularity, the target business information corresponding to the second object is obtained, and the target business information is used as the identification information for identifying the second object.

[0226] This invention also provides an electronic device that may include a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor is configured to perform the image recognition method described above.

[0227] This application also provides a non-transitory computer-readable storage medium, wherein when the instructions in the storage medium are executed by a processor, the processor is able to perform an operation to implement the image recognition method described above in this application.

[0228] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0229] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, apparatus, or computer program products. Therefore, embodiments of the present invention can take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects. Furthermore, embodiments of the present invention can take the form of computer program products implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0230] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0231] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing terminal device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0232] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal equipment, causing a series of operational steps to be performed on the computer or other programmable terminal equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable terminal equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0233] Although preferred embodiments of the present invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of the present invention.

[0234] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes said element.

[0235] The above provides a detailed description of the image recognition method, apparatus, electronic device, and medium provided by the present invention. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of the present invention. 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 the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. An image recognition method, characterized in that, The method includes: The target image and initial identification information for identifying the first object in the target image are displayed on the display interface; When a preset trigger event is triggered on the display interface, in response to a swipe operation performed on the target image based on the trigger event, a second object to be re-identified indicated by the swipe operation is determined; wherein, the trigger event is used to characterize the re-indication of the object to be identified; The second object is identified, and the initial identification information in the display interface is updated to the identification information for identifying the second object; When a preset trigger event is triggered on the display interface, in response to a swipe operation performed on the target image based on the trigger event, the second object to be re-identified indicated by the swipe operation is determined, including: When a preset area icon on the display interface is triggered, in response to a swipe operation performed on the target image, the image area selected by the swipe trajectory of the swipe operation is determined in the target image; wherein, the area icon is used to indicate the object to be identified. The method further includes: identifying the object located in the image region as the second object; The trajectory graphic of the swipe path is displayed on the display interface; In response to a preset operation triggered on the image region where the trajectory graphic is located, the positional relationship between the trigger point that triggered the preset operation and the trajectory graphic is determined; In response to a movement operation performed on the trajectory graphic, the position of the trajectory graphic or the shape of the trajectory graphic is updated based on the positional relationship; Determining an object located in the image region as the second object includes: The object located in the image region framed by the updated trajectory graph is identified as the second object.

2. The method according to claim 1, characterized in that, When a preset trigger event is triggered on the display interface, in response to a swipe operation performed on the target image based on the trigger event, the second object to be re-identified indicated by the swipe operation is determined, including: When a preset location icon on the display interface is triggered, in response to a swipe operation on the location icon, the movement trace of the location icon on the target image is determined; wherein, the location icon is used to indicate the location of the object to be identified; Identify the image region where the end point of the movement trace is located in the target image; The object located in the identified image region is identified as the second object.

3. The method according to claim 1, characterized in that, Determining the image region selected by the swipe trajectory of the swipe operation in the target image includes: When the swipe trajectory is a closed trajectory, the area enclosed by the swipe trajectory is defined as the image area; When the swipe trajectory is not closed, the starting point and the ending point of the swipe trajectory are connected, and the area enclosed by the swipe trajectory after connecting the starting point and the ending point is determined as the image area.

4. The method according to claim 1, characterized in that, In response to a movement operation on the trajectory graphic, updating the position of the trajectory graphic or the shape of the trajectory graphic based on the positional relationship includes: If the location point that triggers the preset operation is located on the trajectory line of the trajectory graphic, the size of the trajectory graphic is updated in response to the movement operation performed on the trajectory graphic. If the location point that triggers the preset operation, as indicated by the positional relationship, is located within the image area bounded by the trajectory graphic, the position of the trajectory graphic is updated in response to a movement operation performed on the trajectory graphic.

5. The method according to claim 1, characterized in that, The identification of the second object includes: Identify the object name of the second object; Based on the identified object name, at least one service information including the second object is determined from a preset set of services; The business popularity corresponding to each of the at least one business information is determined, and based on the business popularity, the target business information corresponding to the second object is obtained, and the target business information is used as the identification information for identifying the second object.

6. The method according to claim 1, characterized in that, The method further includes at least one of the following: When the initial recognition information in the display interface is updated to recognition information for recognizing the second object, a prompt message is output, which includes any one of sound prompt information, vibration prompt information, or sound and light prompt information; If no identification information for recognizing the second object is obtained, guidance information is displayed, and a viewfinder page is displayed in response to a triggering operation of the guidance information; the viewfinder page is used to instruct the re-acquisition of the target image.

7. The method according to claim 1, characterized in that, The method further includes: Perform object recognition on the target image; The initial identification information for identifying the first object in the target image is displayed on the display interface, including: When multiple objects are detected, in response to a selection operation of any one of the multiple objects, the initial identification information of the first object corresponding to the selection operation and the position information of the other objects besides the first object are displayed in the display interface. When a single object is identified, the initial identification information of the single first object is displayed on the display interface.

8. According to claim 1 7. The method described in any one of the following ways is characterized in that initial identification information for identifying the first object in the target image is displayed on the display interface, including: Add a location icon to the location of the identified object, and add an information display layer to the display interface; The intermediate identification information during the identification process of the identified object is iteratively displayed at the location icon; the intermediate identification information includes category information, name information, and identification image; Based on the user's selection of the first object among the identified objects, the recognition result information at the end of the recognition of the first object is displayed on the information display layer; the recognition result information includes the store information and business information to which the first object belongs.

9. The method according to claim 8, characterized in that, The method further includes: When a preset trigger event on the display interface is triggered, the intermediate identification information is hidden; The step of updating the initial recognition information in the display interface to recognition information for recognizing the second object includes: The recognition result information in the information display layer is updated to the recognition information for recognizing the second object.

10. An image recognition device, characterized in that, The device includes: The first display module is used to display the target image and initial identification information for identifying the first object in the target image in the display interface; A response module is configured to, when a preset trigger event is triggered on the display interface, respond to a swipe operation performed on the target image based on the trigger event, and determine the second object to be re-identified indicated by the swipe operation; The triggering event is used to characterize the re-indication of the object to be identified; A re-identification module is used to identify the second object and update the initial identification information in the display interface with the identification information for identifying the second object; When a preset trigger event is triggered on the display interface, in response to a swipe operation performed on the target image based on the trigger event, the second object to be re-identified indicated by the swipe operation is determined, including: When a preset area icon on the display interface is triggered, in response to a swipe operation performed on the target image, the image area selected by the swipe trajectory of the swipe operation is determined in the target image; wherein, the area icon is used to indicate the object to be identified. The step of identifying the object located in the image region as the second object includes: The trajectory graphic of the swipe path is displayed on the display interface; In response to a preset operation triggered on the image region where the trajectory graphic is located, the positional relationship between the trigger point that triggered the preset operation and the trajectory graphic is determined; In response to a movement operation performed on the trajectory graphic, the position of the trajectory graphic or the shape of the trajectory graphic is updated based on the positional relationship; Determining an object located in the image region as the second object includes: The object located in the image region framed by the updated trajectory graph is identified as the second object.

11. An electronic device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executed, implements claim 1. The image recognition method described in any one of the nine claims.

12. A computer-readable storage medium, characterized in that, Its stored computer program causes the processor to execute as claimed in claim 1. The image recognition method described in any one of the nine claims.

Images