Motion information display method and electronic device

Abstract

The application discloses a kind of operation information display method and electronic equipment.Method includes: in response to function trigger operation, show motion recognition page;Wherein, motion recognition page includes target object real-time image information;If the motion attribute information of target object in target object real-time image information is identified to meet preset condition, then show target object real-time image information and the motion attribute information of target object in motion recognition page.The application realizes the same screen partition display of motion image and motion information, and motion image information is used for target object to observe own action posture and calibrate in real time, and motion attribute information synchronously shows relevant information such as motion time, motion type, target object can obtain visual feedback and data feedback simultaneously without switching interface, improve the coherence of training process, realize the adaptive update of interface state, simplify operation path, improve man-machine interaction frequency and user experience.

Description

Technical Field

[0001] This invention relates to the field of interface interaction, and more particularly to a method for displaying motion information and an electronic device. Background Technology

[0002] Existing fitness software often features simplistic user interfaces. Motion recognition information is typically presented as text superimposed in the corner of the camera preview, requiring users to frequently switch focus between observing their own movements and viewing the recognition data. There's a lack of automatic recognition and interface feedback mechanisms when switching training states; users must manually click a button to signal the software to enter rest mode after completing a set of exercises, a cumbersome process that disrupts training continuity. Adding or correcting data often involves jumping to a new page, requiring users to leave the current recognition interface, edit, and then return – abrupt page transitions resulting in a poor training experience. Furthermore, viewing historical sets of data for the entire training session usually requires exiting training mode, preventing quick browsing without interrupting the workout. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to provide a method and electronic device for displaying sports information, so as to improve the readability of data display during fitness.

[0004] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: A method for displaying motion information, the method comprising: In response to a function trigger operation, a motion recognition page is displayed; wherein, the cloud recognition page includes real-time image information of the target object; If the motion attribute information of the target object in the real-time image information of the target object meets the preset conditions, the real-time image information of the target object and the motion attribute information of the target object will be displayed on the motion recognition page.

[0005] To solve the above-mentioned technical problems, another technical solution adopted by the present invention is as follows: An electronic device includes a camera, a display screen, a memory, and a processor. The memory stores a computer program, and when the processor executes the computer program, it controls the electronic device to implement the various steps in the above-described motion information display method.

[0006] The beneficial effects of this invention are as follows: By responding to function trigger operations to display a motion recognition page, and when the recognized motion attribute information meets preset conditions, the real-time image information and motion attribute information of the target object are simultaneously displayed on the motion recognition page. This achieves simultaneous display of motion images and motion information on the same screen, eliminating the need for users to frequently switch visual focus between observing actions and viewing data, thus reducing visual jumps and cognitive burden. Simultaneously, this invention automatically completes the simultaneous presentation of key information when conditions are met, allowing users to simultaneously obtain visual and data feedback without switching interfaces. The automatic display of motion attribute information lays the foundation for subsequent intelligent functions such as asynchronous recognition and automatic switching to resting states. It solves the problem of training continuity interruption caused by frequent page jumps, improves the immersion and operational efficiency of the training process, simplifies the user operation path, lowers the usage threshold, and improves the human-computer interaction experience. Attached Figure Description

[0007] Figure 1 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention; Figure 2 A flowchart illustrating the steps of a motion information display method provided in an embodiment of the present invention; Figure 3 A complete interactive flowchart of a motion information display method provided in an embodiment of the present invention, from entering the software to saving training records; Figure 4 A schematic diagram of the interface of a motion information display method in the initial recognition state according to an embodiment of the present invention; Figure 5 This is a schematic diagram of the interface of a motion information display method provided in an embodiment of the present invention in a scenario where the motion state of a target object is identified; Figure 6 This is a schematic diagram of the interface of a motion information display method provided in an embodiment of the present invention when a target object is identified as entering a resting state. Detailed Implementation

[0008] Definitions: Table 1

[0009] To make the technical problems, technical solutions, and beneficial effects to be solved by this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and are not intended to limit the scope of this application.

[0010] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0011] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0012] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0013] In related technologies, fitness software has a relatively simple interface interaction method. Exercise recognition information is usually simply superimposed on the corner of the camera preview screen in the form of text. Users need to frequently switch visual focus between observing their own movements and viewing the recognition data, resulting in a poor training experience.

[0014] For example, during squat training, users need to simultaneously observe their own posture on the phone screen to ensure it's correct, while also checking the motion recognition results and repetition statistics displayed in the corner of the screen. Because the display area for the recognition information is too small and its position is fixed, users often need to get close to the screen or shift their gaze to see the data clearly, leading to distorted movements or distraction. Furthermore, such software lacks automatic recognition and interface feedback mechanisms when switching training states. After completing a set of exercises, users must manually tap the software to indicate that it's entering rest mode, which is cumbersome and disrupts the continuity of training.

[0015] Furthermore, the relevant technologies use a method of redirecting to a new page to allow users to supplement or correct training data. Users need to leave the current recognition interface, complete the editing, and then return. The page jump is abrupt, causing users to need to reposition their previous training state, which further exacerbates the sense of interruption in the training process and the complexity of the operation.

[0016] To address the aforementioned problems, this application provides a method and electronic device for displaying motion information. The method for displaying motion information according to this application is described in detail below.

[0017] The exercise information display method in this application can be used in fitness software. The electronic device in this application can be a terminal with a camera and a display screen, such as a smartphone or other mobile terminal, a computer, a smart interactive device, a smart home appliance, a holographic display device, etc.

[0018] Please refer to Figure 1 The present invention also provides an electronic device 600, including a camera 601, a display screen 602, a memory 603 and a processor 604, as well as a computer program stored in the memory 601 and running on the processor 603. When the processor 604 executes the computer program, it implements the various steps in the motion information display method described above.

[0019] The following describes in detail a motion information display method of the present invention, with reference to the appendix. Figure 2 This includes steps 110 and 120.

[0020] Step 110: In response to the function trigger operation, display the motion recognition page; wherein, the motion recognition page includes real-time image information of the target object.

[0021] The function trigger operation refers to the user's action of activating the automatic recognition function of the fitness software, which is used to trigger the system to display the exercise recognition page. In a fitness training scenario, this function trigger operation can be the user clicking the function entry button in the application interface.

[0022] A motion recognition page is an interactive interface used to display real-time motion image information and motion attribute information of a target object, assisting users in recording offline training data.

[0023] For example, when a user enters the tab of a fitness app and clicks on the app's automatic exercise recognition function, the system responds to this trigger by displaying an exercise recognition page. This page contains real-time image information of the user, such as training footage captured by the phone's front or rear camera. The "function trigger operation" corresponds to the aforementioned "responding to function trigger operation," meaning the user actively clicks to trigger the system to display the exercise recognition page.

[0024] Step 120: If the motion attribute information of the target object in the real-time image information of the target object meets the preset conditions, the real-time image information and motion attribute information of the target object will be displayed on the motion recognition page.

[0025] Motion attribute information refers to data information associated with the current motion state of the target object, including at least one of the following: motion item identifier, number of times the motion item has been completed, duration of the completed motion item, and motion item parameter information.

[0026] Preset conditions refer to the judgment criteria set in advance by the system to determine whether a valid motion action has been recognized. When the system recognizes that the motion attribute information in the real-time image information of the target object meets the preset conditions, it indicates that the system has detected a valid motion action. At this time, the real-time image information of the target object and the corresponding motion attribute information are displayed on the motion recognition page at the same time, so that users can view their own motion posture and training data in real time.

[0027] For example, the system calls the motion recognition SDK (Software Development Kit) to perform frame-by-frame recognition of real-time motion images of the target object. The SDK first detects the human body in the real-time motion images and identifies the target object's limb posture. After continuously analyzing multiple frames, if the SDK detects that the target object's movement matches a preset motion pattern, it is determined to have identified a valid motion action. At this point, the system determines the action name corresponding to the valid motion action and begins accumulating the number of movements. In response to this recognition result, the system simultaneously displays real-time image information of the target object (i.e., real-time motion footage of the user captured by the camera, allowing the user to observe their own posture) and motion attribute information (including the identified action name, the cumulative number of completed movements, and the training time calculated from the first human body recognition) on the motion recognition page.

[0028] This approach restructures the traditional information recording model, which involves manual operation by the target user and multiple page jumps, into a parallel interaction model where the system automatically identifies and displays the data on the same screen, fundamentally optimizing the process. By using a function trigger as the entry point, the system automatically displays a motion recognition page upon response. When a valid motion is detected, the system independently completes the entire process from image acquisition and motion attribute recognition to the simultaneous display of real-time motion images and attribute information. This reduces the interaction overhead previously scattered across multiple page jumps and manual operations into two main interactions: a single function trigger and a single simultaneous display. This decreases user operation frequency and interface switching latency, effectively improving the overall interaction efficiency and system resource utilization of the fitness software while maintaining compatibility with user training habits, and further enhancing the consistency and accuracy of training data recording.

[0029] In one embodiment of this application, the sports attribute information includes a sports item identifier, and step 120 includes steps 121 and 122.

[0030] Step 121: If the motion item identifier of the target object in the real-time image information of the target object belongs to the preset motion item set, then the real-time image information of the target object and the motion attribute information of the target object are displayed on the motion recognition page.

[0031] The sport identifier refers to the name and / or identifier of the specific sport that the target object is currently performing, which is identified and used by the system. It can be automatically identified and determined by the system or manually selected and modified by the target object.

[0032] The preset set of exercises refers to a pre-configured set of recognizable exercises, such as common training movements like squats, bench presses, deadlifts, lat pulldowns, and bent-over rows. When the exercise identifier recognized by the system belongs to this preset set, it indicates that the system has successfully identified the type of movement the target object is performing. At this point, the motion recognition page simultaneously displays the target object's real-time image information and corresponding motion attribute information, allowing the user to confirm the recognition result and view training data in real time.

[0033] For example, the system calls the motion recognition SDK (Software Development Kit) to recognize the real-time motion image of the target object. When the system recognizes that the action being performed by the target object belongs to a preset set of sports, such as squats, bench presses, deadlifts, etc., the system simultaneously displays real-time motion image information and motion attribute information on the motion recognition page. The motion attribute information includes the name of the sports and the number of times it has been completed.

[0034] Step 122: The sports attribute information includes at least one of the following: sports item identifier, number of times the sports item has been completed, duration of the sports item completion, and preset sports item parameter information.

[0035] The exercise identifier refers to the name and / or identifier of the specific exercise currently being performed by the target object, such as "squat," "bench press," "deadlift," and "lat pulldown." The number of completed exercises refers to the number of valid movements completed by the target object in the current set, as counted in real-time by the system through exercise recognition. The duration of completed exercises refers to the cumulative duration of continuous exercise performed by the target object during this training session, as counted in real-time by the system through exercise recognition, calculated from the first time the target object's body is detected. Exercise parameter information refers to the quantitative training parameters related to the exercise, used to fully record the training data for each set, such as the weight and target number of repetitions. These exercise attribute information collectively constitute a complete record of training data, categorized and displayed on the exercise recognition page so that users can fully understand their current training status.

[0036] For example, when the system detects that a target is performing a squat, the motion recognition page displays the following motion attribute information: the motion name is "squat," the number of repetitions completed is "8," the duration of the exercise is "12 seconds," and the motion parameter information includes the number of repetitions completed is "10" and the weight is "50kg." Specifically, the motion identifier "squat" is displayed in the motion prompt box, the number of repetitions completed ("8") is displayed in larger font in the center of the page, the duration of the exercise ("12 seconds") is displayed in the total training time display area, the weight ("50kg") is displayed in the weight input box in the editable interactive area, and the target number of repetitions ("10") is provided as a training goal for the user's reference.

[0037] In this way, by refining exercise attribute information into multiple dimensions such as exercise identifier, number of times the exercise has been completed, duration of the exercise, and preset exercise parameter information, the system can automatically determine and display the corresponding complete exercise attribute information based on whether the identified exercise identifier belongs to a preset set of exercises. This application achieves a structured display of multi-dimensional exercise attribute information, allowing users to comprehensively understand all the key data of the current training movement on a single page. Simultaneously, through the filtering mechanism of the preset set of exercises, the system can effectively filter out non-fitness movements or invalid exercises, avoiding irrelevant information from interfering with the user, and improving the accuracy of information display and user experience on the exercise recognition page.

[0038] In one embodiment of this application, steps 130 and 131 are also included.

[0039] Step 130: If the motion attribute information of the target object in the real-time image information of the target object does not meet the preset conditions, then display the candidate motion item identifier in the editable area of ​​the motion recognition page, and display the number of times the candidate motion item has been completed or the parameter information of the candidate motion item corresponding to the candidate motion item identifier.

[0040] Preset conditions refer to the judgment criteria pre-defined by the system to determine whether a valid movement is recognized and whether the movement identifier can be determined. If, after analyzing the real-time image information of the target object, the system cannot recognize a valid movement, or the recognized movement identifier does not belong to the preset movement set, it is considered that the preset conditions are not met. In this case, the system cannot automatically determine the name and related parameters of the movement being performed by the target object. Therefore, candidate movement identifiers and their corresponding parameter information need to be displayed in the editable area of ​​the movement recognition page for the target object to select and configure.

[0041] Editable area refers to the interactive area on the motion recognition page where the target object can edit and correct information, such as a half-pull page or a drop-down menu. Candidate motion item identifier refers to the alternative motion names selected by the system from a preset set of motion items that are similar to the current recognition result or are available for the target object to choose from.

[0042] For example, the system might attempt to recognize real-time motion images but fail to identify a target's current action as belonging to a preset set of exercises. For instance, if the user's current action is not within the system's preset range of exercises like squats, bench presses, or deadlifts, the system's recognition fails. In this case, the system displays a list of candidate exercise identifiers in the editable area of ​​the motion recognition page. For example, a dropdown menu might show candidate exercises such as "bent-over rows," "pull-ups," "dips," "seated rows," and "lat pulldowns." It also displays the number of repetitions or parameter information for each candidate exercise. The user can then select the correct exercise based on their actual training content and manually configure the exercise attributes. For example, if the user is performing a "Bulgarian split squat," which is not within the system's preset range of common exercises, the system determines that the recognition result does not meet the preset conditions; that is, the system cannot automatically determine the name of the target's current action and its related parameters. In response to the failed recognition result, the system displays a list of candidate exercise identifiers in the editable area of ​​the exercise recognition page. These include candidate movements similar to or common to the current recognition result, such as "bent-over row," "pull-up," "dips," "seated row," and "lat pulldown." Simultaneously, the system displays the number of times each candidate exercise has been completed or its parameter information. The user then selects "bent-over row" from the candidate list based on their actual training content to manually configure the exercise attributes.

[0043] Step 131: In response to the editing operation on the editable area, display the editing result of at least one of the following in the editable area of ​​the motion recognition page: candidate motion item identifier, number of times the candidate motion item corresponding to the candidate motion item identifier has been completed, or candidate motion item parameter information.

[0044] Editing operations refer to interactive behaviors in which the target object modifies or confirms motion attribute information in the editable area, including clicking to select candidate sports item icons, manually modifying the number of completed times or weight parameters, etc.

[0045] Editable areas refer to interactive areas on a motion recognition page where the target object can edit and correct information, such as half-slide-out pages or drop-down menus.

[0046] The editing result refers to the updated motion attribute information displayed in real-time in the editable area after the target object completes the editing operation. This includes the modified candidate motion item identifier, the adjusted number of completed attempts, or the changed parameter information. The system responds to the target object's editing operation by updating and displaying the editing result in real-time, allowing the target object to confirm the corrections. For example, if a user clicks on the candidate exercise icon "Bendover Row" in the editable area, the system updates the exercise name to the exercise recognition page. Since the system hasn't recognized a valid exercise, it also can't determine the number of times the exercise has been completed. In this case, the editable area provides the user with a recommended number of completed exercises or a target number of completed exercises (e.g., the target number of times for the exercise) for the user to choose from. Users can also manually modify the number of completed exercises in the editable area, for example, changing the system-recognized "5 times" to the actual "8 times," or modifying exercise parameters, such as changing "40kg" to "50kg." The system displays the modified results in real-time in the editable area. The system responds to every user editing operation, displaying the modified results in real-time in the editable area for user confirmation.

[0047] In this way, by proactively displaying an editable area and candidate exercise identifiers when the motion attribute information does not meet preset conditions, this application provides a convenient manual correction entry point for situations where the system fails to identify or is inaccurately identifying motions. This application proactively recommends candidate exercises to users through the editable area, reducing the operational cost of manual input. Simultaneously, users can easily correct the motion name, number of repetitions, and parameter information through simple clicks and value modifications. The system displays the editing results in real time, enabling rapid correction in scenarios of recognition failure, improving the accuracy of motion data recording and the user experience.

[0048] In one embodiment of this application, if the sports item identifier includes a multi-sided sports mode identifier, then modifying the sports attribute information in response to the editing result of the editable area includes steps 132 and 133.

[0049] Step 132: In response to the switching operation of multi-side parameter information in the sports parameter information, switch the sports side label displayed in the editable area.

[0050] Multilateral parameter information refers to parameter information associated with multilateral movement pattern identifiers, including movement side identifiers that need to be distinguished and counted separately for each side of the body (e.g., "L" for left side and "R" for right side) and the corresponding movement attribute information for each side.

[0051] The switching operation refers to the interactive behavior of changing the motion side label of the target object in the editable area, such as clicking the motion side label switching button.

[0052] The movement side identifier is a marker used to identify which side of the body the target object uses for training in multi-sided movements. For example, "L" represents the left side and "R" represents the right side. In response to the target object's switching operation, the system switches the currently displayed movement side identifier in the editable area to the identifier of the other side and simultaneously updates the corresponding movement attribute information of that side, so that the target object can view and edit the training data of different sides separately.

[0053] For example, if a user is performing dumbbell curls, the system identifies the exercise as "dumbbell curls" and recognizes it as a multi-sided movement, requiring differentiation between the left and right sides. The editable area of ​​the exercise recognition page currently displays the side indicator "L (left)" and its corresponding attributes: 10 repetitions completed and 15kg weight. After completing the left-side workout, the user can switch the side indicator in the editable area (an arrow icon or left / right switch control next to the side indicator). The system responds by changing the side indicator from "L (left)" to "R (right)" and updating the number of repetitions from "10" on the left to "8" on the right (or "0" if the right side has just started training). The weight is also updated to "15kg" (unless both sides have the same weight; otherwise, it's updated to the user's preset weight). The user can then view the right-side training data and edit it based on their progress.

[0054] Step 133: The multi-side parameter information includes motion side identifiers of at least two sides with different motion directions.

[0055] Multi-sided parameter information refers to parameter information associated with multi-sided motion mode identifiers, including motion side identifiers of at least two sides with different motion directions and corresponding motion attribute information for each side.

[0056] Different directions of movement refer to differences in the side of the body exerting force or performing the movement during training, such as the left and right sides.

[0057] Movement side identifiers refer to markers used to identify which side of the body a target object uses for training in multi-sided movements. The system uses multi-sided parameter information to achieve refined data management for training movements that require differentiation between left and right sides. At the underlying level, training data for each side is stored separately, while at the front end, the target object can switch between viewing and editing the movement attribute information for different sides.

[0058] For example, for exercises that require distinguishing between left and right sides, such as dumbbell curls, dumbbell presses, and single-leg squats, multi-side parameter information includes two movement side labels: "L (left)" and "R (right)". The system retains separate training data for the left and right sides at the lowest level, including the number of repetitions completed on the left and right sides, the weight parameters for the left and right sides, the completion time for the left and right sides, etc. In the editable area (half-page) of the exercise recognition page, the system has a movement side label switching button, allowing users to switch and view / edit the exercise attribute information for different sides. For example, when the user clicks the "L" label, the editable area displays the training data for the left side (10 repetitions, 15kg); when the user clicks the "R" label, the editable area switches to display the training data for the right side (8 repetitions, 15kg). Users can edit and adjust the data according to the actual completion status of each side. In the training details pull-up page, the system marks "L" or "R" after the group number. For example, "Group 1 L" means the first group on the left and "Group 1 R" means the first group on the right, so that the target can clearly distinguish the training records on the left and right sides respectively.

[0059] In this way, by providing a side label switching function for multi-sided movement identification, this application achieves refined data recording for training movements that require differentiation between the left and right sides. Through the movement side label switching mechanism in the editable area, this application allows users to view, edit, and confirm parameters such as the number of repetitions and movement parameters for both the left and right sides separately. Simultaneously, the system stores training data for both the left and right sides at the underlying level, providing more accurate unilateral data support for subsequent data statistics, muscle fatigue calculation, and training analysis. This meets users' personalized recording needs in unilateral training scenarios and improves the precision and practicality of movement data recording.

[0060] In one embodiment of this application, the sports event parameter information includes the target number of repetitions for the sports event.

[0061] Step 135: In response to the received training termination operation from the target object, the system obtains the number of times the current exercise has been completed and determines whether the number of times the exercise has been completed is less than the target number of times for the exercise. If the determination result is yes, it means that the user has not completed the preset training goal. The system displays candidate exercise icons in the editable area on the exercise recognition page, as well as the number of times the candidate exercise has been completed or the parameter information of the candidate exercise corresponding to the candidate exercise icon.

[0062] Stop training refers to the interactive behavior of the target object actively pausing or ending the current training task during the training process, such as clicking the pause button, exit button or selecting the "end exercise" option on the motion recognition page.

[0063] The target number of repetitions for a sport refers to the preset expected number of times the target subject will complete the current sport, used to measure the achievement of training goals. When the system receives a stop training operation, it automatically obtains the number of repetitions already completed for the current sport and compares it with the preset target number of repetitions. If the number of repetitions completed is less than the target number, it indicates that the target subject has not achieved the preset training goal. At this time, the system actively displays candidate sport icons and corresponding parameter information in the editable area of ​​the motion recognition page, guiding the target subject to switch to other training movements to continue training, avoiding incomplete training records.

[0064] For example, before training begins, the user sets a target of 12 repetitions for the "bench press" exercise on the motion recognition page. During training, the system records and accumulates the number of repetitions completed in each set. After completing the 8th repetition (possibly 2 sets of 4 repetitions each, or 1 set of 8 repetitions), the user clicks the "Stop Training" button (such as a pause or exit button). In response, the system immediately retrieves the current number of repetitions completed (8) and compares it to the preset target of 12. The system determines that 8 is less than 12, indicating that the user has not yet reached the preset target for this training session (4 more repetitions needed). Instead of immediately ending the training, the system actively displays a list of candidate exercise icons in the editable area (half-page) of the motion recognition page, such as "incline bench press" (4 repetitions remaining), "decline bench press" (4 repetitions remaining), and "dumbbell bench press" (4 repetitions remaining), all belonging to the same pushing motion category as the bench press. Meanwhile, the system displays recommended parameters for each candidate exercise, such as recommended number of repetitions "4" (to make up the target difference) and recommended weight "45kg" (similar to or slightly lower than the bench press weight). Users can select one of the candidate exercises to continue training based on their own situation, and the system will switch to that exercise and retain the progress of the target number of repetitions already completed.

[0065] For example, a user sets a target number of repetitions for the exercise "squats" on the exercise recognition page to "15". When the user completes "15" repetitions and clicks the stop training button, the system determines that the number of repetitions completed equals the target number. Therefore, it does not trigger the candidate display in the editable area, directly ends the current training, and saves the record.

[0066] In this way, by determining whether the number of repetitions of an exercise has reached the target number when the user stops training, this application can proactively provide recommendations and guidance for candidate exercises when the user has not completed the training goal. This application identifies the user's training completion status through a judgment mechanism and proactively displays an editable area and candidate exercises when the target has not been met, providing the user with flexible training adjustment options. Users can choose to switch to other training exercises to continue training, avoiding incomplete records caused by directly ending training due to the inability to continue the current exercise, thus improving the completeness of training data and the flexibility of user training.

[0067] In one embodiment of this application, the motion recognition page further includes a motion attribute information switching control. Step 120, which displays real-time image information of the target object and motion attribute information of the target object on the motion recognition page, includes steps 123 and 124.

[0068] Step 123: In response to the trigger operation of the motion attribute information switching control, display the motion attribute information switching list, which includes a list of motion item identifiers, and the list of motion item identifiers includes at least two different candidate motion item identifiers.

[0069] A motion attribute information switching control refers to an interactive element used to trigger the display of a motion attribute information switching list, allowing the target object to switch the current motion item identifier. In one implementation, this switching control can be represented as a switch action button to the right of the action prompt box on the motion recognition page; in another implementation, the aforementioned editable area itself can also serve as a switching control, and the target object can trigger the switching operation by clicking the action name in the editable area.

[0070] The sports attribute information toggle list refers to the interactive interface element displayed in response to the triggering operation of the toggle control, which is used to display a list of candidate sports items to be selected.

[0071] The sports item identifier list refers to the set of at least two different candidate sports item identifiers contained in the toggle list, from which the target object can select the correct action name. In response to the target object's triggering of the toggle control, the system displays this toggle list on the motion recognition page, allowing the target object to quickly correct incorrectly identified action names.

[0072] For example, an icon button is located on the right side of the action prompt box on the motion recognition page; this button serves as the motion attribute information switching control. Furthermore, the editable area (half-page) mentioned in step 130 can also function as a switching control; clicking the currently displayed action name in the editable area triggers the switching operation. When the system recognizes an incorrect action name (for example, the system identifies the user's actual "bent-over row" as "bench press"), the user clicks the switch action button on the right side of the action prompt box (or clicks the action name "bench press" in the editable area). In response to this trigger, the system displays a motion attribute information switching list on the motion recognition page. This list is presented as a pop-up or drop-down card, floating below the action prompt box or above the editable area. The switching list contains a list of exercise identifiers, such as "squat," "bench press," "deadlift," "bent-over row," "pull-up," and "lat pulldown," among other different candidate exercise identifiers. The display height of the toggle list adaptively adjusts based on the number of candidate exercise icons: when there are many candidate exercises, the list height is up to 5.5 cards, allowing users to scroll through all options; when there are few candidate exercises, the list height automatically shrinks to maintain an aesthetically pleasing interface and ease of use. When a user selects "Bent-over Row" as the correct exercise name from the list, the system responds by updating the exercise name in the exercise prompt box from "Bench Press" to "Bent-over Row".

[0073] Step 124: In response to the selection operation of the target sport identifier in the list of sport identifiers, update the sport identifier displayed on the sport identification page to the target sport identifier; the target sport identifier is a candidate sport identifier in the list of sport identifiers.

[0074] The selection operation refers to the interactive behavior of the target object clicking or selecting a candidate sports item icon in the motion attribute information switching list.

[0075] The target item identifier refers to the candidate sport identifier selected by the target object from the list of sport identifiers, which the object wishes to update on the sport recognition page. In response to the target object's selection, the system updates the currently displayed sport identifier on the sport recognition page to the target item identifier, allowing the target object to correct system recognition errors or supplement action names not recognized by the system. In one implementation, the system also records the action manually selected by the target object as the highest priority. During subsequent group recognition, as long as the category recognized on the device is the same as the current group, the system directly uses the manually selected action without waiting for cloud-based recognition results, thereby improving the efficiency and consistency of subsequent group action recognition.

[0076] For example, on the current motion recognition page, the motion prompt box displays the exercise icon as "bench press," but the user is actually performing "bent-over row" (both belong to the "pulling exercises" category, but the specific movements are different). After discovering the system's error, the user clicks the motion attribute information switching control (the icon button to the right of the motion prompt box or the action name in the editable area). The system then displays a list of motion icons, including multiple candidate icons such as "squat," "bench press," "deadlift," "bent-over row," "pull-up," and "lat pulldown." The user selects "bent-over row" as the target exercise icon from the list. In response to this selection, the system immediately updates the "bench press" displayed in the motion prompt box on the motion recognition page to the user-selected "bent-over row." Simultaneously, the system records this manually selected action as the highest priority (P0 level). In subsequent group recognitions, as long as the category recognized on the device is the same as the current group, the system directly uses the manually selected "bent-over row" action without waiting for the cloud-based VLM return result, thus maintaining interface consistency and improving recognition efficiency. The system will also automatically adjust the relevant fields of subsequent training data records based on the updated exercise name "bent-over row".

[0077] In this way, by setting a motion attribute information switching control on the motion recognition page, this application provides users with a convenient manual correction entry point. When the action recognized by the system does not match the actual action, the user does not need to exit the current page or enter multiple menus; they can simply click the switching control and select the correct action from the list to complete the correction. This application achieves quick action switching within the same page through the pop-up interaction of the switching list, reducing the interruption caused by interface jumps. At the same time, the system records the action manually selected by the user as the highest priority, facilitating the inheritance of action recognition in subsequent groups, thus improving the efficiency and accuracy of action correction.

[0078] In one embodiment of this application, step 140 is also included.

[0079] Step 140: If the motion attribute information of the target object in the real-time image information does not meet the preset conditions, a prompt message will be displayed on the motion recognition page; the prompt message is used to prompt the user to edit the motion attribute information.

[0080] Preset conditions refer to the judgment criteria set in advance by the system to determine whether a valid movement has been identified and whether the movement item can be identified. If the system analyzes the real-time image information of the target object and has not yet returned any recognition results, or if the recognition results cannot match any item in the preset movement item set, it is considered that the preset conditions are not met.

[0081] Recognition prompts refer to text prompts displayed by the system when recognition fails or occurs during the recognition process. These prompts inform the target object of its current recognition status and guide it to make manual edits. This information aims to reduce user confusion and anxiety during waiting, guiding users to manually input or correct motion attribute information through the editable area, ensuring the accuracy and completeness of training data recording.

[0082] For example, the system calls the motion recognition SDK to analyze real-time motion images of the target object frame by frame. At the start of training, the SDK has not yet returned any recognition results (it is analyzing). At this time, the system displays the status text "Recognizing..." in the action prompt box on the motion recognition page, indicating that the system is recognizing and prompting the user to wait. After a period of time, if the SDK still cannot return recognition results, or if the recognized action cannot match any item in the preset set of sports (for example, the user is performing a "Bulgarian split squat," which is not within the system's preset range of common actions such as squats, bench presses, deadlifts, and bent-over rows), the system determines that the recognition has failed. At this time, the system displays a recognition prompt message in the action prompt box, such as "No action recognized, please select manually" or "Recognizing, please confirm the action name," to guide the user to manually input or correct the motion attribute information through the editable area (half-page scroll or drop-down menu). According to the prompt message, the user clicks on the candidate sports item list in the editable area, selects the action they are currently performing, and manually fills in the number of repetitions and weight parameters to complete the manual configuration of the motion attribute information.

[0083] In this way, by proactively displaying recognition prompts when motion attribute information does not meet preset conditions, this application can promptly provide feedback to the user on the current recognition status and guide the user to manually intervene when the system fails to recognize or is inaccurate. This application reduces user confusion and anxiety by clearly informing the user of the system status through clear prompts. Simultaneously, the prompts guide the user to edit and correct the motion attribute information through the editable area, organically combining system recognition with manual correction. This ensures the accuracy and completeness of training data recording and improves the user experience in scenarios with recognition anomalies.

[0084] In one embodiment of this application, the real-time image information of the target object includes real-time image information; or, the real-time image information of the target object includes real-time image information and skeletal point information; wherein, when the real-time image information of the target object includes real-time image information and skeletal point information corresponding to the target object in the real-time image information, the display of the motion recognition page includes steps 150 and 151.

[0085] Step 150: After acquiring real-time image information, determine at least one of the following in motion attribute information based on the recognition results of at least one motion frame or motion frame sequence in the real-time image information: motion item identifier, number of times the motion item has been completed, duration of the motion item completion, motion item parameter information, and skeletal point information.

[0086] A motion frame refers to a single frame in a continuous moving image captured by a camera.

[0087] A motion frame sequence refers to an image sequence composed of multiple consecutive motion frames, used to analyze and identify the motion posture and movement changes of a target object. After the system acquires real-time image information of the target object through a camera, it calls the motion recognition SDK to analyze and identify at least one motion frame or motion frame sequence. Based on the recognition results, the system can determine one or more of the following motion attribute information: motion item identifier (e.g., action name), number of times the motion item has been completed (e.g., the number of valid movements completed), duration of the motion item (e.g., the cumulative duration from the first recognition to the start of the human body), motion item parameter information (e.g., weight, target number of repetitions, etc.), and skeletal point information (e.g., joint points and key points of the limbs and torso). The above information together constitutes a complete training data record, which is used for categorized display on the motion recognition page.

[0088] For example, the system continuously captures real-time motion images of the user through the phone's camera, acquiring multiple motion frames and motion frame sequences at a rate of 30 frames per second. The SDK analyzes each frame in real time, detecting human posture and movement trajectory. When the SDK recognizes the first positive sample (i.e., detects a valid motion action), such as detecting a downward shift of the user's center of gravity and a knee flexion angle reaching the threshold for a squat, the system, based on the recognition results of that motion frame and the preceding and following continuous motion frame sequences, identifies the current motion as a "squat." Simultaneously, the system begins counting the number of times the motion has been completed: the first valid squat is recorded as "1 time," the second as "2 times," the third as "3 times," and so on, accumulating progressively. The system also starts timing from the first time the user is detected, recording the duration of the completed motion as "12 seconds." The system determines the motion parameters, including the target number of repetitions (user-preset "10 times") and the weight (user-entered "50 kg"). In addition, the SDK also identifies the user's skeletal information, including the locations of joints such as the shoulder, elbow, knee, and hip joints, as well as key points of the limbs and torso such as the wrist, ankle, and spine.

[0089] Step 151: Display real-time image information. If skeletal point information is identified, overlay the skeletal point information onto the target object in the real-time image information.

[0090] Skeletal point information refers to the location data of joints and key points of the limbs and torso of the target object, identified after analyzing real-time image information of the target object through the motion recognition SDK. Overlay display refers to displaying the skeletal point information as visual markers (such as dots, lines, or skeleton diagrams) overlaid on the corresponding positions of the target object in the real-time image information. The system displays real-time image information of the target object captured by the camera on the motion recognition page. If skeletal point information is identified, it is overlaid on the target object in the real-time image information, allowing the target object to observe its own movement posture in real time, correct itself against standard postures, and improve the standardization of training movements.

[0091] For example, the system displays real-time footage of the user's movements captured by the camera on the motion recognition page, showing the user performing squat training. Once the system calls the motion recognition SDK to analyze the real-time image and identify the human body, it overlays human skeletal points onto the real-time motion image on the display screen. This includes the user's joints (such as shoulder, elbow, knee, and hip joints) and skeletal points of the limbs and torso (such as wrists, ankles, and spine). These skeletal points are displayed as colored dots or linear skeletons overlaid on the corresponding positions on the user's body, allowing the user to visually see if their posture is correct. For example, during a squat, the user can observe whether the knee joint skeletal points are above the toe skeletal points, whether the back skeletal points are straight, and whether the hip joint skeletal points are below the knee joint skeletal points. Before the SDK returns the recognition results, the motion recognition page does not display the number of repetitions; instead, it displays the status text "Recognizing..." in the action prompt box, indicating that the system is recognizing the user and asking the user to wait.

[0092] In this way, by acquiring real-time image information and recognizing it based on motion frames or motion frame sequences, this application achieves multi-dimensional real-time determination of motion attribute information. This application simultaneously determines data from multiple dimensions, including the motion item identifier, the number of times the motion item has been completed, the duration of completion, parameter information, and skeletal point information, providing a rich information foundation for subsequent data display and recognition. Furthermore, by overlaying skeletal point information onto the target object in the real-time image, users can observe their own movement posture in real time and correct it against a standard posture, integrating visual feedback with data feedback, effectively improving the standardization and training effect of training movements. In addition, the overlay display of skeletal points also provides users with an intuitive basis for judging the quality of movement, enhancing the transparency and credibility of the motion recognition process.

[0093] In one embodiment of this application, step 1201 is included before step 120, which displays the real-time image information of the target object and the motion attribute information of the target object on the motion recognition page.

[0094] Step 1201: Acquire real-time image information.

[0095] Real-time image information refers to motion data of a target object captured in real time by the camera of a mobile terminal.

[0096] Acquiring real-time image information is a prerequisite for displaying motion recognition page information and recognizing motion attribute information. The system continuously acquires motion images of the target object through a camera, obtaining continuous motion frames or sequences of motion frames. These are used for real-time preview display on the motion recognition page and for the motion recognition SDK to perform motion analysis, frequency statistics, and skeletal point recognition. The acquisition process starts automatically when entering the motion recognition page and continues throughout the entire training process until training ends or the user exits the recognition page.

[0097] For example, after a user enters the motion recognition page, the system automatically accesses the mobile device's camera, such as the front-facing camera for selfie-style training recording, or the rear-facing camera for assisted filming or fixed-position filming, to begin real-time acquisition of the user's motion image information. The acquisition process continues at a rate of 25 to 30 frames per second to obtain continuous motion frames or sequences for subsequent SDK recognition and analysis. During acquisition, the system displays the camera's captured images in real-time in the first area of ​​the motion recognition page, allowing the user to see their own posture and adjust their movements accordingly during training. The acquisition process continues throughout the entire training cycle, from the start of training until the user finishes or exits the recognition page. The system stops acquiring images when the user completes training and clicks the end button.

[0098] In this way, by acquiring real-time image information before or simultaneously with the display of the motion recognition page, this application achieves real-time acquisition and instant display of motion data. Image acquisition automatically starts upon entering the motion recognition page, requiring no additional user intervention and simplifying the training start process. Simultaneously, the acquired real-time image information is used both for page preview and for SDK recognition and analysis, achieving synchronous acquisition and display and ensuring real-time updates of motion attribute information. Furthermore, it supports flexible selection of front and rear cameras, adapting to the shooting needs of different training scenarios and improving the applicability and convenience of the motion recording function.

[0099] In one embodiment of this application, the motion recognition page includes a first region and a second region; wherein the first region is used to display real-time image information of the target object, and the second region is at least partially overlapping with or adjacent to the first region, and the second region is used to display motion attribute information; the method 100 further includes step 160.

[0100] Step 160: In response to the preset operation on the second area, display the sports details information, which includes the attribute information corresponding to the sports item identifier in the sports attribute information.

[0101] The second region refers to the area in the motion recognition page used to display motion attribute information, which is adjacent to or at least partially overlaps with the first region (the area used to display real-time image information of the target object).

[0102] Preset actions refer to interactive behaviors performed by the target object on the second area to trigger the display of motion details, such as pull-up gestures, clicking the pull-up arrow, or tapping.

[0103] Exercise details refer to more detailed exercise attribute information corresponding to the current exercise identifier, including but not limited to training name, completed duration and required duration, training volume, calories burned, and training data for each set organized by movement. The system displays exercise details in response to the target user's preset operation on the second area, allowing the target user to view complete training data for the current exercise without leaving the exercise identification page.

[0104] For example, the first area of ​​the motion recognition page is the top area, occupying about two-thirds of the screen. It displays real-time footage of the user's movements captured by the camera, allowing the user to observe their own posture and movements. The second area is the bottom area, occupying about one-third of the screen. It displays motion attribute information, such as the motion item identifier, the number of times the motion item has been completed, and motion parameter information. After completing a training session, the user wants to view the complete data for that session, so they perform a swipe-up gesture on the second area. The system responds to this preset action, smoothly popping up a training details swipe-up page from the bottom of the screen. This page displays the motion details in half-screen or full-screen format, including the complete attribute information corresponding to the current motion item identifier "squat": training name, completed duration and required completion time, training volume, calories burned, and training data for each set organized by motion. The user can view detailed data on this swipe-up page and return to the motion recognition page to continue training by swiping down.

[0105] In this way, by dividing the motion recognition page into a first area displaying real-time image information and a second area displaying motion attribute information, this application achieves simultaneous partitioned display of motion images and motion information on the same screen. Through clear area division, this application allows users to clearly distinguish between visual feedback areas and data feedback areas, reducing visual interference caused by information clutter. Simultaneously, by displaying motion details in response to preset operations in the second area, users can view complete training data for the current exercise without exiting the motion recognition page, enabling convenient interaction for accessing details at any time during training. The smooth transition from half-pull-out cards to pull-up pages, coupled with the visual transition effect of fading old content and revealing new content, further enhances the smoothness of page transitions and the user experience.

[0106] In one embodiment of this application, when the second region at least partially overlaps with the first region, the second region is a region overlay page; step 160, in response to a preset operation on the second region, displays motion details information, including steps 161 and 162.

[0107] Step 161: In response to the first preset operation on the second area, display motion details information.

[0108] The first preset action refers to a lightweight interactive behavior performed by the target object on the second area to quickly view exercise details, such as tapping, clicking, or short pressing. Exercise details refer to brief exercise attribute information related to the current exercise, including but not limited to the name of the current set of movements, the number of times the exercise has been completed, and input boxes for exercise parameter information. In response to the target object's first preset action on the second area, the system directly displays the exercise details in the second area without requiring page switching or expanding to a larger area, thus meeting the target object's need to quickly view the current set of training data.

[0109] For example, the second area is a semi-pull-out card-style overlay page, floating below the first area (the real-time motion image display area). Initially, only the top gripper area of ​​the card is visible, displaying a small amount of basic information. During or after a training session, if a user wants to quickly view the current set's training data without interrupting the training flow, they can tap or click on the top of the card or any part of the card in the second area (the first preset operation). In response to this tap, the system smoothly expands the semi-pull-out card from a collapsed state to a semi-expanded state without switching pages, directly displaying the exercise details, including the name of the current set's movements, the number of repetitions, and input boxes for exercise parameters. Users can quickly browse this information to confirm the system's accuracy and edit directly in the input boxes if necessary. After viewing, users can tap the top of the card again or click on a blank area to collapse the card back to its initial state and continue training.

[0110] Step 162: In response to a second preset operation on the second region, display a region overlay page, the region overlay page including motion details information; and display an extension control, the extension control can be used to increase the area of ​​the region overlay page and add attribute information displayed in the motion details information.

[0111] The second preset operation refers to the interactive behavior performed by the target object on the second area to expand the area's overlay page to view more detailed motion information, such as a pull-up gesture, swiping up, or clicking the pull-up arrow. The area overlay page refers to the expanded page displayed in response to the second preset operation; its area is larger than the initial state, used to display more comprehensive motion details. Extended controls refer to interactive elements set within the area overlay page, used to further increase the page's area and display more attribute information. The system responds to the target object's second preset operation on the second area by displaying the area overlay page and extended controls to meet the target object's need for in-depth access to complete training data.

[0112] For example, when a user performs a pull-up gesture on the second area (the second preset action), such as swiping up a half-pull-up card or clicking the pull-up arrow on the card, the system responds by expanding the half-pull-up card into a larger pull-up page (the area overlay page). This page displays more comprehensive exercise details, including the training name, completed and required completion time, training volume, calories burned, and training data for each set organized by movement (e.g., Set 1: Squat, weight 50kg, 8 reps, 60 seconds rest; Set 2: Squat, weight 55kg, 6 reps, 90 seconds rest; Set 3: Bench Press, weight 40kg, 10 reps, 75 seconds rest, etc.). At the bottom or top right corner of the pull-up page, there are also expansion controls, such as an "expand" button or an upward arrow icon. When the user clicks this expansion control, the system responds by further increasing the area of ​​the area overlay page, expanding the pull-up page to full screen or near full screen, displaying more attribute information, such as historical set records, detailed exercise descriptions, links to exercise demonstration videos, and nutritional supplement suggestions. Users can expand the floating area pages level by level as needed, achieving a seamless transition from quick browsing to detailed viewing. After viewing, users can collapse the pages level by swiping down or clicking the close button to return to the motion recognition page and continue training. In this way, by distinguishing between the first and second preset operations, this application provides users with different levels of detailed information viewing methods. This application satisfies users' lightweight needs for quickly viewing information of the current group through a tap operation, and satisfies users' in-depth needs for viewing complete training details through an up-swipe operation. At the same time, through the setting of extended controls, users can dynamically adjust the area of ​​the floating area pages according to their needs, achieving a smooth transition from a half-pull-out card to a full detail page. This hierarchical interaction design ensures both high efficiency for daily use and flexibility for in-depth viewing. Users can seamlessly switch from quick browsing to detailed viewing within the same interface, effectively improving the efficiency of information acquisition and the interactive experience.

[0113] In one embodiment of this application, step 170 is also included.

[0114] Step 170: If the training end operation of the target object is received, obtain the number of times the exercise has been completed, and determine whether the number of times the exercise has been completed is equal to the target number of times the exercise is in the exercise parameter information. If so, generate and display the corresponding exercise record based on the target object's exercise attribute information within the preset time.

[0115] End training refers to the interactive behavior of the target object actively ending the current training task, such as clicking the exit button in the upper left corner of the motion recognition page and selecting the "End Exercise" option, or clicking the "End Training" button.

[0116] The preset duration refers to the entire time period from the start to the end of this training session. When the system receives the target user's "end training" command, it automatically retrieves the number of times the current exercise has been completed and compares it to the preset target number of times for that exercise. If the number of completed times equals the target number, it indicates that the target user has achieved the preset training goal. Based on all the target user's exercise attributes within the preset duration (including exercise identifier, number of times completed for each set, exercise parameters for each set, rest periods between sets, etc.), the system generates a complete training record and displays it on the data processing page for user confirmation and saving, providing achievement feedback and positive incentives.

[0117] For example, before starting training, a user presets a target number of repetitions ("10") for the exercise "squats" through the settings on the motion recognition page. During training, the system uses the motion recognition SDK to record and accumulate the number of repetitions completed in each set. When the user completes the last set of squats, the system has accumulated 10 repetitions. The user clicks the exit button in the upper left corner of the motion recognition page, and a confirmation layer pops up, offering two options: "End Exercise" and "End Without Saving." If the user selects "End Exercise," the system responds by retrieving the current number of repetitions ("10") and comparing it with the preset target number ("10"). If the system determines that the number of repetitions equals the target number, the user has achieved the preset goal for this training session. Based on all the user's exercise attributes within the preset duration, including exercise identifier, number of repetitions completed for each set, parameter information for each set, and rest duration between sets, the system generates a complete training record, which is then displayed to the user for confirmation and saving on the data processing page. Users can view the complete training data on the data processing page and save the record after confirming that it is correct.

[0118] In this way, by determining whether the number of repetitions completed for the exercise equals the target number of repetitions when the training ends, this application can automatically identify whether the user has completed the preset training goal. This application generates a complete exercise record when the user achieves the training goal, providing achievement feedback and positive incentives. Simultaneously, the system generates records based on exercise attribute information within a preset time period, ensuring the integrity and accuracy of training data and facilitating subsequent viewing and statistical analysis by the user. When the user does not achieve the goal, the system guides the user to choose to continue training or adjust the training plan through an editable area, providing a flexible training end processing mechanism and improving the intelligence and user-friendliness of exercise data recording.

[0119] The following describes in detail the application embodiments of this application in conjunction with steps S201 to S208. Please refer to... Figure 3This application can apply the above scheme to the independent recording method of sports software.

[0120] S201. The user enters the Gym Tab interface in the application and clicks the entry point for the automatic recognition function of the fitness app. The system determines whether the user is using this function for the first time. If it is the first time using it, the system first displays an informed consent page. At the bottom of this page, there is a radio button with the text "Agreed to Collecting Camera Recording Data". The user must check this radio button for the "Get Started" button to change from grayed-out to highlighted. After the user checks the radio button and clicks the "Get Started" button, the system jumps to the new user guide page, which includes instructions on how to place the phone.

[0121] The system uses animations to demonstrate the correct phone placement: the face should be oriented at the 12 o'clock position, and the phone can be placed between 5 and 11 o'clock or between 13 and 17 o'clock; the user's limbs should be spread out and still within the frame; the camera should be positioned at eye level with the core torso area. The guide page also includes exercise parameters, the number of repetitions completed, and motion editing guidance, guiding the user to edit the weight on their phone after each set, with GIF animations illustrating how to do so; the page text will prompt, "You need to manually enter the weight; if the recognition is inaccurate, you can manually calibrate." After the user clicks the "Get Started" button on the guide page, the system redirects to the motion recognition page. The guide page shrinks to a small icon displayed in the upper left corner of the screen for easy reference. If the user is not using the system for the first time, the system will treat clicking the automatic recognition function entry in the exercise software as the function trigger operation, directly redirecting to the motion recognition page. This is equivalent to step 110 above.

[0122] S202. After entering the motion recognition page, the system begins to capture real-time motion images using the phone's camera. The system calls the motion recognition SDK to recognize the real-time motion images. When the system recognizes a human body, it overlays and displays human skeletal points, including the user's joints and the skeletal points of the limbs and torso, onto the real-time motion image on the display screen. The system does not display the number of movements until the SDK returns the recognition results. The status text "Recognizing..." is displayed in the action prompt box, indicating that the system is recognizing; the system also displays the total training time, which is calculated from the first time a human body is recognized. This is equivalent to step 1201 above.

[0123] S203 and SDK continuously identify real-time motion images, performing action recognition and count. When the SDK returns the first positive sample (i.e., a valid motion action is detected), the system begins to determine the action name and update the interface display.

[0124] The determination of action names follows these priority rules: 1. P0 level (highest weight): Sub-category action manually selected by the user. When the user manually modifies the action, as long as the major category recognized by the device is the same as the previous group, the system directly reuses the sub-category action selected by the user, without waiting for the return result from the cloud VLM (Visual Language Model).

[0125] 2. P1 Level (Second in Weight): Subclass actions returned by the cloud-based VLM. If the previous group of actions was identified and determined by VLM, and the major category identified on the client side in this group is the same as the previous group, the system first pre-populates the P1 subclass actions of the previous group to maintain interface continuity, but the backend still waits for the latest VLM result for this group. If the VLM in this group returns successfully, the system immediately refreshes the interface with the latest VLM result and records it; if the VLM does not return or times out, the pre-populated P1 subclass actions shall prevail.

[0126] 3. P2 level (lowest weight): The default subclass sorting under the major category identified by the edge model, used only as a fallback.

[0127] When the major category returned by the endpoint changes, the system clears the historical state and re-initiates cloud-based VLM recognition; before the VLM returns, the default subclass at level P2 is temporarily used. The system's network strategy is: major category recognition is completed on the endpoint and does not depend on the network. The video stream is cached in a scrolling window for the most recent 5 seconds. When a positive sample appears for the first time, the system sends the previous 5 seconds of video to the cloud-based VLM for minor category analysis, with the estimated cloud analysis time being 6 to 15 seconds.

[0128] After the SDK returns the number of movements, the system displays the recognized movement count in large font in real time in the center of the screen. Simultaneously, the name of the currently recognized sub-category of movement is displayed in the action prompt box, with a button to switch actions on the right side of the prompt box. The system updates the count returned by the SDK in real time; each time the user completes a valid movement, the displayed count increases accordingly. This is equivalent to steps 120, 121, and 122 above.

[0129] S204. The system determines whether a user has entered a rest state in two ways: passive rest and active rest. The criteria for passive rest are: the SDK returns the first negative sample, indicating invalid movement or stillness, provided that all previous samples were positive. The system must prevent short pauses from being misjudged as rest, which is achieved by setting a time threshold or a sample count threshold.

[0130] The trigger conditions for voluntary rest include: the user opening the training details page (pull up), clicking the switch action button, and opening the numeric keypad to enter the exercise parameters on a partially scrolled page. When voluntary rest is triggered, the system stops uploading video streams to the SDK, but does not turn off the camera.

[0131] Once the system determines that it has entered a resting state (whether passive or active rest), it immediately refuses to accept any delayed VLM return results. If VLM data is still not obtained at the time of settlement, the system directly records the group as the default subclass of P2. This is equivalent to steps 130 and 140 above.

[0132] S205. When the exercise set is determined to be completed, the system enters a rest state, changes the number of repetitions displayed on the screen to rest timer information, and automatically pops up a half-finish page. The half-finish page displays information about the current set, including: set number, name of the identified sub-type of movement (switchable), number of repetitions completed (editable), and weight input box (for user to fill in).

[0133] The naming rules for group numbers are as follows: if the current group and the previous group belong to the same subclass of action, the group number continues to increment; if the current group and the previous group are not the same subclass of action, the group number for this action starts from 1; if the user manually modifies the action, the group number is recalculated based on the modified action. If the system recognizes an action as belonging to the same subclass of action as the previous group, the weight input box is automatically filled with the weight value previously entered by the user.

[0134] Users can perform the following operations: Activate the keyboard, enter weight information, and click on a blank space to confirm. At this time, the keyboard will collapse, and the background training volume and calorie data will be updated accordingly; Activate the keyboard but do not enter any information, and the system will not upload videos to the SDK; Activate the keyboard but do not enter any information and pull down half of the page, which will retain its original style; Ignore the half-pull-down page and proceed directly to the next set of exercises. When the system recognizes that the user has started the next set, it will automatically pull down half of the page and return to the exercise recognition page; Users can also pull up half of the page or click the up arrow to enter the training details pull-up page; The rest timer is calculated based on an absolute timestamp, and the timer will not stop even if the application is switched to the background. This is equivalent to steps 131, 160, and 161 above.

[0135] S206. When the system recognizes an incorrect action, the user can click the "Switch Action" button on the right side of the action prompt box. The system will display a list of all subcategories of actions under that category, with the first one selected by default. The user can then click the option to select the correct action. The action selection bar can display up to 5.5 cards high. When there are 5 or fewer actions but more than 1, the height will automatically adjust according to the number of actions. If there is only one option, only that option will be displayed, with the "View All Exercises" button below it.

[0136] If none of the available actions are the user's current action, the user can click the "Other" option to select the correct action from the complete library of recognizable subcategories. After the user makes their selection, the system will redirect them back to the half-page. If the selected action is different from the previous one, the user must manually enter the weight; if it is the same as the previous action, the system will automatically enter the previous weight value; if the previous weight was 0, the current weight will also be automatically entered as 0.

[0137] The system's action library divides actions into two categories: unilateral actions that require differentiation between left and right and bilateral actions that do not require differentiation between left and right.

[0138] For unilateral exercises that require left-right differentiation (e.g., dumbbell lunges, unilateral rows), the system retains the L (left) or R (right) label when recording at the bottom layer. The recognition page uniformly displays the number of repetitions (Reps) completed for the exercise, without distinguishing between L / R. If the user performs 10 repetitions on the left side first, followed by 10 repetitions on the right side without rest in between, the total number of repetitions displayed on the half-pull page is 20Reps, and the data returned by the SDK is labeled L*20reps. If the user only performs 10 repetitions on the right side and then rests, the half-pull page displays 10Reps, and the data returned by the SDK is labeled R*10reps. On the training details pull-up page, the system labels L or R after the group number. For bilateral exercises that do not require left-right differentiation (e.g., double-ended rope pulldowns, barbell bench presses), the system records them in the format "1 set:kg", without labeling L / R. This is equivalent to steps 123, 124, 132, and 133 above.

[0139] S207. Users can access the training details page by swiping up halfway or clicking the up arrow. This page is used to view complete training information for the current workout. The name of the workout is displayed at the top of the page; the default name is "Gym&FreeWeight," which can be edited by the user, with a maximum of 50 characters. The page also displays the completed and remaining time, real-time heart rate, training volume, and calories burned. The page organizes training data by movement. When the SDK returns information, the system updates the corresponding movement cards in the background. Connected movements of the same subcategory are merged into different groups under the same movement module. Each movement module includes: movement name and cover (clicking will take you to the movement details page), and three dots on the far right (clicking will delete the movement, requiring secondary confirmation). Each set of training data is displayed in tabular form, including: set number (Set), weight (kg), number of repetitions completed (Reps), and rest time between sets (Rest). Weight and number of repetitions completed are editable, while rest time between sets is not. Groups can be deleted by swiping left.

[0140] Weight editing follows a backward-propagating principle, meaning it only affects weights within the same action: when a user manually modifies the weight in one group, it can affect subsequent groups in batches, but it must not affect data that already exists in that group. Confirmed data on the training page and manually entered data directly on the details page should retain their confirmed results and not be unconditionally overwritten. The page background is a frosted glass background, and the camera remains on, but video is not transmitted to the SDK.

[0141] When the application switches to the background during recognition, the system immediately terminates the current group, uses the data returned by the SDK to calculate the motion attribute information and the number of times the exercise has been completed, and the page enters a rest state. The rest timer is calculated based on an absolute timestamp, and the camera is no longer accessed. When the application returns to the foreground, the system continues to display the rest timer and automatically pops up a half-open page, waiting for the user to enter the weight. When the application switches to the background during rest, the system stops accessing the camera. When the application returns to the foreground, the system remains in a rest state, and the rest timer continues to run based on an absolute timestamp. If the keyboard was active before switching to the background, it can be hidden when returning to the foreground, requiring the user to reactivate it.

[0142] Each time the system generates a new set of data, it writes the data to the local cache in real time. If the current set has never received any positive SDK samples, the records for that set are discarded when the application is forcibly closed. If there are training caches that have not ended normally, the system will display a prompt window when the user cold starts, titled "Incomplete training found," with the text "You had an unfinished training record last time, do you want to continue?", providing three buttons: "Continue Training" (loads the local cache and jumps to the training details pull-up page), "End and Save" (directly settles the training and enters the data processing page), and "Discard" (deletes the data and returns to the application homepage). If the system exits abnormally on the data processing page, the system will display a prompt window when the user cold starts, titled "You have incomplete data editing!", providing two buttons: "Continue Editing" and "Discard." This is equivalent to steps 162 and 170 above.

[0143] S208. When the user clicks the exit button in the upper left corner of the interface, a confirmation window pops up, offering two options: "End Exercise" and "End Without Saving." The user can close the pop-up by clicking on a blank area. If the user selects "End Exercise," the system redirects to the data processing page. This page reuses the manual training results page, allowing users to manually add groups and exercises. Groups for which the user has not entered a weight are not selected by default; they are automatically selected after the user enters a weight.

[0144] After user confirmation, the system saves the training record. Offline training data is incorporated into statistics such as muscle fatigue calculation, total data center training volume, total duration, and total energy expenditure. This data is automatically included in the offline training data collection but is not counted in the activity and badge systems. This is equivalent to steps 135 and 170 above.

[0145] Through the above application examples, the traditional information recording mode, which is dominated by manual user operation and involves multiple page jumps, is reconstructed into a parallel interaction mode where the system automatically identifies and displays information on the same screen. This architecture simplifies the user operation path, reduces interface switching latency, and effectively improves the consistency and accuracy of training data recording. This helps users focus more on the training itself and ultimately improves the human-computer interaction efficiency and user experience of fitness software.

[0146] Please refer to Figure 4 , Figure 4 This is a schematic diagram of an interface for a motion information display method provided in an embodiment of this application.

[0147] like Figure 4 As shown, the motion recognition page 301 includes a first region and a second region. The first region is used to display real-time image information 302 of the target object, and the second region is set adjacent to the first region to display motion attribute information 303 of the target object.

[0148] The motion recognition page displays real-time image information 302 and motion attribute information 303 of the target object. The real-time image information 302 includes real-time motion footage of the user captured by the mobile terminal's camera, and the motion attribute information 303 includes at least one of the following: motion item identifier, number of times the motion item has been completed, duration of the completed motion item, and motion item parameter information.

[0149] The total training time display area 304 shows the total training time for this session. In the image, the total training time display area 304 shows "00:00:04", indicating that 4 seconds of training have been completed since the first human body was recognized. This time is calculated from the moment the system first recognizes the target human body and is continuously accumulated and updated during the training process.

[0150] The action prompt box 305 is used to display the name of the currently recognized action, corresponding to the "sports item identifier" in claim 2. In the figure, the action prompt box 305 displays "Recognizing...", indicating that the system is recognizing the action and has not yet returned the recognition result. Before the SDK returns the recognition result, the system displays the status text "Recognizing..." in the action prompt box 305 to indicate to the target object that the system is currently performing action recognition analysis.

[0151] The action prompt box 305 has a switch action button 306 on the right. This button is an icon button. The target object can click the switch action button 306 to trigger the switch action operation. In response to the trigger operation, the system displays a list of motion attribute information switching. The switching list includes a list of motion item identifiers. The list of motion item identifiers contains at least two different candidate motion item identifiers for the target object to select the correct action name.

[0152] Please refer to Figure 5 , Figure 5 This is a schematic diagram illustrating a motion information display method provided in this application under the scenario of recognizing the motion state of a target object.

[0153] like Figure 5 As shown, the motion recognition page 301 includes a first region and a second region. The first region is used to display real-time image information 302 of the target object, and the second region is set adjacent to the first region to display motion attribute information 303.

[0154] The top of the motion recognition page displays a status bar and page icons, with "Gym Workout" indicating that the user is currently on a fitness training page and "SIDEKICK" indicating that the motion recognition function of the software is being used.

[0155] The target object real-time image information 302 displays real-time motion footage of the user captured by the mobile terminal's camera. The motion attribute information 303 includes an action name display area 304, a number of motion counts display area 305, an action switch button 306, and a total training time display area 307.

[0156] The action name display area 304 is used to display the name of the currently identified action. In the figure, it is shown as "Smith Seated Shoulder Press", indicating that the system has successfully identified the valid action being performed by the target object.

[0157] The exercise count display area 305 shows the number of exercises completed in the current group in a larger font. The exercise count display area 305 in the image shows specific numbers, indicating that the system has been counting and displaying the number of times the user has completed the exercise in real time.

[0158] To the right of the action name display area 304, there is a switch action button 306, which is an icon button. The target object can click the switch action button 306 to trigger the action switching operation. In response to the trigger operation, the system displays a list of motion attribute information for the target object to select the correct action name.

[0159] The total training time display area 307 shows the total training time for this session. In the image, the total training time display area 307 shows "00:02:31", indicating that 2 minutes and 31 seconds of training have been completed since the first human body was recognized. This time is calculated from the moment the system first recognizes the target human body and is continuously accumulated and updated during the training process. After the system recognizes the human body, it overlays and displays skeletal point information, including the user's joint points and skeletal point annotations for the limbs and torso, onto the user's image in the real-time image information 302.

[0160] Please refer to Figure 6 , Figure 6 This is a schematic diagram illustrating a motion information display method provided in this application embodiment when a target object is identified as entering a resting state.

[0161] like Figure 6 As shown, the real-time image information 302 of the target object displays the user's real-time movement image captured by the mobile terminal's camera. During rest periods, the camera remains on, and the system stops uploading video streams to the SDK, but the user can still see their own image in the first area to observe their posture or prepare for the next training session.

[0162] The editable interactive area 303 is a half-pull page that automatically pops up from the bottom of the screen. When the exercise is determined to be completed, the system enters a resting state and automatically pops up this half-pull page, which is the editable interactive area 303. The editable interactive area 303 is used to display editable exercise attribute information.

[0163] The total training time display area 304 shows the total training time of this training session. The image shows "00:02:31", which means that 2 minutes and 31 seconds of training have been completed since the first human body was recognized.

[0164] The exercise name display area 305 shows the exercise name for the current set, displayed as "Smith Seated Shoulder Press" in the image, indicating the type of exercise in this set. This exercise name is editable; users can click to switch to other exercises.

[0165] To the right of the action name display area 305, there is a switch action button 306, which is an icon button. The target object can click the switch action button 306 to trigger the action switching operation. In response to the trigger operation, the system displays a list of motion attribute information switching. This switching list includes a list of motion item identifiers, which contains at least two different candidate motion item identifiers for the target object to select the correct action name.

[0166] The rest timer display area 308 shows the current rest time, displayed as "00:02" in the image, indicating that the user has rested for 2 seconds. When the system determines that the exercise session has ended and the user has entered a rest state, the original display area 307 showing the total training time of the exercise sessions is transformed into the rest timer display area 308, and the displayed content changes from the total training time of the exercise sessions to rest timer information. The rest timer is calculated based on an absolute timestamp, so the timer will not stop even if the application is switched to the background.

[0167] The weight display area 309 shows the weight value entered by the user in the current group. In the image, it is displayed as "59," indicating a weight of 59 kg. The weight unit "kg" is labeled below. If the current group and the previous group belong to the same subcategory of actions, the weight input box will automatically fill in the weight value entered by the user previously. The user can activate the keyboard to enter the weight information and click on a blank space to confirm. If the action modified by the user belongs to the same major category as the previous group, the system will automatically use the weight value from the previous group; if they do not belong to the same major category, the user will be prompted to manually enter the weight.

[0168] The repetition setting area 310 is used to display and set the number of repetitions. In the image, it is displayed as "Reps 17," indicating that the user has completed 17 repetitions. The user can edit this number, and the system updates the exercise-related information in response to the user's modification. This number is counted in real time by the SDK, and the user can also manually adjust it through the repetition setting area 310 in the editable interactive area 303.

[0169] In summary, the present invention provides a method for displaying motion information and an electronic device.

[0170] By responding to function triggers and displaying a motion recognition page, and when the motion attribute information of the target object in the real-time image information meets preset conditions, the motion recognition page simultaneously displays the real-time image information and motion attribute information of the target object, achieving simultaneous display of motion images and motion information on the same screen. Unlike traditional solutions that simply overlay the recognition information in the corner of the preview screen, users do not need to frequently switch their visual focus between observing the action and viewing the data, reducing visual jumps and cognitive burden.

[0171] By dividing the motion recognition page into a first area and a second area, the first area displays real-time image information of the target object, while the second area displays motion attribute information, thus clearly distinguishing between the visual feedback area and the data feedback area. When skeletal point information is recognized, it is overlaid on the target object in the real-time image information, allowing users to observe and correct their own movements in real time. Simultaneously, the second area displays motion attribute information such as the motion item identifier, the number of times the motion item has been completed, and the duration of completion. The two types of information complement each other and are presented collaboratively, allowing users to obtain both visual and data feedback simultaneously without switching interfaces, effectively solving the problem of training continuity interruption caused by frequent page jumps.

[0172] Upon recognizing a valid movement, the system enters a resting state, switching the movement attribute information displayed on the movement recognition page to rest timing information. An editable interactive area automatically pops up from the bottom of the screen, enabling automatic recognition and interface feedback for training state switching. Users only need to complete weight replenishment and repetition correction within the editable interactive area without leaving the current recognition page, making the operation more convenient and fluid. By setting up a movement attribute information switching control, users can quickly complete movement correction within the same page. The system records the user-selected movements as the highest priority, facilitating the inheritance of movement recognition for subsequent groups.

[0173] This invention solves the technical problems of existing fitness software, such as inconvenient display of motion recognition information, manual operation required for switching training states, interruption of training continuity by data editing, and cumbersome motion correction process. It has great practical value in improving the human-computer interaction efficiency of fitness software, ensuring the continuity and accuracy of training data recording, and improving user experience.

[0174] The above are merely embodiments of the present invention and do not limit the patent scope of the present invention. Any equivalent modifications made based on the content of the present invention's specification and drawings, or direct or indirect applications in related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for displaying motion information, characterized in that, include: In response to a function trigger operation, a motion recognition page is displayed; wherein, the motion recognition page includes real-time image information of the target object; If the motion attribute information of the target object in the real-time image information of the target object meets the preset conditions, the real-time image information of the target object and the motion attribute information of the target object will be displayed on the motion recognition page.

2. The motion information display method according to claim 1, characterized in that, The sports attribute information includes the sports event identifier; If the motion attribute information of the target object in the real-time image information of the target object meets the preset conditions, then the real-time image information of the target object and the motion attribute information of the target object are displayed on the motion recognition page, including: If the motion identifier of the target object in the real-time image information of the target object is identified as belonging to a preset set of motion items, then the real-time image information of the target object and the motion attribute information of the target object are displayed on the motion recognition page; The sports attribute information includes at least one of the following: sports event identifier, number of times the sports event has been completed, duration of the completed sports event, and preset sports event parameter information.

3. The motion information display method according to claim 2, characterized in that, Also includes: If the motion attribute information of the target object in the real-time image information of the target object does not meet the preset conditions, then the candidate motion item identifier is displayed in the editable area of ​​the motion recognition page, and the number of times the candidate motion item has been completed or the parameter information of the candidate motion item corresponding to the candidate motion item identifier is displayed. In response to an editing operation on the editable area, the editing result of at least one of the candidate sports item identifier, the number of times the candidate sports item corresponding to the candidate sports item identifier has been completed, or the parameter information of the candidate sports item is displayed in the editable area of ​​the motion recognition page.

4. The motion information display method according to claim 3, characterized in that, If the sports item identifier includes a multi-sided sports mode identifier, then modifying the sports attribute information in response to the editing result of the editable area includes: In response to a switching operation on the multi-side parameter information in the sports parameter information, the sports side identifier displayed in the editable area is switched; The multi-side parameter information includes motion side identifiers for at least two sides with different motion directions.

5. The motion information display method according to claim 3, characterized in that, The sports event parameter information includes the target number of repetitions for the sports event; In response to the received training termination operation of the target object, the number of times the exercise has been completed is obtained, and it is determined whether the number of times the exercise has been completed is less than the target number of times the exercise has been completed. If so, the process of displaying candidate exercise identifiers in the editable area of ​​the exercise recognition page, and displaying the number of times the candidate exercise has been completed or the candidate exercise parameter information corresponding to the candidate exercise identifier is executed.

6. The motion information display method according to claim 1, characterized in that, The motion recognition page also includes a motion attribute information switching control; The display of the real-time image information and motion attribute information of the target object on the motion recognition page includes: In response to the triggering operation of the motion attribute information switching control, a motion attribute information switching list is displayed. The motion attribute information switching list includes a list of sports item identifiers, and the list of sports item identifiers includes at least two different candidate sports item identifiers. In response to the selection operation of the target sport identifier in the list of sport identifiers, the sport identifier displayed on the sport identification page is updated to the target sport identifier; the target sport identifier is one of the candidate sport identifiers in the list of sport identifiers.

7. The motion information display method according to claim 1, characterized in that, Also includes: If the motion attribute information of the target object in the real-time image information of the target object does not meet the preset conditions, a prompt message will be displayed on the motion recognition page. The prompt message is used to prompt the user to edit the motion attribute information.

8. The motion information display method according to claim 1, characterized in that, The real-time image information of the target object includes real-time image information; or, the real-time image information of the target object includes real-time image information and skeletal point information. Wherein, the real-time image information of the target object includes real-time image information and skeletal point information corresponding to the target object in the real-time image information, the motion recognition display page includes: After acquiring the real-time image information, based on the recognition result of at least one motion frame or motion frame sequence in the real-time image information, at least one of the following in the motion attribute information is determined: the motion item identifier, the number of times the motion item has been completed, the duration of the completed motion item, the motion item parameter information, and the skeletal point information. The real-time image information is displayed. If the skeletal point information is identified, the skeletal point information is overlaid on the target object in the real-time image information.

9. The motion information display method according to claim 8, characterized in that, Before displaying the real-time image information and motion attribute information of the target object on the motion recognition page, the process also includes: The real-time image information is acquired.

10. The motion information display method according to claim 1, characterized in that, The motion recognition page includes a first region and a second region; Wherein, the first region is used to display real-time image information of the target object, the second region is at least partially overlapping with or adjacent to the first region, and the second region is used to display the motion attribute information; The method further includes: In response to a preset operation on the second area, sports details information is displayed, including attribute information corresponding to the sports item identifier in the sports attribute information.

11. The motion information display method according to claim 10, characterized in that, When the second region at least partially overlaps with the first region, the second region is a region overlay page; The response to a preset operation on the second region, displaying motion details information includes: In response to a first preset operation on the second region, the motion details information is displayed; In response to a second preset operation on the second region, a floating page of the region is displayed, the floating page of the region including the motion details information; and an extension control is displayed, the extension control being used to increase the area of ​​the floating page of the region and to add attribute information displayed in the motion details information.

12. The motion information display method according to claim 2, characterized in that, Also includes: If the training end operation of the target object is received, the number of times the exercise has been completed is obtained, and it is determined whether the number of times the exercise has been completed is equal to the target number of times the exercise is in the exercise parameter information. If so, the corresponding exercise record is generated and displayed according to the exercise attribute information of the target object within a preset time period.

13. An electronic device comprising a camera, a display screen, a memory, and a processor, wherein the memory stores a computer program, and the processor, when executing the computer program, controls the electronic device to implement a motion information display method according to any one of claims 1 to 12.

Images