A rope skipping posture risk identification method and device based on a swing ball trajectory

Abstract

The application discloses a ropeless skipping posture risk identification method and device based on a swinging ball trajectory. The method collects swinging ball trajectory signals, handle inertia signals and foot landing impact signals, extracts features such as relative height curves, trajectory plane changes, left and right swinging ball phase relationships, trajectory closure, trajectory plane inclination and relative height peak and valley differences; combines motion safety prior constraints and user individual action trajectory baseline models to identify posture risks such as swinging ball deviation, non-closed trajectory, left and right asynchrony, hand and foot rhythm mismatch, empty swinging, half jumping, abnormal landing impact and post-fatigue action deformation, and prevents a self-learning model from learning bad actions as normal actions through a safety boundary limiting mechanism to output personalized correction prompts. The application does not collect images, sounds or face information and is not used for medical diagnosis or treatment effect judgment.

Description

Technical Field

[0001] This invention relates to the fields of intelligent sports equipment, cordless jump rope, wearable sensors, inertial sensor signal processing, trajectory quality evaluation, machine learning, and sports training assistance. In particular, it relates to a method and device for identifying posture risks in cordless jump rope based on the trajectory of a swing ball, the inertial signal of the handle, the impact signal of the foot landing, prior constraints on sports safety, and a baseline model of the user's individual movement trajectory. Background Technology

[0002] Cordless jump ropes typically simulate the swinging sensation of traditional jump ropes through handles, connecting ropes, and swing balls. They are characterized by their small footprint, ease of indoor training, and reduced risk of tripping. Existing cordless or smart jump rope devices primarily focus on estimating the number of repetitions, frequency, duration, and energy consumption. Some solutions use handle rotation, inertial sensors, or foot contact signals for motion recognition and counting correction.

[0003] However, since there is no actual rope passing under the feet in a cordless jump rope, the rotation of the handles does not necessarily indicate that the user has completed a stable, coordinated, and training-compliant jump. The trajectory of the swing ball can reflect the user's wrist swing pattern, left-right hand coordination, swing amplitude, rhythm stability, and upper limb compensation. Relying solely on handle rotation or simple thresholds makes it difficult to identify postural risks such as ball deviation, incomplete trajectory, asynchronous left-right swing, hand-foot rhythm mismatch, empty swing, half-jump, abnormal landing impact, and post-fatigue movement distortion.

[0004] Meanwhile, different users vary in height, weight, wrist exertion methods, jump range, landing habits, and exercise rhythm. Using only fixed thresholds can easily lead to false alarms or missed alarms; relying entirely on users' historical data for learning may result in long-term poor movement habits being learned as normal movements. Therefore, a posture risk identification scheme for evaluating the trajectory quality of a swing ball in cordless jump rope is needed, enabling the system to build individualized movement models under prior constraints of exercise safety and output personalized correction prompts. Summary of the Invention

[0005] The technical problem to be solved by this invention is to overcome the problem that existing cordless jump rope devices rely on counting, handle movement or fixed thresholds, and have difficulty in identifying posture risks using the quality of the swing ball trajectory. The invention provides a method and device for identifying posture risks in cordless jump rope based on the swing ball trajectory, which enables the system to identify and correct posture risks during cordless jump rope training by combining handle inertial signals, foot impact signals, prior constraints on motion safety and baseline models of individual user movements.

[0006] To address the aforementioned technical problems, this invention provides a method for identifying posture risks in cordless jump rope based on the trajectory of a swing ball, comprising the following steps: acquiring the trajectory signal of the swing ball in the cordless jump rope, wherein the trajectory signal includes at least the relative height curve of the swing ball, the trajectory plane change, and the phase relationship between the left and right swing balls; acquiring the inertial signal of the cordless jump rope handle and the impact signal of the foot landing; synchronizing the trajectory signal of the swing ball, the inertial signal of the handle, and the impact signal of the foot landing in time to establish a unified time axis, and performing sliding window segmentation, noise reduction, normalization, peak detection, motion cycle identification, and phase relationship calculation.

[0007] The quality features of the swing ball trajectory are extracted based on the swing ball trajectory signal. These features include trajectory closure, trajectory plane inclination, relative height peak-to-valley difference, and left-right swing ball phase difference. Prior constraints for motion safety are constructed, including upper limit of landing impact, upper limit of impact difference between left and right feet, upper limit of swing ball trajectory plane inclination, upper limit of left-right swing ball phase difference, range of phase difference between handle rotation event and foot landing event, upper limit of consecutive abnormal impact number, and range of motion fluctuation after fatigue.

[0008] Stable and valid motion samples are selected based on the consistency verification results of the swing ball trajectory quality characteristics, handle rotation event, and foot landing event. These stable and valid motion samples are those that simultaneously satisfy the swing ball trajectory quality condition, the timing matching of the handle rotation event and the foot landing event, the consistency of the handle rotation period and the foot landing period, and do not violate prior constraints on motion safety.

[0009] A baseline model of individual user movement trajectory is established based on stable and effective movement samples. The baseline model of individual user movement trajectory includes the range of normal ball swing trajectory closure, the range of normal trajectory plane inclination, the range of normal relative height peak-to-valley difference, the range of normal left-right ball swing phase difference, the range of normal hand-foot phase difference, the range of normal landing impact, and the range of normal rhythm fluctuation.

[0010] During real-time training, the current motion characteristics are compared with prior constraints on motion safety and the user's individual motion trajectory baseline model to identify postural risk states. These postural risk states include ball deviation, incomplete trajectory closure, abnormal swing height, asynchronous left and right ball swings, hand-foot rhythm mismatch, excessive upper limb compensation, empty swing, half-jump, excessive landing impact, uneven force distribution, and motion deformation due to fatigue.

[0011] Personalized correction prompts are output based on the posture risk identification results. These personalized correction prompts include reducing the swing ball's deviation, maintaining synchronization between the left and right hands, maintaining a complete swing cycle, synchronizing the handle swing with the landing rhythm, reducing landing impact, adjusting the left and right forces, reducing the jumping frequency, or pausing to rest.

[0012] A self-learning update mechanism is established to update the baseline model of individual user motion trajectories based on stable and valid motion samples, user confirmation markers, prompt response results, or statistical results of motion stability after training. A safety boundary constraint mechanism is set to ensure that the updated results of the individual user motion trajectory baseline model do not exceed the prior constraints of motion safety. When a candidate update result causes the landing impact range, the impact difference range between the left and right feet, the tilt range of the swing ball trajectory plane, the phase difference range between the left and right swing balls, or the phase difference range between the handle rotation event and the foot landing event to exceed the corresponding prior constraints of motion safety, the candidate update result is rejected or restricted to within the prior constraints of motion safety.

[0013] The present invention also provides a ropeless jump rope posture risk identification device based on the trajectory of a swing ball, including a swing ball trajectory acquisition module, a handle inertia acquisition module, a foot impact acquisition module, a time synchronization and feature extraction module, a trajectory quality evaluation module, a prior constraint module, a consistency verification module, an individual motion trajectory baseline model module, a posture risk identification module, a self-learning update module, a safety boundary limitation module, and a correction prompt module.

[0014] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described method.

[0015] Compared with the prior art, the present invention has at least the following beneficial effects: First, the present invention does not take ordinary jump rope counting as the core, but uses the trajectory of the swing ball in the ropeless jump rope as an important basis for identifying posture risks; Second, the present invention evaluates the quality of the swing ball trajectory through features such as relative height curve, trajectory plane change, left and right swing ball phase relationship, trajectory closure, trajectory plane inclination, and relative height peak-valley difference, and can identify ropeless jump rope-specific problems such as swing ball deviation, trajectory non-closure, and left and right swing asynchrony.

[0016] Third, by combining the swing ball trajectory signal, handle inertia signal, and foot landing impact signal, this invention can distinguish between stable and effective movements, empty swings, half jumps, hand-foot rhythm mismatches, and abnormal landing impacts. Fourth, this invention limits the update boundary of the user's individual movement trajectory baseline model through prior constraints on movement safety, avoiding the learning of long-term poor movements as normal movements. Fifth, this invention does not collect images, sounds, or facial information, unlike jump rope recognition schemes based on video human body key points, which helps reduce privacy risks and deployment costs. Sixth, this invention outputs sports training assistance prompts and posture risk prompts, and is not used for medical diagnosis or treatment effect judgment. Attached Figure Description

[0017] Figure 1 This is an overall structural diagram of the device provided in an embodiment of the present invention.

[0018] Figure 2This is a flowchart for evaluating the trajectory quality of a swing ball, provided in an embodiment of the present invention.

[0019] Figure 3 A flowchart for screening stable and effective action samples provided in an embodiment of the present invention.

[0020] Figure 4 A flowchart for updating the baseline model of individual user motion trajectories provided in an embodiment of the present invention.

[0021] Figure 5 A flowchart illustrating posture risk identification and correction prompts provided in an embodiment of the present invention. Detailed Implementation

[0022] The specific embodiments of the present invention will now be described with reference to the accompanying drawings. It should be understood that the following embodiments are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.

[0023] In one embodiment, the cordless jump rope includes a left handle, a right handle, a connecting rope, and a swing ball connected to the distal end of the connecting rope. The swing ball can be a gravity ball, a counterweight ball, or an adjustable weight ball. The system acquires the trajectory status of the swing ball during training through a swing ball trajectory acquisition module.

[0024] In one embodiment, the trajectory signal of the swing ball is acquired by an inertial sensor disposed inside the swing ball. The inertial sensor includes at least one of a three-axis accelerometer, a three-axis gyroscope, a magnetometer, or an attitude sensor, for acquiring the swing ball's acceleration, angular velocity, attitude change, trajectory period, and relative height change.

[0025] In another embodiment, the trajectory signal of the swing ball is calculated based on the handle inertia signal, the length of the connecting rope, the handle attitude, and the swing dynamics model. This method can estimate the relative height curve, trajectory plane change, and trajectory period of the swing ball without adding internal sensors to the swing ball.

[0026] In another embodiment, the trajectory signal of the swing ball is obtained through magnetic induction between the swing ball and the handle, distance sensing, Bluetooth ranging, ultra-wideband ranging, photoelectric detection, or a combination thereof. The system constructs the relative height curve of the swing ball, the radial swing amplitude, and the phase relationship between the left and right swing balls based on the relative distance change and time series.

[0027] In one embodiment, the handle inertia acquisition module is disposed within the cordless jump rope handle and is used to acquire triaxial acceleration, triaxial angular velocity, handle rotation period, peak angular velocity, peak acceleration, and wrist swing amplitude. The system identifies handle rotation events and handle period stability based on the handle inertia signals.

[0028] In one embodiment, the foot impact acquisition module is disposed in the foot, shoe clip, ankle ring, insole, sports mat, or mobile terminal, and is used to collect data on takeoff status, peak landing impact, impact difference between the left and right feet, landing cycle, and changes in foot acceleration. The foot impact acquisition module may include an acceleration sensor, a pressure sensor, an impact sensor, or a combination thereof.

[0029] In one embodiment, the time synchronization and feature extraction module maps the ball trajectory signal, handle inertia signal, and foot impact signal to a unified time axis. Synchronization methods include Bluetooth timestamp synchronization, unified time synchronization via mobile terminal, local clock offset correction, or post-processing alignment based on characteristic peaks.

[0030] In one embodiment, the trajectory quality evaluation module extracts trajectory closure, trajectory plane inclination, relative height peak-to-valley difference, and left-right swing phase difference from the swing ball trajectory signal. Trajectory closure characterizes whether the swing ball trajectory approaches a complete closed curve within a single motion cycle; trajectory plane inclination characterizes the degree of deviation of the swing trajectory relative to the normal swing plane; relative height peak-to-valley difference characterizes the vertical swing amplitude of the swing ball; and left-right swing phase difference characterizes the coordination of the left and right hand swings.

[0031] In one embodiment, the consistency verification module filters stable and valid motion samples based on the quality characteristics of the swing ball trajectory, the handle rotation event, and the foot landing event. If the current motion meets the quality conditions of the swing ball trajectory, the handle rotation event and the foot landing event satisfy the temporal matching relationship within the dynamic time window, and do not violate the prior constraints of motion safety, then the motion is marked as a stable and valid motion sample. If the current motion involves a dodge, a half-jump, a swing ball deviation, an incomplete trajectory, or an abnormal landing impact, then it is not used as a normal sample to update the user's individual motion trajectory baseline model.

[0032] In one embodiment, a baseline model of an individual user's motion trajectory is established based on stable and valid motion samples. This model can be implemented using statistical models, moving average models, exponential moving average models, clustering models, classification models, lightweight neural network models, or combinations thereof. The model records the range of normal ball-swinging trajectory closure, normal trajectory plane tilt, normal relative height peak-to-valley difference, normal left-right ball-swinging phase difference, normal hand-foot phase difference, normal landing impact range, and normal rhythm fluctuation range.

[0033] In one embodiment, the posture risk identification module compares the current action features with both motion safety prior constraints and the user's individual motion trajectory baseline model. If the current action violates the motion safety prior constraints, a posture risk warning is output; if the current action does not violate the motion safety prior constraints but its deviation from the user's individual motion trajectory baseline model exceeds the adaptive anomaly threshold, a personalized correction warning is output; if the current action is close to the individual baseline and does not violate the motion safety prior constraints, it is recorded as a stable action candidate sample.

[0034] In one embodiment, the self-learning update module generates candidate update results based on stable and valid action samples, user confirmation markers, prompt response results, or post-training action stability statistics. The safety boundary restriction module verifies the candidate update results; if a candidate update result violates the prior constraints on motion safety, it rejects the candidate update result or restricts it within the safety boundary.

[0035] In one embodiment, the correction prompt module outputs different prompts based on the type of risk. If the trajectory plane is too tilted, it prompts to reduce the swing ball's deflection; if the phase difference between the left and right swing balls is too large, it prompts to keep the left and right hands synchronized; if the trajectory closure is low, it prompts to maintain a complete swing cycle; if the hand and foot rhythm is mismatched, it prompts to keep the handle swing synchronized with the landing rhythm; if the landing impact peak is too high, it prompts to reduce the jump height or slow down the rhythm; if the trajectory fluctuation increases in the later stage of training, it prompts to reduce the jump frequency or pause for rest.

[0036] This invention does not collect images, sounds, or facial information, and does not determine posture based on camera body key points, video frames, or audio content. The output of this invention is for sports training assistance and posture risk warnings, and is not used for medical diagnosis or treatment effect assessment.

Claims

1. A method for identifying postural risks in ropeless jump rope based on the trajectory of a swinging ball, characterized in that, include: The process involves: acquiring the trajectory signal of a swing ball in a cordless jump rope, including at least the relative height curve of the swing ball, trajectory plane changes, and the phase relationship between the left and right swing balls; acquiring the inertial signal of the rope handle and the impact signal of the feet landing; performing time synchronization and motion feature extraction on the trajectory signal, handle inertial signal, and foot landing impact signal; extracting the trajectory quality features of the swing ball based on the trajectory signal, including trajectory closure, trajectory plane inclination, relative height peak-to-valley difference, and phase difference between the left and right swing balls; constructing prior constraints for motion safety; selecting stable and valid motion samples based on the consistency verification results of the trajectory quality features, handle rotation events, and foot landing events; and establishing a baseline model of the user's individual motion trajectory based on the stable and valid motion samples. During real-time training, the current action features are compared with the motion safety prior constraints and the user's individual action trajectory baseline model to identify posture risk states. Personalized correction prompts are output based on the recognition results; Furthermore, a safety boundary constraint mechanism is used to ensure that the update results of the baseline model of the individual user's motion trajectory do not exceed the prior constraints of motion safety.

2. The method according to claim 1, characterized in that, The trajectory signal of the swing ball is acquired by an inertial sensor installed inside the swing ball, or calculated based on the inertial signal of the handle, the length of the connecting rope, the handle posture and the swing dynamics model, or obtained through at least one of the following methods: magnetic induction between the swing ball and the handle, distance sensing, Bluetooth ranging, ultra-wideband ranging, photoelectric detection.

3. The method according to claim 1, characterized in that, The handle inertial signal includes three-axis acceleration, three-axis angular velocity, handle rotation period, peak angular velocity, peak acceleration, and wrist swing amplitude. The foot landing impact signal includes take-off state, landing impact peak, impact difference between left and right feet, landing period, and foot acceleration variation.

4. The method according to claim 1, characterized in that, The time synchronization and motion feature extraction include performing unified time axis mapping, sliding window segmentation, denoising, normalization, peak detection, motion cycle recognition, and phase relationship calculation on the swing ball trajectory signal, the handle inertial signal, and the foot landing impact signal.

5. The method according to claim 1, characterized in that, The prior constraints for sports safety include the upper limit of landing impact, the upper limit of impact difference between left and right feet, the upper limit of tilt of the trajectory plane of the swing ball, the upper limit of phase difference between left and right swing balls, the phase difference range between handle rotation event and foot landing event, the upper limit of the number of consecutive abnormal impacts, and the range of motion fluctuation after fatigue.

6. The method according to claim 1, characterized in that, The stable and valid motion sample is a motion sample that simultaneously satisfies the following conditions: the quality condition of the swing ball trajectory, the timing matching condition of the handle rotation event and the foot landing event within the dynamic time window, the periodic consistency condition of the handle rotation period and the foot landing period, and does not break the prior constraints of motion safety.

7. The method according to claim 1, characterized in that, The user's individual motion trajectory baseline model includes the range of normal ball swing trajectory closure, normal trajectory plane tilt range, normal relative height peak-to-valley difference range, normal left-right ball swing phase difference range, normal hand-foot phase difference range, normal landing impact range, and normal rhythm fluctuation range. Furthermore, the user's individual motion trajectory baseline model is updated based on stable and effective motion samples, user confirmation markers, prompt response results, or post-training motion stability statistics.

8. The method according to claim 1, characterized in that, The posture risk states include ball deviation, trajectory non-closed, abnormal swing height, asynchronous left and right ball swings, hand-foot rhythm mismatch, excessive upper limb compensation, empty swing, half jump, excessive landing impact, uneven force on the left and right sides, and movement deformation after fatigue. The safety boundary restriction mechanism includes rejecting the candidate update result or restricting the candidate update result to the sports safety prior constraints when the candidate update result of the user's individual movement trajectory baseline model causes the landing impact range, the left and right foot impact difference range, the swing ball trajectory plane tilt range, the left and right swing ball phase difference range, or the phase difference range between the handle rotation event and the foot landing event to exceed the corresponding sports safety prior constraints.

9. A cordless jump rope posture risk recognition device based on the trajectory of a swinging ball, characterized in that, The device is used to perform the method according to any one of claims 1 to 8, and includes: a swing ball trajectory acquisition module for acquiring the trajectory signal of a ropeless jump rope swing ball; a handle inertia acquisition module for acquiring the inertia signal of a ropeless jump rope handle; a foot impact acquisition module for acquiring the foot landing impact signal; a time synchronization and feature extraction module for performing time synchronization and motion feature extraction on the swing ball trajectory signal, handle inertia signal, and foot landing impact signal; a trajectory quality evaluation module for extracting trajectory closure, trajectory plane inclination, relative height peak-to-valley difference, and left and right swing ball phase difference; a priori constraint module for constructing motion safety prior constraints; a consistency verification module for screening stable and effective motion samples; an individual motion trajectory baseline model module for establishing a user individual motion trajectory baseline model; a posture risk identification module for identifying posture risk states; a self-learning update module for updating the user individual motion trajectory baseline model; a safety boundary restriction module for restricting the update result of the user individual motion trajectory baseline model from exceeding the motion safety prior constraints; and a correction prompt module for outputting personalized correction prompts.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 8.

Images