KR-20260062905-A - Method, Device and System for Multi-Body-Segment Sports Motion Analysis, Automatic Strength-Weakness Classification and Post-Session Report Generation Using Multiple Inertial Measurement Units

KR20260062905AKR 20260062905 AKR20260062905 AKR 20260062905AKR-20260062905-A

Abstract

The present invention relates to a method, apparatus, and system for collecting sports motion data through a plurality of IMUs worn on one or more body parts of a sports athlete or hobbyist, an IMU optionally coupled to a sports tool, and an impact detection means; continuously analyzing motion sequences and synchronization between multiple body parts, as well as tool-body relative posture, position, and trajectory over the entire stroke; automatically classifying strengths and weaknesses by comparing Z-scores with expert standard patterns or personalized standard values; and generating a post-analysis report. It is applicable to tool sports such as billiards, golf, tennis, baseball, and bowling, as well as physical sports such as martial arts and gymnastics. Unlike existing prior art which is limited to measuring the moment of impact or classifying shots, the present invention is differentiated in that it provides integrated continuous relative analysis over the entire stroke and a Z-score-based post-analysis report.

Inventors

강태욱

Assignees

강태욱

Dates

Publication Date: 20260507
Application Date: 20260330

Claims (16)

A step of simultaneously collecting continuous data of sports movements from multiple inertial measurement units (IMUs) worn on one or more parts of the body of a sports athlete or hobby user; A step of synchronizing the above plurality of IMU data based on a common time standard; A step of automatically detecting the start time of movement for each body part and analyzing the movement sequence and synchronization relationship between multiple body parts; A step of calculating the Z-score of feature values by operation interval by comparing with expert standard patterns or personalized standard values, and automatically classifying strength items and improvement items; and A step of generating a post-hoc analysis report including classification results, sequence and alignment analysis results, and action-result correlations; A sports motion analysis method including
Multiple IMU modules worn on one or more parts of the body of a sports athlete or hobbyist; A tool attachment module that is optionally detachably coupled to a sports tool and includes a tool IMU and an impact detection means; A processing unit that receives and synchronizes the above-mentioned plurality of IMU data via wireless communication, analyzes the sequence and synchronization between multiple body parts, classifies strengths and weaknesses based on Z-scores relative to a reference pattern, and generates a post-analysis report; and A communication unit that transmits analysis results to a user terminal; A sports motion analysis device including
Sports motion analysis device according to claim 2; and A user terminal app that displays a post-analysis report including a session summary, pros and cons by checkpoint, multiple body part sequence and synchronization, tool-body alignment time series, trajectory comparison diagram, motion timeline, cumulative growth trend, and AI coaching messages; A sports motion analysis system including
A sports motion analysis method according to claim 1, wherein the sequence analysis automatically detects the motion start time (t_onset_i) that first exceeds a threshold value in the acceleration magnitude time series of each body part, calculates the start order and time interval (?t_ij) between parts, and classifies deviations by comparing with a reference sequence.
A sports motion analysis method according to claim 4, wherein the time interval (?t = t_wrist - t_ankle) between the ankle weight transfer start time (t_ankle) and the wrist downswing start time (t_wrist) is calculated when applying a golf swing, and the case where ?t < 0 is classified as an upper body leading error.
A sports motion analysis method according to claim 1, wherein the synchronization analysis measures coordination by calculating the Pearson correlation coefficient (r_sync) between the velocity vector time series of multiple body parts.
A sports motion analysis method according to claim 1, further comprising a tool IMU coupled to a sports tool and an impact detection means, and continuously calculating the relative posture (Q_rel = Q_body^-1 ? Q_tool), relative position change (d_rel), and trajectory synchronization (r_traj) of the entire stroke section from the quaternions of the body IMU and the tool IMU.
A sports motion analysis method according to claim 7, wherein the alignment breakdown point (t_break) at which the relative posture deviation first exceeds a reference threshold (?_threshold) is calculated and displayed in a post-analysis report.
A sports motion analysis method according to claim 7, wherein the impact sensing means comprises one or more of a piezoelectric sensor, a strain gauge, a high-G accelerometer, and a combination thereof.
A sports motion analysis method according to claim 1, wherein the skilled user standard pattern is set as the expected value and standard deviation calculated from multiple session data of skilled users wearing the same IMU device, and the personalized standard value is automatically updated as the expected value and standard deviation of the top 30% performance data of the individual user's last N sessions.
A sports motion analysis method according to claim 1, wherein the advantage item is classified as |Z_i| ≤ 1.0 and the improvement item as |Z_i| > 1.5, and the improvement priority is calculated by applying a motion-result Pearson correlation coefficient-based weight (W_i=1.5 if |r_corr| ≥ 0.5).
A sports motion analysis method according to claim 1, wherein the IMU operates in a game rotation vector mode using only an accelerometer and a gyroscope, excluding a magnetometer, in an electromagnetic interference environment.
A sports motion analysis method according to claim 1, wherein the post-analysis report includes a timeline visualizing the entire motion in score-based colors, a graph of growth trends across multiple sessions, and a one-sentence execution guideline for the item with the highest priority for improvement.
A sports motion analysis device according to claim 2, wherein the tool attachment module is detachably coupled to any one of a billiard cue, a golf club, a tennis racket, a baseball bat, or a bowling finger ring with a weight of 20g or less.
A sports motion analysis system according to claim 3, wherein the user terminal app supports one or more sport modes among billiards, golf, tennis, baseball batting, bowling, martial arts, and gymnastics, and when a sport is selected, checkpoint items, reference patterns, and analysis algorithms are automatically switched.
A sports motion analysis method according to claim 1, wherein at least one of the plurality of IMUs is worn on the wrist and at least one is worn on the ankle to analyze the sequence and synchronization relationship of upper and lower body movements.

Description

Method, Device and System for Multi-Body-Segment Sports Motion Analysis, Automatic Strength-Weakness Classification and Post-Session Report Generation Using Multiple Inertial Measurement Units The present invention relates to sports motion analysis technology, and more specifically, to a method, apparatus, and system for collecting sports motion data through a plurality of inertial measurement units (IMUs) worn on one or more parts of the body of a sports athlete or hobbyist, an IMU optionally coupled to a sports tool, and an impact detection means; continuously analyzing motion sequences, synchronization relationships between multiple body parts, and relative posture, position, and trajectory based on a tool-body dual IMU across the entire stroke by comparing them with expert standard patterns or personalized standard values; automatically classifying strengths and weaknesses based on Z-scores; and generating a post-analysis report. The present invention is applicable to tool sports such as billiards, golf, tennis, baseball, and bowling, as well as physical sports such as martial arts and gymnastics. In sports, the accuracy of movement is a key factor in determining athletic performance. Hobby and amateur users find it difficult to identify which parts of their movements are correct and which are incorrect, and guidance from professional coaches is not always available. To address this, wearable IMU-based sports motion analysis technology has been researched, but conventional technology has the following significant limitations. [Prior Art 1] Billiards Field Korean Registered Utility Model KR20180000932U (Billiard Cue for Posture Correction) is limited to an aiming correction method that illuminates the first object ball with a laser pointer, and completely lacks IMU sensor-based motion data collection, analysis, and feedback functions. Chinese Utility Model CN210051520U (Billiard Cue Performance Measurement Platform) is a laboratory-mounted device, is not wearable, and cannot analyze body movements. In the field of billiards, no IMU-based wrist-cue relative posture analysis system exists domestically or internationally. [Prior Art 2] Golf Field - Most Prior Art There are numerous prior technologies in the field of golf. U.S. Patent US10709945B2 (Golf Swing Monitoring System) discloses a system for collecting golf swing data and detecting impact using a wrist-worn wearable. However, this patent is limited to the detection of impact events and the logging of swing data, and does not disclose continuous analysis of alignment angles, relative position changes, and trajectory deviations over the entire stroke using the relative quaternion of a tool-body dual IMU. U.S. Patent US12285667 (Club Face IMU Attachment Device, registered April 2025) relates to a method of measuring loft and lie angles by attaching an IMU to a club face, and is unrelated to data synchronization with a body-worn IMU and continuous analysis of relative posture. The technology developed by the Stanford School of Medicine research team (two upper and lower spinal IMUs, technology transfer in progress) is specialized for measuring spinal rotation parameters such as S-factor, O-factor, and X-factor, but does not include analysis of wrist and ankle motion sequences or continuous analysis of club-wrist relative posture. In addition, golf swing segment classification studies using a single wrist IMU (Kim & Park, 2020; Scientific Reports, 2024) are limited to motion measurements of a single body part and do not cover sequence analysis between multiple body parts or tool-body relative analysis. [Prior Art 3] Tennis Field Wrist-worn IMU products (such as Armbeep and Babolat Pop) measure the number of shots, wrist speed, and impact strength, but they do not have the function to continuously analyze the relative posture between the racket and the wrist throughout the entire stroke. Although a study (Agarwal & Vishwakarma, 2025) was published in which IMUs were mounted on the wrist and racket respectively to collect data with ESP32 and classify shot types with LSTM, this is limited to shot classification and does not disclose continuous analysis of alignment angle, position change, and trajectory deviation over the entire range using racket-wrist relative quaternions, automatic classification of strengths and weaknesses based on Z-score, and generation of post-hoc analysis reports. [Prior Art 4] Baseball Field The study on baseball pitching analysis using five IMUs (wrist, forearm, upper arm, chest, and waist) (Lapinski et al., 2019) is a specialized research device intended for the purpose of predicting injury risk and is not a low-cost wearable system for analyzing hobbyist or amateur batting. Furthermore, wrist-ankle motion sequence analysis and bat-wrist relative posture continuous analysis in batting have not been disclosed in prior art. [Prior Art 5] Common Limitations of Existing Prior Art All of the above prior art has the following common li