WO2017181717A1 - 电子教练实现方法及系统 - Google Patents

电子教练实现方法及系统 Download PDF

Info

Publication number
WO2017181717A1
WO2017181717A1 PCT/CN2016/112255 CN2016112255W WO2017181717A1 WO 2017181717 A1 WO2017181717 A1 WO 2017181717A1 CN 2016112255 W CN2016112255 W CN 2016112255W WO 2017181717 A1 WO2017181717 A1 WO 2017181717A1
Authority
WO
WIPO (PCT)
Prior art keywords
user
motion
current
posture
exercise
Prior art date
Application number
PCT/CN2016/112255
Other languages
English (en)
French (fr)
Inventor
罗合见
师丹玮
Original Assignee
深圳泰山体育科技股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳泰山体育科技股份有限公司 filed Critical 深圳泰山体育科技股份有限公司
Publication of WO2017181717A1 publication Critical patent/WO2017181717A1/zh

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A63—SPORTS; GAMES; AMUSEMENTS
    • A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
    • A63B24/00—Electric or electronic controls for exercising apparatus of preceding groups; Controlling or monitoring of exercises, sportive games, training or athletic performances
    • A—HUMAN NECESSITIES
    • A63—SPORTS; GAMES; AMUSEMENTS
    • A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
    • A63B24/00—Electric or electronic controls for exercising apparatus of preceding groups; Controlling or monitoring of exercises, sportive games, training or athletic performances
    • A63B24/0003—Analysing the course of a movement or motion sequences during an exercise or trainings sequence, e.g. swing for golf or tennis
    • A63B24/0006—Computerised comparison for qualitative assessment of motion sequences or the course of a movement
    • A—HUMAN NECESSITIES
    • A63—SPORTS; GAMES; AMUSEMENTS
    • A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
    • A63B24/00—Electric or electronic controls for exercising apparatus of preceding groups; Controlling or monitoring of exercises, sportive games, training or athletic performances
    • A63B24/0062—Monitoring athletic performances, e.g. for determining the work of a user on an exercise apparatus, the completed jogging or cycling distance

Definitions

  • the invention relates to the field of human-computer interaction, in particular to an electronic coach implementation method and system.
  • the invention provides a method and a system for implementing an electronic coach, which solves the problem that the fitness guidance cannot be performed in real time in the prior art, and whether the movement of the athlete is standard in real time is determined in the exercise.
  • an electronic coach implementation method including:
  • S1 Obtain identity information of the logged-in user, and retrieve an action plan that matches the identity of the user; the exercise plan includes each exercise guide information that is sequentially arranged;
  • step S5 the standard posture guidance information of the current action is fed back to the user, guiding the user to redo the current action, and then performing step S3;
  • each motion guidance information includes the number of allowed repetitions of the current motion.
  • the method further includes: checking whether the number of times the user repeats the current motion has reached the preset allowed repetition number, and if so, It is determined that the user has completed the motion guidance information of this section, and then step S6 is performed; if not, step S5 is performed.
  • comparing the current action gesture with the pre-stored standard motion gesture in the step S4 includes:
  • the key joint points on the human body are determined, and the rotation angles of the key joint points in the current motion posture are calculated;
  • the rotation angle of each key joint point in the current motion posture is compared with the rotation angle of the joint point corresponding to the pre-stored standard motion posture.
  • feeding back the standard posture guidance information of the current action to the user in step S5 includes: playing the video to correct the current operation error of the user, and demonstrating the standard action posture.
  • step S6 it is determined in step S6 that if the current exercise ends, the user also records the motion information to update the motion plan accordingly.
  • the invention also provides an electronic coach implementation system, comprising:
  • a login unit configured to obtain identity information of the login user, and retrieve a motion plan that matches the identity of the user; the motion plan includes each exercise guide information that is sequentially arranged;
  • a motion guiding unit configured to guide user motion from the first piece of motion information from a sequence of motion guidance information that has not been executed
  • a shooting unit for taking a moving image of the user
  • the action determining unit is configured to extract a current motion posture in the user motion image captured by the imaging unit, compare the current motion posture with the pre-stored standard motion posture, and determine the relationship between the two Whether the difference is within the preset range; if otherwise, feedback to the user action error reminder information and activate the gesture guiding unit, if yes, start the progress unit;
  • the posture guiding unit feeds back the standard posture guiding information of the current action to the user, and guides the user to redo the current action;
  • the progress unit determines whether the current execution is the last motion guidance information sent by the motion guidance unit, and if so, determines that the motion is ended, if otherwise, the motion guidance unit is activated to guide the user to continue the motion.
  • each motion guidance information includes an allowed repetition number of the current motion
  • the electronic coach implementation system further includes a counting unit, and the motion determination unit starts the counting unit before starting the posture guiding unit,
  • the counting unit is configured to check whether the number of times the user repeats the current action has reached the preset allowable repetition number, and if yes, determine that the user has completed the motion guidance information of the article and start the progress unit; if not, the posture guiding unit is activated. .
  • the action determining unit includes:
  • a node determining module for determining a key joint point on the human body according to the current action
  • a rotation angle calculation module for calculating a rotation angle of each key joint point in the current motion posture
  • the comparison determination module is configured to compare the rotation angle of each key joint point in the current motion posture with the rotation angle of the joint point corresponding to the pre-stored standard motion posture, and determine whether the difference value is within a preset fault tolerance range.
  • the action determining unit includes an error prompting module for playing a video to correct a user's current motion error.
  • the electronic coach implementation system further includes a recording module, configured to record the user's motion information after the progress unit determines that the motion ends, to perform the motion scheme update accordingly.
  • the beneficial effects of the present invention are: the electronic coach implementation method and system provided by the present invention are obtained After the user identity information, the user's matching exercise plan is retrieved, and the user is allowed to exercise according to the exercise plan. During the movement process, the image shooting method is used to compare the user's motion movement with the pre-stored standard motion, and correct the non-standard motion in real time.
  • an electronic coach is provided for the bodybuilder by means of visual recognition, and the action of the exerciser in the movement can be automatically compared and corrected, thereby real-time guiding the fitness activity according to the exercise prescription.
  • FIG. 1 is a flow chart of a preferred embodiment of an electronic coach implementation method of the present invention
  • FIG. 2 is a block diagram of a preferred embodiment of an electronic coaching implementation system of the present invention
  • Figure 3 is a schematic view of the main joint of the human body
  • Figure 4a is a schematic view of the joint of the left hand portion of the human body in a horizontal state
  • Figure 4b is a schematic view of the joint of the left elbow joint of the human body after rotation
  • Figure 5a is a schematic view of a human joint of a Taiji movement in a standard posture
  • Figure 5b is a schematic view of the human joint that the user performs the Tai Chi exercise.
  • the electronic coach implementation system of the present invention comprises a registration unit 1, a motion guiding unit 2, a shooting unit 3, an action determining unit 4, a posture guiding unit 5 and a progress unit 6, if necessary,
  • a counting unit 7 can also be included.
  • step S1 ie
  • S1 Obtain identity information of the logged-in user, and retrieve an action plan that matches the identity of the user.
  • the login unit 1 obtains the identity information of the user, and the identity information includes information such as the user's ID card, a face, and the like that can distinguish the identity of the user.
  • the login unit 1 can be an identity scanner, a face scanner or an integration of a plurality of identity information scanning instruments.
  • the identity information has been registered and stored, and under the user identity, an activity plan matching its identity is stored. Stored data is stored in the cloud or on a local server.
  • the identity information such as the obtained ID card and face is compared with the stored identity information, and after confirming the identity, the identity of the user is retrieved.
  • Exercise program The exercise plan includes each exercise guide information arranged in sequence.
  • the exercise plan is a targeted fitness plan issued for the user's physical condition. Based on the exercise plan, the exercise guide information is decomposed and generated. For example, the exercise program is to complete a set of 42-style Tai Chi exercise. After decomposing, the motion guidance information formed is: "The following actions are performed in sequence: the starting position (standing on the foot) - the starting position (the left foot is open) - the starting position (the knees pressing the palm) -.
  • the closing trend by breaking up the hand) - closing the trend (receiving the feet and standing)"; and as the sports program for the rowing machine, the exercise lasts for 30 minutes.
  • the exercise guidance message is: "The whole body endurance exercise is performed for 5 minutes in turn (the legs are straight, the waist is reclined, and the arms are stretched back to the sides) -.
  • - the lower body is vigorously exercised for 5 minutes ( With your arms stretched out, your back will be tilted back and your legs will be bent straight.”
  • step S2 After the login unit 1 retrieves the motion plan of the logged-in user, after the user prepares for the exercise, the system performs step S2, that is,
  • the motion guidance unit 2 guides the user motion from the first motion guidance information from the motion guidance information sequence that has not been executed.
  • the exercise program Take the exercise program as an example of “complete a set of 42-style Taijiquan exercise”, if the user has completed the exercise guidance letter In the “starting (and standing)" action in the interest, in the sequence of motion guidance information that has not been executed, the first motion guidance information is "starting (left foot open)", then the motion guiding unit 2 starts guiding the user Do the "moving action (left foot open)" movement.
  • step S3 ie
  • the shooting of the user's moving image is performed by the photographing unit 3, which may be an image sensor of a charge coupled device (CCD) or a complementary metal oxide semiconductor (CMOS) type.
  • the photographing unit 3 may be an image sensor of a charge coupled device (CCD) or a complementary metal oxide semiconductor (CMOS) type.
  • CCD charge coupled device
  • CMOS complementary metal oxide semiconductor
  • the captured image information is transmitted to the action determining unit 4 via the imaging unit 3, at which time the system executes step S4, ie
  • step S4 Extracting the current motion posture in the captured user motion image, comparing the current motion posture with the pre-stored standard motion posture, and determining whether the difference between the two is within a preset range; if not, feeding back to the user action
  • the error reminding message is executed and step S5 is performed; if yes, step S6 is performed.
  • the motion determining unit 4 extracts the current motion of the user from the user motion image captured by the imaging unit 3, and compares the current motion gesture with the pre-stored standard motion gesture to determine whether the difference between the two is within a preset range. Regarding how to extract the action gesture from the moving image, there is a mature technical solution in the field of image processing, which will not be described here.
  • the joint relationship of the human body is first described. As shown in FIG. 3, the joints of the human body are divided into five parts:
  • Body part middle of the sac, the spine, the center of the shoulder, the head, and the middle part of the body is the middle part of the body.
  • the sacral is the parent joint of the spine
  • the spine is the parent joint of the shoulder center
  • the shoulder center is the parent joint of the head, that is, the parent joint of each joint is the former joint
  • the shoulder center is the parent joint of the left shoulder and the right shoulder.
  • the father's joints are left and right.
  • Left hand left shoulder, left elbow, left wrist, where the left shoulder is the root joint of the left hand, and the parent joint of each joint is the previous joint.
  • Left leg left ankle, left knee, left ankle, left foot, wherein the left ankle is the root joint of the left leg, and the parent joint of each joint is the former joint.
  • Right leg right ankle, right knee, right ankle, right foot, where the right ankle is the root joint of the right leg, and the parent joint of each joint is the former joint.
  • Human motion is essentially the motion of each joint, that is, the rotation of each joint relative to the parent joint, and the displacement in the jaw.
  • the rotation of the parent joint rotates with all its descendant joints (relative to the joint rotated by the parent joint) without affecting the rotation of the other parent joints; the joints of the joints are part of the joint, and if only the joint rotates, The descendant joint itself does not rotate, and its relationship is as follows:
  • FIG. 4a taking the left hand movement of the human body as an example.
  • the left hand is in a horizontal state
  • the A point represents the left shoulder
  • the B point represents the left elbow
  • the C point represents the left wrist.
  • a three-dimensional Cartesian coordinate system is established, the X axis is along the horizontal direction of the left hand, the Y axis is the vertical left hand direction, and the Z axis is the vertical XY plane direction.
  • the angle of rotation of the joint is represented by (x, y, z), where x, y, z represent the angle of rotation of the joint relative to its parent joint about the x, y, and z axes, respectively.
  • the rotation angle information of the three points A, B, and C are A(0,0,0), B(0,0,0), C(0,0,0, respectively. ).
  • the left hand is rotated by the left elbow in the horizontal state, specifically, the left elbow is rotated 45 degrees around the Y axis, and rotated 45 degrees around the Z axis, as shown in FIG. 4b, then the rotation angles of the three points A, B, and C are The information changes to A(0,0,0), B(0,45,45), C(0,0,0), respectively, that is, the left wrist of the left elbow joint is not rotated.
  • the action determining unit 4 compares the current action posture with the standard motion posture, and determines whether the motion meets the requirements.
  • the basic principle is to determine the key joint points on the human body according to the current motion, and then detect the rotation and standard of the key joint points. Whether the motion posture corresponds to the error of the rotation of the joint is within the specified tolerance range.
  • the action determining unit 4 specifically includes a joint point determining module (not shown) for determining a key joint point on the human body according to the current motion; a rotation angle calculating module (not shown) , used to calculate the rotation of each key joint point in the current action posture a rotation angle; and a comparison determination module (not shown) for comparing a rotation angle of each key joint point in the current motion posture with a rotation angle of a joint point corresponding to the pre-stored standard motion posture, and determining the difference Whether the value is within the preset tolerance range.
  • Pi.x, Pi.y, Pi.z respectively represent the rotation angle of the i-th joint in the key joint of the user's motion relative to the parent joint about the x-axis, the y-axis, and the z-axis;
  • Ti.x, Ti. y, Ti.z respectively indicate the rotation angle of the i-th joint of the standard motion key joint with respect to the parent joint about the x-axis, the y-axis, and the z-axis;
  • error indicates the error of the user's action and the standard motion;
  • Indicates the most allowable fault tolerance threshold expressed by Tolerance.
  • the standard of the left hand is taken as an example.
  • the off node determination module determines that the left hand key nodes are the left shoulder, the left elbow, and the left wrist.
  • the rotation angle of point a in the left shoulder is (47.6, 40.2, 11.1)
  • the angle of rotation of point b in the left elbow is (-53.2, -20.7, -3.9).
  • the left wrist is the rotation angle of point c in the figure (4.3, -9.8, 23.2); when the action made by the user is as shown in Fig.
  • the rotation angle calculation module detects the rotation angle of the left shoulder, that is, the point a' in the figure. (28.8, 22.1, 2.5), the angle of rotation of the left elbow, that is, the point b' in the figure is (-54.5, -20.2, -8.1), and the rotation angle of the c' point in the left wrist is (21.9, 8.5, 30.8); After the calculation of the comparison decision module,
  • the error on the left shoulder is:
  • the error of the left elbow is:
  • the error of the left wrist is:
  • the error of the user's current action and the standard action posture calculated by the comparison determination module is: 27.4774 .
  • the motion error of the user's right hand, left leg, right leg and body part is successively determined, and then with the preset fault tolerance value.
  • the action determining unit 4 further includes an error prompting module (not shown), and the error prompting module may be a display for playing the video to correct the current operation error of the user, and in addition, the login may be displayed.
  • the system separately performs different processing:
  • step S5 is performed, ie,
  • the gesture guiding unit 5 will perform the current action.
  • the standard attitude guidance information is fed back to the user, guiding the user to redo the current action.
  • the system further includes performing step S41 before performing step S5, that is,
  • the counting unit 7 in the system is used to check whether the number of times the user repeats the current action has reached a preset number of times.
  • the motion determination unit 4 starts the counting unit 7 before starting the posture guiding unit 5, and repeats the number of times the current motion is performed for the user.
  • the preset allowed number of repetitions has been reached for inspection.
  • the number of allowed repetitions of the current action may be preset in each motion guidance information.
  • the check by the counting unit 7 may avoid the occurrence of the preset standard motion posture when the user repeats the current motion.
  • the current action is repeated continuously and the next action cannot be entered. In this way, even if the user can not reach the preset standard action posture by performing a certain motion instruction, the next motion guide can be continued, and the next motion action can be performed, thereby bringing a better user experience.
  • step S6 When it is detected that the difference between the current action posture and the standard motion posture of the user is within a preset fault tolerance range, that is, when the current motion standard of the user is determined, the system performs step S6, that is,
  • step S6 Determine whether the current exercise instruction is the last exercise instruction information. If yes, go to step S7; if no, go back to step S2.
  • the progress unit 6 determines whether the currently executed motion guidance information is the last motion guidance information issued by the motion guiding unit 4, and if so, determines that the current motion is over. For example, when the user completes the last movement instruction information in the forty-two style Taijiquan exercise program, the action of “receiving the foot”, it is determined that the user has completed the exercise plan “completed a set of forty Two-style Taijiquan exercise.” Conversely, if it is determined that the last instruction information is not currently being executed, the motion guiding unit 4 is activated to guide the user to continue the motion, that is, the first motion guidance information is executed from the motion guidance information sequence that has not been executed.
  • the action posture completes the movement guidance information "receive the potential (breaking down)" action in the 42-style Taijiquan exercise program. At this time, it is determined that the current execution is not the last motion guidance information, and the motion guidance information sequence that has not been executed yet. In the middle, there is also the movement guidance information of “receiving the situation (receiving the feet and standing together)”. Therefore, the user is continuously guided to do the action of “receiving the situation (receiving the feet and standing together)”.
  • step S6 determines in step S6 that if the current exercise ends, the system further records the user's motion information to update the motion plan accordingly.
  • the system further includes a recording module (not shown) for recording the user's movement information after the end of the user's current motion, uploading to the local or cloud server, and exercising the motion plan Update.
  • a recording module (not shown) for recording the user's movement information after the end of the user's current motion, uploading to the local or cloud server, and exercising the motion plan Update.
  • the electronic coach implementation method and system provided by the present invention after acquiring the user identity information, retrieves the user-matched exercise plan, and allows the user to exercise according to the exercise plan.
  • the image shooting method is used to compare the user's motion movement with the pre-stored standard motion, and correct the non-standard motion in real time.
  • an electronic coach is provided for the bodybuilder by means of visual recognition, and the action of the exerciser in the movement can be automatically compared and corrected, thereby real-time guiding the fitness activity according to the exercise prescription.

Landscapes

  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Physical Education & Sports Medicine (AREA)
  • Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
  • Medical Treatment And Welfare Office Work (AREA)

Abstract

电子教练实现方法及系统包括：S1、获取登录用户的身份信息，调取与用户身份相匹配的运动方案；S2、从尚未执行的运动指导信息序列中，自首条运动指导信息开始引导用户运动；S3、拍摄用户运动图像；S4、提取用户当前动作姿态并与标准动作姿态进作比较，判断两者的差异是否在预设范围内；否则反馈给用户动作错误提醒信息并执行步骤S5；是则执行步骤S6；S5、将当前动作的标准姿态信息反馈给用户，引导用户重做当前动作，然后执行步骤S3；S6、判定当前执行的是否是最后一条运动指导信息，是则判定本次运动结束；否则返回执行步骤S2。该方法和系统可以对健身者运动中的动作进行自动比对和纠正，实现实时指导健身。

Description

电子教练实现方法及系统 技术领域
本发明涉及人机互动领域，特别涉及一种电子教练实现方法及系统。
背景技术
随着生活水平的逐步提高，人们越来越重视自身健康。其中，通过运动健身的方式来促进身体健康是最根本的途径。
与开展无序、无计划性的自行健身相比，当用户在专业健身教练的指导下，开展针对性的健身活动时，可以获得更好的健身效果。但现有健身指导方式需要用户到专业的健身房，通过与教练之间进行面对面的交流，获得健身指导，且存在以下不足：
(1)用户在运动过程中无法随时获得教练的指导与督促；
(2)由于教练不是实时指导，用户无法实时知道自己的动作是否标准；
现有健身指导方式存在的不足，大大制约了人们参与健身活动的积极性，因此，在健身活动中，提供一种智能化的健身指导方式是有必要的。
发明内容
本发明提供一种电子教练实现方法及系统，解决现有技术中无法实时开展健身指导，以及在运动中对运动者动作是否标准实时作出判断的问题。
为解决上述技术问题，本发明提供了一种电子教练实现方法，包括：
S1、获取登录用户的身份信息，调取与该用户身份相匹配的运动方案；所述运动方案包括依序排列的各运动指导信息；
S2、从尚未执行的运动指导信息序列中，自首条运动指导信息开始引导用户运动；
S3、拍摄用户运动图像；
S4、提取出拍摄到的用户运动图像中的当前动作姿态，将当前动作姿态与预存的标准动作姿态进行比较，判断两者之间的差异是否在预设范围内；若否，反馈给用户动作错误提醒信息并执行步骤S5；若是，则执行步骤S6；
S5、将当前动作的标准姿态引导信息反馈给用户，引导用户重做当前动作，然后执行步骤S3；
S6、判定当前执行的是否是最后一条运动指导信息，若是，则判定本次运动结束；若否，则返回执行步骤S2。
进一步，每条运动指导信息内包含了当前动作的允许重复次数，所述步骤S4中执行步骤S6之前还包括：检查用户重复做当前动作的次数是否已达到预设的允许重复次数，若是，则判定用户已执行完毕本条运动指导信息，接着执行步骤S6；若否，则执行步骤S5。
进一步，所述步骤S4中将当前动作姿态与预存的标准动作姿态进行比较包括：
根据当前动作定出人体上的关键关节点，计算出当前动作姿态中各关键关节点的旋转角度；
将当前动作姿态中各关键关节点的旋转角度与预存的标准动作姿态中相对应的关节点的旋转角度进行比较。
进一步，步骤S5中将当前动作的标准姿态引导信息反馈给用户包括：播放视频纠正出用户当前动作错误之处，并演示标准动作姿态。
进一步，步骤S6中判定若本次运动结束之后还包括，对用户该次运动信息进行记录以据此进行运动方案更新。
本发明还提供了一种电子教练实现系统，包括：
登录单元，用于获取登陆用户的身份信息，并调取与该用户身份相匹配的运动方案；所述运动方案包括依序排列的各运动指导信息；
运动指导单元，用于从尚未执行的运动指导信息序列中，自首条运动信息开始引导用户运动；
拍摄单元，用于拍摄用户运动图像；
动作判定单元，用于提取出拍摄单元拍摄到的用户运动图像中的当前动作姿态，将当前动作姿态与预存的标准动作姿态进行比较，判断两者之间的 差异是否在预设范围内；若否则反馈给用户动作错误提醒信息并启动姿势引导单元，若是则启动进度单元；
姿势引导单元，将当前动作的标准姿态引导信息反馈给用户，引导用户重做当前动作；
进度单元，判定当前执行的是否是运动指导单元所发出的最后一条运动指导信息，若是则判定本次运动结束，若否则启动运动指导单元以引导用户继续运动。
进一步，每条运动指导信息内包含了当前动作的允许重复次数，所述电子教练实现系统还包括一计数单元，所述动作判定单元在启动姿势引导单元之前先启动所述计数单元，
所述计数单元用于检查用户重复做当前动作的次数是否已达到预设的允许重复次数，若是，则判定用户已执行完毕本条运动指导信息，并启动进度单元；若否，则启动姿势引导单元。
进一步，所述动作判定单元包括：
关节点确定模块，用于根据当前动作定出人体上的关键关节点；
旋转角度计算模块，用于计算出当前动作姿态中各关键关节点的旋转角度；以及
比较判定模块，用于将当前动作姿态中各关键关节点的旋转角度与预存的标准动作姿态中相对应的关节点的旋转角度进行比较，判断差异值是否在预设的容错范围内。
进一步，所述动作判定单元包括错误提示模块，用于播放视频纠正出用户当前动作错误之处。
进一步，所述电子教练实现系统还包括一记录模块，用于在进度单元判定本次运动结束后，对用户该次运动信息进行记录以据此进行运动方案更新。
本发明的有益效果是：本发明提供的电子教练实现方法及系统，在获取 用户身份信息后，调取出用户匹配的运动方案，让用户根据运动方案做运动。运动过程中，采用图像拍摄的方式，对用户运动动作与预存的标准动作进行比较，实时纠正不标准动作。采用本发明方案，通过视觉识别的方式，为健身者提供了一名电子教练，可以对健身者运动中的动作进行自动比对和纠正，从而实时指导其按照运动处方开展健身活动。
附图说明
为了更清楚地说明本发明实施例中的技术方案，下面将对实施例描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图，其中：
图1是本发明电子教练实现方法的一个优选实施方式的流程图；
图2是本发明电子教练实现系统一个优选实施例的模块图；
图3是人体主要关节示意图；
图4a是人体左手部处于水平状态的关节示意图；
图4b是人体左肘关节旋转后的关节示意图；
图5a是标准姿态下的太极起势动作的人体关节示意图；
图5b是用户做太极起势动作的人体关节示意图。
具体实施方式
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅是本发明的一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
请参阅图2，本发明电子教练实现系统，包括登录单元1、运动指导单元2，拍摄单元3，动作判定单元4，姿势引导单元5和进度单元6，若有必要， 还可以包括计数单元7。
请结合参阅图1，用户通过登录单元1进入电子教练实现系统。此时，系统进行步骤S1，即
S1、获取登录用户的身份信息，调取与该用户身份相匹配的运动方案。
登录单元1获取用户的身份信息，该身份信息包括用户的身份证、人脸等可以区分用户身份的信息。相应的，登录单元1可以是身份扫描器、人脸扫描器或者是多种身份信息扫描仪器的集成体。该身份信息已经进行过注册和存储，并在该用户身份下，存储有与其身份相匹配的运动方案。存储数据储存在云端或本地服务器中。当用户通过身份证、人脸扫描等方式进行登陆时，通过对获取的身份证、人脸等身份信息与存储的身份信息进行比对，确认身份后，调取出与该用户身份相匹配的运动方案。该运动方案包括依序排列的各运动指导信息。运动方案是针对用户身体体质状况出具的针对性健身计划，基于该健身计划，分解生成各运动指导信息。如该运动方案为完成一套四十二式太极拳锻炼。经过分解后，形成的运动指导信息为，“依次进行如下动作：起势(并脚站立)—起势(左脚开立)—起势(屈膝按掌)—.......—收势(分手下按)—收势(收脚并立)”；又如该运动方案为划船器锻炼，锻炼时长30分钟。经过分解后，运动指导信息为，“依次进行全身耐力锻炼5分钟(腿部伸直，腰部后仰，双臂向两侧后划)—.......—下半身爆发力锻炼5分钟(双臂伸直后腰部后仰，双腿弯曲蹬直)”。
在登录单元1调取出登录用户的运动方案，用户做好运动准备后，系统执行步骤S2，即
S2、从尚未执行的运动指导信息序列中，自首条运动指导信息开始引导用户运动。
根据登录单元1调取出的登录用户的运动方案，运动指导单元2从尚未执行的运动指导信息序列中，自首条运动指导信息开始引导用户运动。以运动方案为“完成一套四十二式太极拳锻炼”为例，若用户已完成运动指导信 息中的“起势(并脚站立)”动作，则在尚未执行的运动指导信息序列中，首条运动指导信息为“起势(左脚开立)”，则运动指导单元2开始引导用户做“起势(左脚开立)”这一运动动作。
在用户根据运动指导单元2引导做运动过程中，系统执行步骤S3，即
S3、拍摄用户运动图像。
拍摄用户运动图像由拍摄单元3完成，该拍摄单元3可以是采用电荷耦合器件(CCD)或互补金属氧化物半导体(CMOS)类型的图像传感器。
拍摄得到的图像信息经拍摄单元3传输给动作判定单元4，此时，系统执行步骤S4，即
S4、提取出拍摄到的用户运动图像中的当前动作姿态，将当前动作姿态与预存的标准动作姿态进行比较，判断两者之间的差异是否在预设范围内；若否，反馈给用户动作错误提醒信息并执行步骤S5；若是，则执行步骤S6。
动作判定单元4从拍摄单元3拍摄到的用户运动图像中提取出用户当前动作，并将当前动作姿态与预存的标准动作姿态进行比较，判断两者之间的差异是否在预设范围内。关于如何从运动图像中提取出动作姿态，在图像处理领域已有成熟的技术方案，在这里不再赘述。
在阐述动作判定单元4如何将提取出的用户当前动作姿态与标准动作姿态进行比较之前，先对人体的关节关系进行说明，如图3所示，人体的关节分为五大部分：
身体部：胯中、脊柱、肩中心、头部，其中胯中为身体部的根关节。另外，胯中是脊柱的父关节，脊柱是肩中心的父关节，肩中心是头部的父关节，即后面每个关节的父关节为前一个关节；肩中心为左肩和右肩的父关节、胯中为左胯和右胯的父关节。
左手部：左肩、左肘、左腕，其中左肩为左手部的根关节，后面每个关节的父关节为前一个关节。
右手部：右肩、右肘、右腕，其中右肩为右手部的根关节，后面每个关 节的父关节为前一个关节。
左腿部：左胯、左膝、左踝、左脚，其中左胯为左腿部的根关节，后面每个关节的父关节为前一个关节。
右腿部：右胯、右膝、右踝、右脚，其中右胯为右腿部的根关节，后面每个关节的父关节为前一个关节。
人体运动本质上是各个关节的运动，即各个关节相对于父关节的旋转，以及胯中的位移。
父关节的旋转带着它的所有子孙关节(相对于父关节旋转的关节)一起旋转，而不会影响其它父关节的旋转；关节的子孙关节作为此关节的一部分，若只是此关节旋转，其子孙关节本身并不旋转，其关系说明如下：
请参照图4a，以人体左手部动作为例，此时左手部处于水平状态，A点代表左肩，B点代表左肘，C点代表左腕。建立三维直角坐标系，X轴沿左手部水平方向，Y轴为垂直左手部方向，Z轴为垂直XY平面方向。关节的旋转角度用(x，y，z)表示，其中x，y，z分别表示该关节相对于其父关节绕x、y、z轴的旋转角度。在图4a所示左手部处于水平状态时，A、B、C三点的旋转角度信息分别为A(0，0，0)、B(0，0，0)、C(0，0，0)。当左手部由处于水平状态进行左肘旋转，具体为左肘绕Y轴旋转45度，绕Z轴旋转45度后，如图4b所示，则此时A、B、C三点的旋转角度信息分别变化为A(0，0，0)、B(0，45，45)、C(0，0，0)，即作为左肘子孙关节的左腕并不发生旋转。
因此，动作判定单元4将当前动作姿态与标准动作姿态进行比较，判定动作是否符合要求，其基本原理是根据当前动作，定出人体上的关键关节点，然后检测这些关键关节点的旋转与标准动作姿态对应关节的旋转的误差是否在指定的容错范围内。因此，具体的，动作判定单元4具体包括一关节点确定模块(图中未示出)，用于根据当前动作定出人体上的关键关节点；一旋转角度计算模块(图中未示出)，用于计算出当前动作姿态中各关键关节点的旋 转角度；以及一比较判定模块(图中未示出)，用于将当前动作姿态中各关键关节点的旋转角度与预存的标准动作姿态中相对应的关节点的旋转角度进行比较，判断差异值是否在预设的容错范围内。
其中，检测用户动作是否标准的算法公式如下：
Figure PCTCN2016112255-appb-000001
公式中，Pi.x、Pi.y、Pi.z:分别表示用户动作关键关节中第i个关节相对于其父关节绕x轴、y轴、z轴的旋转角度；Ti.x、Ti.y、Ti.z:分别表示标准动作关键关节中第i个关节相对于其父关节绕x轴、y轴、z轴的旋转角度；error表示用户此动作与标准动作的误差；预先设定一表示允许的最在容错阈值，用Tolerance表示，则，如果error<Tolerance，则表明用户动作与标准动作动作间的误差在预设范围内，符合动作要求；如果error>＝Tolerance，则表明用户动作与标准动作动作间的误差不在预设范围内，不符合动作要求。
下面继续以太极的起势动作为例，对判断当前动作是否符合标准进行说明。由于对如何从提取的人体姿态中，计算出各个关节点的旋转角度信息，已有大量技术文献公开，如微软的Kinect技术，因此，该部分在这里不再赘述。
以检测太极的起势动作中，左手部是否标准为例。关节点确定模块定出左手部关键节点为左肩、左肘和左腕，检测时，只对该三个关节点进行旋转角度检测。标准动作姿态下，如图5a所示，左肩即图中a点的旋转角度为(47.6，40.2，11.1)，左肘即图中b点的旋转角度为(-53.2，-20.7，-3.9)，左腕即图中c点的旋转角度为(4.3，-9.8，23.2)；当用户做出的动作如图5b所示，旋转角度计算模块分别检测出左肩即图中a'点的旋转角度为(28.8，22.1，2.5)，左肘即图中b'点的旋转角度为(-54.5，-20.2，-8.1)，左腕即图中c'点的旋转角度为(21.9，8.5，30.8)；在经过比较判定模块的计算后，
左肩的误差为：
sqrt((47.6-28.8)*(47.6-28.8)+(40.2-22.1)*(40.2-22.1)+(11.1-2.5)*(11.1-2.5))＝27.4774
左肘的误差为：
sqrt((-53.2+54.5)*(-53.2+54.5)+(-20.7+20.2)*(-20.7+20.2)+(-3.9+8.1)*(-3.9+8.1))＝4.4249
左腕的误差为：
sqrt((4.3-21.9)*(4.3-21.9)+(-9.8-8.5)*(-9.8-8.5)+(23.2-30.8)*(23.2-30.8))＝26.5030
根据前文所述的检测用户动作是否标准的算法公式，则用户在做太极的起势动作时，就左手部动作姿态，经比较判定模块计算出的用户当前动作与标准动作姿态的误差为：27.4774。同理，相继求出用户右手部，左腿部，右腿部以及身体部(根据动作类型及严格程度的需要，选择需要检测的人体关节部分)的动作误差，然后与预设的容错值进行比较，判断用户当前动作与标准动作姿态间的误差是否在预设范围内。
进一步的，动作判定单元4还包括一错误提示模块(图中未示出)，该错误提示模块可以是一显示器，用于播放视频纠正出用户当前动作错误之处，除此，还可以显示登录用户的运动方案，以及播放声音和/或画面的方式向用户反馈动作错误提醒信息。
根据用户当前动作与标准动作姿态相比后的判断结果，系统分别作出不同处理：
(1)当检测出用户当前动作姿态与标准动作姿态间的差异值不在预设的容错范围内，即判断出用户当前动作不标准时，在反馈给用户动作错误提醒信息后，执行步骤S5，即
S5、将当前动作的标准姿态引导信息反馈给用户，引导用户重做当前动作，然后返回执行步骤S2。
根据动作判定单元4给过来的启动信号，姿势引导单元5将当前动作的 标准姿态引导信息反馈给用户，引导用户重做当前动作。
进一步的，当检测出用户当前动作姿态与标准动作姿态间的差异值不在预设的容错范围内，系统在执行步骤S5之前还包括执行步骤S41，即
S41、检查用户重复做当前动作的次数是否已达到预设的允许重复次数。
相对应的，系统中的计数单元7，用于检查用户重复做当前动作的次数是否已达到预设的次数。当检测出用户当前动作姿态与标准动作姿态间的差异值不在预设的容错范围内时，动作判定单元4在启动姿势引导单元5之前先启动计数单元7，对用户重复做当前动作的次数是否已达到预设的允许重复次数进行检查。每条运动指导信息内中可预先设置好当前动作的允许重复次数，当用户在重复做该动作时，通过计数单元7的检查，可以避免出现当用户重复当前动作始终无法达到预设标准动作姿态时，一直不停地重复当前动作而无法进入下一条动作的情况。这样，即使用户执行某条运动指导无法达到预设标准动作姿态时，仍然可以继续进行下一条运动指导，做下一个运动动作，从而带来更好的用户体验。
(2)当检测出用户当前动作姿态与标准动作姿态间的差异值在预设的容错范围内，即判断出用户当前动作标准时，系统执行步骤S6，即
S6、判定当前执行的是否是最后一条运动指导信息。若是，则执行步骤S7；若否，则返回执行步骤S2。
S7、判定本次运动结束。
根据动作判定单元4给过来的启动信号，进度单元6判定当前执行的是否是运动指导单元4所发出的最后一条运动指导信息，若是，则判定本次运动结束。如当用户以符合要求的动作姿态完成了四十二式太极拳运动方案中的最后一条运动指导信息“收势(收脚并立)”动作，则判定用户已完成运动方案“完成一套四十二式太极拳运动”。相反的，若判定当前执行的不是最后一条指导信息，则启动运动指导单元4以引导用户继续运动，即从尚未执行的运动指导信息序列中，执行首条运动指导信息。如当前用户以符合要求的 动作姿态完成了四十二式太极拳运动方案中的运动指导信息“收势(分手下按)”动作，此时，判定出当前执行并不是最后一条运动指导信息，尚未执行的运动指导信息序列中，还有“收势(收脚并立)”这一条运动指导信息，因此，继续引导用户做“收势(收脚并立)”这一动作。
进一步的，系统在执行步骤S6中判定若本次运动结束之后还包括，对用户该次运动信息进行记录以据此进行运动方案更新。
相对应的，系统还包括一记录模块(图中未示出)，用于在在判定用户本次运动结束后，对用户该次运动信息进行记录，上传到本地或云端服务器，对其运动方案进行更新。
本发明提供的电子教练实现方法及系统，在获取用户身份信息后，调取出用户匹配的运动方案，让用户根据运动方案做运动。运动过程中，采用图像拍摄的方式，对用户运动动作与预存的标准动作进行比较，实时纠正不标准动作。采用本发明方案，通过视觉识别的方式，为健身者提供了一名电子教练，可以对健身者运动中的动作进行自动比对和纠正，从而实时指导其按照运动处方开展健身活动。
以上所述仅为本发明的示例性实施例，并非因此限制本发明的专利保护范围，凡是利用本发明说明书及附图内容所作的等效结构或等效流程变换，或直接或间接运用在其他相关的技术领域，均同理包括在本发明的专利保护范围内。

Claims (10)

  1. 一种电子教练实现方法，其特征在于，包括：
    S1、获取登录用户的身份信息，调取与该用户身份相匹配的运动方案；所述运动方案包括依序排列的各运动指导信息；
    S2、从尚未执行的运动指导信息序列中，自首条运动指导信息开始引导用户运动；
    S3、拍摄用户运动图像；
    S4、提取出拍摄到的用户运动图像中的当前动作姿态，将当前动作姿态与预存的标准动作姿态进行比较，判断两者之间的差异是否在预设范围内；若否，反馈给用户动作错误提醒信息并执行步骤S5；若是，则执行步骤S6；
    S5、将当前动作的标准姿态引导信息反馈给用户，引导用户重做当前动作，然后执行步骤S3；
    S6、判定当前执行的是否是最后一条运动指导信息，若是，则判定本次运动结束；若否，则返回执行步骤S2。
  2. 根据权利要求1所述的电子教练实现方法，其特征在于，每条运动指导信息内包含了当前动作的允许重复次数，所述步骤S4中执行步骤S5之前还包括：检查用户重复做当前动作的次数是否已达到预设的允许重复次数，若是，则判定用户已执行完毕本条运动指导信息，接着执行步骤S6；若否，则执行步骤S5。
  3. 根据权利要求1或2所述的电子教练实现方法，其特征在于，所述步骤S4中将当前动作姿态与预存的标准动作姿态进行比较包括：
    根据当前动作定出人体上的关键关节点，计算出当前动作姿态中各关键关节点的旋转角度；
    将当前动作姿态中各关键关节点的旋转角度与预存的标准动作姿态中相对应的关节点的旋转角度进行比较。
  4. 根据权利要求1或2所述的电子教练实现方法，其特征在于，所述步骤S5中将当前动作的标准姿态引导信息反馈给用户包括：播放视频纠正出用 户当前动作错误之处，并演示标准动作姿态。
  5. 根据权利要求1或2所述的电子教练实现方法，其特征在于，步骤S6中判定若本次运动结束之后还包括，对用户该次运动信息进行记录以据此进行运动方案更新。
  6. 一种电子教练实现系统，其特征在于，包括：
    登录单元，用于获取登陆用户的身份信息，并调取与该用户身份相匹配的运动方案；所述运动方案包括依序排列的各运动指导信息；
    运动指导单元，用于从尚未执行的运动指导信息序列中，自首条运动信息开始引导用户运动；
    拍摄单元，用于拍摄用户运动图像；
    动作判定单元，用于提取出拍摄单元拍摄到的用户运动图像中的当前动作姿态，将当前动作姿态与预存的标准动作姿态进行比较，判断两者之间的差异是否在预设范围内；若否则反馈给用户动作错误提醒信息并启动姿势引导单元，若是则启动进度单元；
    姿势引导单元，将当前动作的标准姿态引导信息反馈给用户，引导用户重做当前动作；
    进度单元，判定当前执行的是否是运动指导单元所发出的最后一条运动指导信息，若是则判定本次运动结束，若否则启动运动指导单元以引导用户继续运动。
  7. 根据权利要求6所述的电子教练实现系统，其特性在于，每条运动指导信息内包含了当前动作的允许重复次数，所述电子教练实现系统还包括一计数单元，所述动作判定单元在启动姿势引导单元之前先启动所述计数单元，所述计数单元用于检查用户重复做当前动作的次数是否已达到预设的允许重复次数，若是，则判定用户已执行完毕本条运动指导信息，并启动进度单元；若否，则启动姿势引导单元。
  8. 根据权利要求6或7所述的电子教练实现系统，其特征在于，所述动 作判定单元包括：
    关节点确定模块，用于根据当前动作定出人体上的关键关节点；
    旋转角度计算模块，用于计算出当前动作姿态中各关键关节点的旋转角度；以及
    比较判定模块，用于将当前动作姿态中各关键关节点的旋转角度与预存的标准动作姿态中相对应的关节点的旋转角度进行比较，判断差异值是否在预设的容错范围内。
  9. 根据权利要求6或7所述的电子教练系统，其特征在于，所述动作判定单元包括错误提示模块，用于播放视频纠正出用户当前动作错误之处。
  10. 根据权利要求6或7所述的电子教练实现系统，其特征在于，所述电子教练实现系统还包括一记录模块，用于在进度单元判定本次运动结束后，对用户该次运动信息进行记录以据此进行运动方案更新。
PCT/CN2016/112255 2016-04-19 2016-12-27 电子教练实现方法及系统 WO2017181717A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201610243003.7A CN105903157B (zh) 2016-04-19 2016-04-19 电子教练实现方法及系统
CN201610243003.7 2016-04-19

Publications (1)

Publication Number Publication Date
WO2017181717A1 true WO2017181717A1 (zh) 2017-10-26

Family

ID=56746594

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/112255 WO2017181717A1 (zh) 2016-04-19 2016-12-27 电子教练实现方法及系统

Country Status (2)

Country Link
CN (1) CN105903157B (zh)
WO (1) WO2017181717A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114220300A (zh) * 2021-02-01 2022-03-22 黄华 利用增强现实佩戴设备可视化智能互动教考系统和方法
CN114403858A (zh) * 2021-12-29 2022-04-29 杭州程天科技发展有限公司 一种人体运动功能评估方法、设备及系统
WO2022171971A1 (fr) * 2021-02-15 2022-08-18 Pleyo Système d'entraînement interactif

Families Citing this family (33)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105903157B (zh) * 2016-04-19 2018-08-10 深圳泰山体育科技股份有限公司 电子教练实现方法及系统
JP6871708B2 (ja) * 2016-10-06 2021-05-12 りか 高木 代償運動の原因部位を特定するための方法、システム、プログラム、及びコンピュータ装置、並びに、代償運動を解消するための方法、及びシステム
CN106448295A (zh) * 2016-10-20 2017-02-22 泉州市开拓者智能科技有限公司 基于捕捉的远程教学系统及其方法
CN106984027B (zh) * 2017-03-23 2019-07-26 华映科技（集团）股份有限公司 一种动作对比分析方法、装置及一种显示器
CN107153812A (zh) * 2017-03-31 2017-09-12 深圳先进技术研究院 一种基于机器视觉的运动辅助方法及系统
CN107463785A (zh) * 2017-08-17 2017-12-12 广州视源电子科技股份有限公司 健身管理方案的确定方法、装置、设备及存储介质
CN107551521B (zh) * 2017-08-17 2020-05-08 广州视源电子科技股份有限公司 健身指导方法及装置、智能设备及存储介质
CN107485828A (zh) * 2017-09-22 2017-12-19 四川长虹电器股份有限公司 家庭健身系统
CN107909060A (zh) * 2017-12-05 2018-04-13 前海健匠智能科技（深圳）有限公司 基于深度学习的健身房健身动作识别方法及装置
CN110064176A (zh) * 2018-01-23 2019-07-30 鎏玥(上海)健康咨询有限公司 一种智能健身系统
CN108319922A (zh) * 2018-02-06 2018-07-24 锐马（福建）电气制造有限公司 一种扎马步指导方法
CN108428472A (zh) * 2018-02-27 2018-08-21 深圳市知本石墨烯医疗科技有限公司 运动锻炼的数据处理方法、系统及智能瑜伽垫
CN108492311A (zh) * 2018-03-15 2018-09-04 北京小米移动软件有限公司 通过电子设备进行动作校正的方法及装置
CN110298221B (zh) * 2018-03-23 2022-02-11 上海史贝斯健身管理有限公司 自助健身方法、系统、电子设备、存储介质
CN110298220B (zh) * 2018-03-23 2021-11-09 上海史贝斯健身管理有限公司 动作视频直播方法、系统、电子设备、存储介质
CN110298218B (zh) * 2018-03-23 2022-03-04 上海史贝斯健身管理有限公司 交互式健身装置和交互式健身系统
CN110428486B (zh) * 2018-04-28 2022-09-27 上海史贝斯健身管理有限公司 虚拟互动的健身方法、电子设备及存储介质
CN109407829A (zh) * 2018-09-18 2019-03-01 孔军民 一种应用于健身器材的人机交互系统及交互方法
CN109248414B (zh) * 2018-09-30 2021-08-10 深圳市科迈爱康科技有限公司 运动训练提醒方法、装置、设备以及可读存储介质
CN109461488A (zh) * 2018-10-31 2019-03-12 上海掌门科技有限公司 便携式激励运动设备控制方法及其设备、终端和服务器
CN111353345B (zh) * 2018-12-21 2024-04-16 上海史贝斯健身管理有限公司 提供训练反馈的方法、装置、系统、电子设备、存储介质
CN109820690B (zh) * 2019-03-11 2021-06-25 贵阳市第四人民医院 一种穿戴式肘关节康复训练系统
CN110147468A (zh) * 2019-04-15 2019-08-20 深圳壹账通智能科技有限公司 基于人工智能的视频教学方法及相关设备
CN111652078A (zh) * 2020-05-11 2020-09-11 浙江大学 一种基于计算机视觉的瑜伽动作指导系统和方法
CN113721758A (zh) * 2020-05-26 2021-11-30 华为技术有限公司 一种健身指导方法与电子设备
CN111985393A (zh) * 2020-08-18 2020-11-24 深圳市瓴鹰智能科技有限公司 一种纠正运动姿态的智能镜子及其运动姿态纠正方法
CN112386882A (zh) * 2020-10-14 2021-02-23 珠海格力电器股份有限公司 虚拟运动系统、方法、计算机设备及计算机可读存储介质
CN112201323A (zh) * 2020-10-16 2021-01-08 童心堂健康科技(北京)有限公司 基于人工智能的运动监测方法、装置和计算机设备
CN113096152B (zh) * 2021-04-29 2022-04-01 北京百度网讯科技有限公司 多对象运动分析方法、装置、设备和介质
CN113641856A (zh) * 2021-08-12 2021-11-12 三星电子（中国）研发中心 用于输出信息的方法和装置
CN113694501A (zh) * 2021-08-31 2021-11-26 重庆科华安全设备有限责任公司 一种应用于引体向上的智能指导系统
CN114642867B (zh) * 2022-04-15 2023-07-18 胜兰科技(杭州)有限公司 一种具有划船运动姿态自动纠错功能的ai教练机系统
CN116611970B (zh) * 2023-07-20 2023-11-07 中国人民解放军空军特色医学中心 结合人脸和姿态识别的群体训练动作纠正系统及方法

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101174346A (zh) * 2006-11-01 2008-05-07 姚旺 健身房的电子教练方法及装置
US20080170123A1 (en) * 2007-01-12 2008-07-17 Jacob C Albertson Tracking a range of body movement based on 3d captured image streams of a user
CN101564594A (zh) * 2008-04-25 2009-10-28 财团法人工业技术研究院 交互式肢体动作复健方法与系统
CN103127665A (zh) * 2011-11-25 2013-06-05 索尼公司 电子健身训练器和操作电子健身训练器的方法
CN104581079A (zh) * 2015-01-21 2015-04-29 山东大学 一种辅助用户使用健身器材的装置和方法
CN105903157A (zh) * 2016-04-19 2016-08-31 深圳泰山体育科技股份有限公司 电子教练实现方法及系统

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101202994B (zh) * 2006-12-14 2012-10-24 北京三星通信技术研究有限公司 辅助用户健身的方法和装置
EP2347363A1 (en) * 2008-10-27 2011-07-27 Koninklijke Philips Electronics N.V. User feedback engine
US8897491B2 (en) * 2011-06-06 2014-11-25 Microsoft Corporation System for finger recognition and tracking
TWI464696B (zh) * 2012-09-12 2014-12-11 Ind Tech Res Inst 動作比對方法與動作比對系統

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101174346A (zh) * 2006-11-01 2008-05-07 姚旺 健身房的电子教练方法及装置
US20080170123A1 (en) * 2007-01-12 2008-07-17 Jacob C Albertson Tracking a range of body movement based on 3d captured image streams of a user
CN101564594A (zh) * 2008-04-25 2009-10-28 财团法人工业技术研究院 交互式肢体动作复健方法与系统
CN103127665A (zh) * 2011-11-25 2013-06-05 索尼公司 电子健身训练器和操作电子健身训练器的方法
CN104581079A (zh) * 2015-01-21 2015-04-29 山东大学 一种辅助用户使用健身器材的装置和方法
CN105903157A (zh) * 2016-04-19 2016-08-31 深圳泰山体育科技股份有限公司 电子教练实现方法及系统

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114220300A (zh) * 2021-02-01 2022-03-22 黄华 利用增强现实佩戴设备可视化智能互动教考系统和方法
WO2022171971A1 (fr) * 2021-02-15 2022-08-18 Pleyo Système d'entraînement interactif
FR3119777A1 (fr) * 2021-02-15 2022-08-19 Pleyo Système d’entraînement interactif
CN114403858A (zh) * 2021-12-29 2022-04-29 杭州程天科技发展有限公司 一种人体运动功能评估方法、设备及系统
CN114403858B (zh) * 2021-12-29 2023-10-03 杭州程天科技发展有限公司 一种人体运动功能评估方法、设备及系统

Also Published As

Publication number Publication date
CN105903157A (zh) 2016-08-31
CN105903157B (zh) 2018-08-10

Similar Documents

Publication Publication Date Title
WO2017181717A1 (zh) 电子教练实现方法及系统
US11776421B2 (en) Systems and methods for monitoring and evaluating body movement
Zhao et al. Rule-based human motion tracking for rehabilitation exercises: realtime assessment, feedback, and guidance
KR100772497B1 (ko) 골프 클리닉 시스템 및 그것의 운용방법
Anilkumar et al. Pose estimated yoga monitoring system
JP2020174910A (ja) 運動支援システム
Hua et al. Evaluation of machine learning models for classifying upper extremity exercises using inertial measurement unit-based kinematic data
JP2022043264A (ja) 運動評価システム
KR102320960B1 (ko) 사용자 신체 맞춤형 홈 트레이닝 동작 안내 및 교정 시스템
Brooks et al. Quantifying upper-arm rehabilitation metrics for children through interaction with a humanoid robot
CN114022512A (zh) 运动辅助方法、装置及介质
CN113409651B (zh) 直播健身方法、系统、电子设备、存储介质
Lin et al. The effect of real-time pose recognition on badminton learning performance
Ma et al. A wushu posture recognition system based on mediapipe
WO2019183733A1 (en) Method and system for motion capture to enhance performance in an activity
US20230285806A1 (en) Systems and methods for intelligent fitness solutions
CN112541382A (zh) 一种辅助运动的方法、系统以及识别终端设备
WO2021261529A1 (ja) 身体運動支援システム
WO2022030619A1 (ja) 指導支援システム
Cai et al. PoseBuddy: Pose estimation workout mobile application
Gharasuie et al. Performance monitoring for exercise movements using mobile cameras
Sharma et al. Digital Yoga Game with Enhanced Pose Grading Model
Zhang et al. The Application of Computer-Assisted Teaching in the Scientific Training of Sports Activities
CN113705536B (zh) 连续动作打分方法、装置及存储介质
Hande et al. Yoga Postures Correction and Estimation using Open CV and VGG 19 Architecture

Legal Events

Date Code Title Description
NENP Non-entry into the national phase

Ref country code: DE

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 16899296

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 16899296

Country of ref document: EP

Kind code of ref document: A1