CN107491648A - Hand recovery training method based on Leap Motion motion sensing control devices - Google Patents

Hand recovery training method based on Leap Motion motion sensing control devices Download PDF

Info

Publication number
CN107491648A
CN107491648A CN201710736343.8A CN201710736343A CN107491648A CN 107491648 A CN107491648 A CN 107491648A CN 201710736343 A CN201710736343 A CN 201710736343A CN 107491648 A CN107491648 A CN 107491648A
Authority
CN
China
Prior art keywords
hand
data
motion
patient
training
Prior art date
Legal status (The legal status 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 status listed.)
Pending
Application number
CN201710736343.8A
Other languages
Chinese (zh)
Inventor
张林宣
姚荣
龙腾
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Tsinghua University
Original Assignee
Tsinghua University
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 Tsinghua University filed Critical Tsinghua University
Priority to CN201710736343.8A priority Critical patent/CN107491648A/en
Publication of CN107491648A publication Critical patent/CN107491648A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
    • G06F3/011—Arrangements for interaction with the human body, e.g. for user immersion in virtual reality
    • G06F3/014—Hand-worn input/output arrangements, e.g. data gloves
    • A—HUMAN NECESSITIES
    • A63—SPORTS; GAMES; AMUSEMENTS
    • A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
    • A63B23/00—Exercising apparatus specially adapted for particular parts of the body
    • A63B23/035—Exercising apparatus specially adapted for particular parts of the body for limbs, i.e. upper or lower limbs, e.g. simultaneously
    • A63B23/12—Exercising apparatus specially adapted for particular parts of the body for limbs, i.e. upper or lower limbs, e.g. simultaneously for upper limbs or related muscles, e.g. chest, upper back or shoulder muscles
    • A63B23/16—Exercising apparatus specially adapted for particular parts of the body for limbs, i.e. upper or lower limbs, e.g. simultaneously for upper limbs or related muscles, e.g. chest, upper back or shoulder muscles for hands or fingers
    • 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
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent
    • 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
    • A63B2024/0012—Comparing movements or motion sequences with a registered reference
    • A—HUMAN NECESSITIES
    • A63—SPORTS; GAMES; AMUSEMENTS
    • A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
    • A63B2220/00—Measuring of physical parameters relating to sporting activity
    • A63B2220/80—Special sensors, transducers or devices therefor

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Theoretical Computer Science (AREA)
  • General Engineering & Computer Science (AREA)
  • General Health & Medical Sciences (AREA)
  • Physics & Mathematics (AREA)
  • Orthopedic Medicine & Surgery (AREA)
  • Physical Education & Sports Medicine (AREA)
  • General Physics & Mathematics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • Biophysics (AREA)
  • Biomedical Technology (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Artificial Intelligence (AREA)
  • Human Computer Interaction (AREA)
  • Rehabilitation Tools (AREA)

Abstract

The invention discloses a kind of hand recovery training method based on Leap Motion motion sensing control devices, comprise the following steps：Step A, the typing standard hand motion data into main frame；Step B, the standard hand motion that patient plays according to display completes rehabilitation exercise motion, obtains the data of patient hand's action in real time during patient completes rehabilitation exercise motion and is transferred to main frame；Step C, the data acted to the patient hand collected are handled, and extract valid data therein, and then extraction obtains training characteristics data；Step D, the action completed to patient is evaluated.The present invention carries out real-time display using the combination of virtual reality technology and body feeling interaction technology to the hand motion of patient, patient can see the training process of oneself in real time in the environment of virtual reality, improve the training enthusiasm of patient, the passive type for becoming traditional is trained for active training, while raising resumes training effect, the rehabilitation cost of patient is reduced.

Description

Hand recovery training method based on Leap Motion motion sensing control devices
Technical field
The present invention relates to Rehabilitation equipment and information intelligent process field, relates in particular to one kind and is applied to cerebral apoplexy The hand training rehabilitation training method based on Leap Motion motion sensing control devices of Rehabilitation.
Background technology
Cerebral apoplexy, it is a kind of acute cerebrovascular diseases also known as " apoplexy " or " cerebrovas-cularaccident ".Its pathogenic factor one As be blood vessel of the part to blood supply in brain（Mainly artery）It is oppressed or bursts, within a few minutes, the nerve in the region is thin Born of the same parents will be affected it is even dead within a few houres, and this will result directly in can not by the body part of brain region control Normal work.Paralytic's democratic meeting after illness is greatly affected, and is mainly reflected in as hand is powerless, hemiplegia etc. The appearance of symptom, patient can not complete the necessary behavior of life of a part, and quality of life is greatly diminished, and is aggravated significantly Its family and personal burden.Existing research shows：In the existing paralytic in China, along with dyskinesia Patient be up to 75%, and there is hand exercise problem on obstacle in most patient wherein, hand be in normal life most One of conventional body part, the quality of life of patient obviously can be largely affected by the problem of hand.
At present in the hand rehabilitation project field in China, for the patient of the symptom such as hand hemiplegia, powerless, swelling, mainly Acted using doctor's guidance, the mode of patient's active training carries out hand rehabilitation, is also proposed the instrument of many hand rehabilitations Aid in Rehabilitation.But these means occupy more medical resource, in the epoch that nowadays this medical resource is in short supply, this A little rehabilitation maneuvers waste medical resource, have also aggravated the cost of patient, have brought many inconvenience.But if patient lacks doctor The guidance of teacher or good aid help hand to temper recovery, can not be accurately performed hand rehabilitation action again, and right Also lack feedback in the execution of patient, easily cause training effect difference or even aggravate the consequence of the state of an illness.
The Chinese patent of Application No. 201310133113.4 discloses a kind of upper extremity exercise based on Kinect sensor Rehabilitation training system and its training method, using virtual reality technology, patient is set to have broken away from biography using body-sensing man-machine interaction mode Limitation of the interactive devices such as the mouse and keyboard of system to body position, without wearing complicated motion capture equipment；Patient passes through Word and the method for voice message are trained the real time correction of action, training effect Real-time Feedback, and patient can not control Autonomous completion is trained with the help for the treatment of teacher.But it is only applicable to upper limb shoulder, elbow, wrist simple joint mobility and resumed training and comprehensive The training of the control coordination ability is closed, the rehabilitation of patients with cerebral apoplexy hand is not helped then.
The content of the invention
The technical problem to be solved in the present invention is to provide a kind of hand rehabilitation instruction based on Leap Motion motion sensing control devices Practice method, while guiding patient completes rehabilitation action, the hand motion data of patient are obtained by body feeling interaction equipment, in void Intend real-time display hand realistic operation and standard operation under the environment of reality, judge patient's by being compared to two actions Rehabilitation situation simultaneously provides assessment.
In order to solve the above technical problems, the technical solution adopted by the present invention is：One kind is based on Leap Motion motion sensing controls The hand recovery training method of device, based on hand rehabilitation training system, the hand rehabilitation training system includes being provided with Unity The main frames of 3D softwares, Leap Motion motion sensing controls devices and the display being connected with main frame, its feature exist In comprising the following steps：
Step A, the typing standard hand motion data into main frame；
Step B, the standard hand motion that patient plays according to display completes rehabilitation exercise motion, and rehabilitation training is completed in patient Leap Motion motion sensing controls device obtains the data of patient hand's action and is transferred to main frame in real time in action process；
Step C, the data that main frame acts to the patient hand collected are handled, and extract valid data therein, Then the hand-characteristic data in valid data are extracted to obtain training characteristics data；
Step D, the standard feature data in training characteristics data and standard hand motion data are compared, the action completed to patient Evaluated.
The method have the benefit that：1st, using the combination of virtual reality technology and body feeling interaction technology to patient's Hand motion carries out real-time display, and patient can see the training process of oneself in real time in the environment of virtual reality, can be with The training enthusiasm of patient is largely improved, the passive type for becoming traditional is trained for active training, and raising resumes training effect While fruit, the rehabilitation cost of patient is reduced；2nd, for traditional hand rehabilitation aid it is intelligent insufficient the problem of, this is System employs rehabilitation training situation of the nerual network technique to patient and assessed, and system is ensure that using the characteristic of neutral net Can be perfect in continuous training, " intelligence " provide feedback for the rehabilitation training of patient, ensure that the rehabilitation training of patient Quality.
The present invention is described in detail below in conjunction with the accompanying drawings.
Brief description of the drawings
Fig. 1 is the structural representation of the hand rehabilitation training system of the invention based on Leap Motion motion sensing control devices；
Fig. 2 is actual hand skeletal graph；
Fig. 3 is the hand skeletal graph after system processing；
Fig. 4 is the hand bone vectogram after system processing；
Fig. 5 is the angle between bone direction vector；
Fig. 6 is the flow chart of the hand recovery training method embodiment two of the invention based on Leap Motion motion sensing control devices.
In the accompanying drawings：1 is main frame, and 2 be Leap Motion motion sensing control devices, and 3 be display.
Embodiment
Referring to accompanying drawing 1, the invention provides a kind of hand rehabilitation training system based on Leap Motion motion sensing control devices System, key are：Including being provided with the main frame 1 of Unity 3D softwares, realizing the Leap of body feeling interaction and data acquisition Motion motion sensing controls device 2 and the display 3 being connected with main frame 1.
Based on above-mentioned hand rehabilitation training system, the invention provides one kind to be based on Leap Motion motion sensing control devices Hand recovery training method.
Embodiment one：The assessment to patient motion is completed using static action and dynamic action in the present embodiment.
The hand recovery training method of the present embodiment comprises the following steps：
Step A, Leap Motion motion sensing controls device 2 and supporting Unity 3D softwares typing into main frame 1 are passed through Standard hand motion data.In this step, Leap Motion motion sensing controls device 2 with the speed of 60 frame per second to calculate owner Machine 1 sends frame data, includes hand solid data and finger solid data in the frame data of each frame.Moved by standard hand Making data can be with extraction standard characteristic.
Step B, the standard hand motion that patient plays according to display 3 completes rehabilitation exercise motion, and health is completed in patient Refreshment practices Leap Motion motion sensing controls device 2 in action process and obtains the data of patient hand's action in real time and be transferred to calculating Machine host 1.In this step, Leap Motion motion sensing controls device 2 is also to be sent out with the speed of 60 frame per second to main frame 1 Frame data are sent, include hand solid data and finger solid data in the frame data of each frame.
Step C, the data that main frame 1 acts to the patient hand collected are handled, and are extracted therein effective Data, then the hand-characteristic data in valid data are extracted to obtain training characteristics data.In this step, referring to attached Fig. 2-4, actual hand bone are distributed as shown in Figure 2, and hand has 19 bones, also have two bones in wrist, but this For two bones when patient carries out actual rehabilitation training, bone direction is very close, only need to can be regarded as a direction, institute In the method, the two bones have only been considered into a direction, the hand skeletal graph after system processing is as shown in Figure 3. Specific way is that the coordinate of elbow is subtracted with the coordinate of wrist, then is carried out unitization.Specific formula for calculation is as follows：
In formulaWrist vector after expression processing,The coordinate of patient's wrist is represented,Represent patient's ancon Coordinate.Hand bone vectogram after system processing is as shown in Figure 4.
The various entities for each frame data bag hand that Leap Motion motion sensing controls device 2 is sent to main frame 1 Data, include the direction vector of bone, and system is used for subsequent treatment by obtaining 20 unit bone vectors after processing.
Similarly, same operation extraction standard characteristic is carried out to standard hand motion data.
Step D, training characteristics data and standard feature data are compared, the action completed to patient is evaluated.In this step The frame data of same time hand standard operation and patient motion are taken out in rapid, static action is carried out and assesses.
Static state action assessment comprises the following steps：
Step（1）, calculate the hand of the same bone of hand in the frame data of same time hand standard operation and patient motion The cosine value of the angle theta of portion's bone vector, wherein coordinate value of the bone in hand standard operation frame data are（x0, y0, z0）, the coordinate value in patient motion frame data be（x1, y1, z1）.Referring to accompanying drawing 5,Hand standard in a certain frame is represented to move The a certain bone vector made,The same bone vector of patient motion in same frame is represent,ForWithAngle.
Wherein cos (θ)=x0*x1+y0*y1+z0*z1。
Step（2）, calculate the cosine summation of 20 hand bones vector in frame data,
Wherein；
Step（3）, calculate static evaluation result, in order to ensure that the full marks of end product are 100, in the result of this summation superior 5,
Final static evaluation end value is：
。
The frame data of every 20 frame form a set, ask for the static action assessed value of the frame data of each frame respectively, so This 20 values are averaged afterwards and complete assessed value as a dynamic action, then complete to comment by obtained all dynamic actions Valuation is averagely obtained final dynamic action and completes assessed value.Complete to assess by dynamic action, can judge to suffer from substantially Person acts the trend completed.
Embodiment two：, may when the training action somewhat hysterisis criterion of patient acts using the method for embodiment one Provide relatively low scoring.In the present embodiment, using the hand data processing based on machine learning, to reach reduction training action The influence of hysteresis, provide accurate assessment result.
Referring to accompanying drawing 6, the hand recovery training method of the present embodiment comprises the following steps：
Step A, Leap Motion motion sensing controls device 2 and supporting Unity 3D softwares typing into main frame 1 are passed through Standard hand motion data.In this step, Leap Motion motion sensing controls device 2 with the speed of 60 frame per second to calculate owner Machine 1 sends frame data, includes hand solid data and finger solid data in the frame data of each frame.Moved by standard hand Making data can be with extraction standard characteristic.
Step B, the standard hand motion that patient plays according to display 3 completes rehabilitation exercise motion, and health is completed in patient Refreshment, which is practiced, to be obtained the data of patient hand's action by Leap Motion motion sensing controls device 2 in action process and is transferred to calculating Machine host 1.In this step, Leap Motion motion sensing controls device 2 sends frame with the speed of 60 frame per second to main frame 1 Data, include hand solid data and finger solid data in the frame data of each frame.
Step C, the data that main frame 1 acts to the patient hand collected are handled, and are extracted therein effective Data, then the hand-characteristic data in valid data are extracted to obtain training characteristics data.
Training characteristics data are obtained to sdpecific dispersion algorithm CDK with the k steps in RBM training algorithms in this step and standard is special Levy data.The k steps used in the system are chosen for 0.005-0.02 to sdpecific dispersion algorithm parameter learning rate parameter η, initialize Visible layer nvWith hidden layer dimension nhRespectively 600 and 1-3, cycle of training, J was 20000-50000, passed through setting for these parameters It is fixed, proved by the comparison of many kinds of parameters, the parameter in the range of this, which is chosen, to be carried well to the characteristic realization of hand Take, applied well in neutral net afterwards.Wherein, k steps are to the preferred η of sdpecific dispersion algorithm learning rate parameter 0.01, initialization visible layer nvWith hidden layer dimension nhRespectively preferably 600 and 2, cycle of training J preferably 50000.
Step D, after the hand data after previous step is compressed, by two characteristics after compression and provide Scale trains BP neural network, after network training terminates, the validity of network is verified as training data, If effective to given test data, terminate training, if invalid to given test data, continue training or again Training.
In this step, assessed using BP neural network execution, action assessment comprises the following steps：
Step（1）, with the premnmx functions in matlab be done directly normalization require, even if ensure unit it is inconsistent Data can also become in a section, be easy to subsequent treatment, different size of input data is played equal effect（Due to instruction Practice the uncertainty of data, in fact it could happen that the proportion of input data differs, and normalization can avoid this problem）, it is easy to directly Activation primitive after use（In general network activation function all has certain requirement for domain, and can in input data There can be data can not be directly substituted into activation primitive, normalization can solve this problem）.
, the data before wherein x expression normalization, the data after y expression normalization, WithThe maximum and minimum value of x in this group of data is represented respectively.
Step（2）, due to the codomain of hyperbola S type activation primitives be（- 1,1）, select hyperbola S type activation primitive conducts The activation primitive of network.
Step（3）, utilize the newff in matlab, train functions complete BP neural network definition and training；Use Triple BP neural networks are trained, and in network parameter, input layer number is 1-3, node in hidden layer 20-100, is exported Node layer number is 1, and training precision 0.005-0.02, learning rate is arranged to 0.005-0.02, enters in the availability to network Row checking shows, by such parameter setting, is all controlled within 3 points for most error originated from input（100 points of systems）, Success rate is reaching more than 85%.More specifically, input layer number preferably 1, node in hidden layer preferably 40, exports node layer Number preferably 1, training precision preferably 0.01, learning rate set preferably 0.01.
Step（4）, after training terminates, carries out validation verification with sim functions, if estimation results and standard results error In allowed band, then terminate to train, system effectiveness is verified, and otherwise continues to be trained network.
Flow chart in the present embodiment as shown in Figure 6, BP neural network training terminate after, according to the training action of patient Scoring, as illustrated by dotted line in fig. 6, the training program of patient can be updated, patient completes required movement training.
Finally it should be noted that：The above embodiments are merely illustrative of the technical scheme of the present invention and are not intended to be limiting thereof；To the greatest extent The present invention is described in detail with reference to preferred embodiments for pipe, those of ordinary skills in the art should understand that：Still The embodiment of the present invention can be modified or equivalent substitution is carried out to some technical characteristics；Without departing from this hair The spirit of bright technical scheme, it all should cover among the claimed technical scheme scope of the present invention.

Claims (9)

  1. A kind of 1. hand recovery training method based on Leap Motion motion sensing control devices, based on hand rehabilitation training system, institute State the main frame that hand rehabilitation training system includes being provided with Unity 3D softwares（1）, Leap Motion motion sensing control devices （2）And and main frame（1）Connected display（3）, it is characterised in that comprise the following steps：
    Step A, to main frame（1）Interior typing standard hand motion data；
    Step B, patient is according to display（3）The standard hand motion of broadcasting completes rehabilitation exercise motion, and rehabilitation is completed in patient Leap Motion motion sensing control devices during training action（2）Patient hand's action data is obtained in real time and is transferred to computer Main frame（1）；
    Step C, main frame（1）The patient hand's action data collected is handled, extracts valid data therein, Then the hand-characteristic data in valid data are extracted to obtain training characteristics data；
    Step D, the standard feature data in training characteristics data and standard hand motion data are compared, the action completed to patient Evaluated.
  2. 2. the hand recovery training method according to claim 1 based on Leap Motion motion sensing control devices, its feature exist In in stepb, Leap Motion motion sensing control devices（2）With the speed of 60 frame per second to main frame（1）Send frame number According to the frame data of each frame include hand solid data and finger solid data.
  3. 3. the hand recovery training method according to claim 2 based on Leap Motion motion sensing control devices, its feature exist In the frame data for taking out same time hand standard operation and patient motion in step D, carry out static action and assess, it is described quiet State action assessment comprises the following steps：
    Step（1）, calculate the hand of the same bone of hand in the frame data of same time hand standard operation and patient motion The cosine value of the angle theta of portion's bone vector, wherein coordinate value of the bone in hand standard operation frame data are（x0, y0, z0）, the coordinate value in patient motion frame data be（x1, y1, z1）,
    Wherein cos (θ)=x0*x1+y0*y1+z0*z1；
    Step（2）, calculate the cosine summation of 20 hand bones vector in frame data,
    Wherein；
    Step（3）, calculate static evaluation result,
    It is worth and is, wherein ScorecosMost Big value is 100.
  4. 4. the hand recovery training method according to claim 3 based on Leap Motion motion sensing control devices, its feature exist A set is formed in the frame data of every 20 frame, the static action assessed value of the frame data of each frame is asked for respectively, then by this 20 values are averaged and complete assessed value as a dynamic action, then enter obtained all dynamic actions completion assessed value Row averagely obtains final dynamic action and completes assessed value.
  5. 5. the hand recovery training method according to claim 1 based on Leap Motion motion sensing control devices, its feature exist In obtaining training characteristics data and standard feature data with to sdpecific dispersion algorithm in step C, to sdpecific dispersion algorithm learning Rate parameter η is 0.005-0.02, initialization visible layer nvFor 600, hidden layer dimension nhFor 1-3, cycle of training, J was 20000- 50000。
  6. 6. the hand recovery training method according to claim 5 based on Leap Motion motion sensing control devices, its feature exist In being 0.01 to sdpecific dispersion algorithm learning rate parameter η, visible layer n is initializedvWith hidden layer dimension nhRespectively 600 and 2, instruction It is 50000 to practice cycle J.
  7. 7. the hand recovery training method according to claim 5 based on Leap Motion motion sensing control devices, its feature exist In utilizing BP neural network execution to assess in step D, action assessment comprises the following steps：
    Step（1）, with the premnmx functions in matlab be done directly normalization require,
    , the data before wherein x expression normalization, the data after y expression normalization,WithThe maximum and minimum value of x in this group of data is represented respectively；
    Step（2）, activation primitive of the selection hyperbola S types activation primitive as network；
    Step（3）, utilize the newff in matlab, train functions complete BP neural network definition and training；
    Step（4）, after training terminates, carries out validation verification with sim functions, if estimation results and standard results error are fair Perhaps in the range of, then terminate to train, system effectiveness is verified, and otherwise continues to be trained network.
  8. 8. the hand recovery training method according to claim 7 based on Leap Motion motion sensing control devices, its feature exist It is trained in using triple BP neural networks, network parameter sets as follows：Input layer number is 1-3, node in hidden layer For 20-100, output layer nodes are 1, and training precision 0.005-0.01, learning rate is arranged to 0.005-0.02.
  9. 9. the hand recovery training method according to claim 8 based on Leap Motion motion sensing control devices, its feature exist Input layer number is 1 in network parameter, node in hidden layer 40, and output layer nodes are 1, training precision 0.01, Learning rate is arranged to 0.01.
CN201710736343.8A 2017-08-24 2017-08-24 Hand recovery training method based on Leap Motion motion sensing control devices Pending CN107491648A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710736343.8A CN107491648A (en) 2017-08-24 2017-08-24 Hand recovery training method based on Leap Motion motion sensing control devices

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710736343.8A CN107491648A (en) 2017-08-24 2017-08-24 Hand recovery training method based on Leap Motion motion sensing control devices

Publications (1)

Publication Number Publication Date
CN107491648A true CN107491648A (en) 2017-12-19

Family

ID=60645829

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710736343.8A Pending CN107491648A (en) 2017-08-24 2017-08-24 Hand recovery training method based on Leap Motion motion sensing control devices

Country Status (1)

Country Link
CN (1) CN107491648A (en)

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108648795A (en) * 2018-03-30 2018-10-12 常州市第人民医院 A kind of ankle pump sport monitoring device and computer readable storage medium
CN108682450A (en) * 2018-06-11 2018-10-19 郑州大学 Online finger motion function evaluating system
CN108744414A (en) * 2018-04-09 2018-11-06 广东斐瑞智能技术有限公司 Mobilizing physiotherapy instrument and Motion evaluation method
CN109192272A (en) * 2018-11-26 2019-01-11 燕山大学 Based on the Leap Motion healing hand function training system combined with VR and its implementation
CN109199712A (en) * 2018-10-15 2019-01-15 郑州大学 A kind of evaluation and test of intelligent hand motor function and recovery training wheel chair
CN110070065A (en) * 2019-04-30 2019-07-30 李冠津 The sign language systems and the means of communication of view-based access control model and speech-sound intelligent
CN110075486A (en) * 2019-05-31 2019-08-02 东北大学 A kind of rehabilitation training of upper limbs system and method using virtual reality technology
CN110478883A (en) * 2019-08-21 2019-11-22 南京信息工程大学 A kind of body-building movement teaching and correction system and method
WO2020253645A1 (en) * 2019-06-17 2020-12-24 麦克赛尔株式会社 Finger motion exercise assistance apparatus and method
CN113274612A (en) * 2021-04-16 2021-08-20 顺德职业技术学院 VR interactive installation is felt to recovered body of hand
CN118506978A (en) * 2024-05-31 2024-08-16 广州美术学院 Hand function rehabilitation training system, method and equipment for cerebral apoplexy patient

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104007819A (en) * 2014-05-06 2014-08-27 清华大学 Gesture recognition method and device and Leap Motion system
CN105844100A (en) * 2016-03-24 2016-08-10 乐视控股（北京）有限公司 Method and system for carrying out rehabilitation training through television and somatosensory accessory

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104007819A (en) * 2014-05-06 2014-08-27 清华大学 Gesture recognition method and device and Leap Motion system
CN105844100A (en) * 2016-03-24 2016-08-10 乐视控股（北京）有限公司 Method and system for carrying out rehabilitation training through television and somatosensory accessory

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
刘志辉 等: "《基于Leap Motion的脑卒中上肢功能康复主动运动系统》", 《东华大学学报（自然科学版）》 *

Cited By (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108648795A (en) * 2018-03-30 2018-10-12 常州市第人民医院 A kind of ankle pump sport monitoring device and computer readable storage medium
CN108744414A (en) * 2018-04-09 2018-11-06 广东斐瑞智能技术有限公司 Mobilizing physiotherapy instrument and Motion evaluation method
CN108682450A (en) * 2018-06-11 2018-10-19 郑州大学 Online finger motion function evaluating system
CN109199712A (en) * 2018-10-15 2019-01-15 郑州大学 A kind of evaluation and test of intelligent hand motor function and recovery training wheel chair
CN109192272A (en) * 2018-11-26 2019-01-11 燕山大学 Based on the Leap Motion healing hand function training system combined with VR and its implementation
CN110070065A (en) * 2019-04-30 2019-07-30 李冠津 The sign language systems and the means of communication of view-based access control model and speech-sound intelligent
CN110075486A (en) * 2019-05-31 2019-08-02 东北大学 A kind of rehabilitation training of upper limbs system and method using virtual reality technology
WO2020253645A1 (en) * 2019-06-17 2020-12-24 麦克赛尔株式会社 Finger motion exercise assistance apparatus and method
CN110478883A (en) * 2019-08-21 2019-11-22 南京信息工程大学 A kind of body-building movement teaching and correction system and method
CN110478883B (en) * 2019-08-21 2021-04-13 南京信息工程大学 Body-building action teaching and correcting system and method
CN113274612A (en) * 2021-04-16 2021-08-20 顺德职业技术学院 VR interactive installation is felt to recovered body of hand
CN118506978A (en) * 2024-05-31 2024-08-16 广州美术学院 Hand function rehabilitation training system, method and equipment for cerebral apoplexy patient

Similar Documents

Publication Publication Date Title
CN107491648A (en) Hand recovery training method based on Leap Motion motion sensing control devices
US11944446B2 (en) Apparatus, method, and system for pre-action therapy
Avola et al. An interactive and low-cost full body rehabilitation framework based on 3D immersive serious games
Ding et al. Continuous estimation of human multi-joint angles from sEMG using a state-space model
Barzilay et al. Adaptive rehabilitation games
CN104524742A (en) Cerebral palsy child rehabilitation training method based on Kinect sensor
Garrido et al. Balance disorder rehabilitation through movement interaction
Dao et al. Interactive and connected rehabilitation systems for e-Health
CN110404243A (en) A kind of method of rehabilitation and rehabilitation system based on posture measurement
Brooks et al. Quantifying upper‐arm rehabilitation metrics for children through interaction with a humanoid robot
Tripathi et al. Advancing remote healthcare using humanoid and affective systems
García-Vergara et al. Super pop VR TM: An adaptable virtual reality game for upper-body rehabilitation
Orejuela-Zapata et al. Self-help devices for quadriplegic population: A systematic literature review
CN112230777A (en) Cognitive training system based on non-contact interaction
Hoda et al. Cloud-based rehabilitation and recovery prediction system for stroke patients
Yun et al. Animation fidelity in self-avatars: impact on user performance and sense of agency
Muhammad et al. Mono camera-based human skeletal tracking for squat exercise Abnormality detection using double Exponential smoothing
Zaher et al. A framework for assessing physical rehabilitation exercises
CN110473602B (en) Body state data collection processing method for wearable body sensing game device
Jawed et al. Rehabilitation posture correction using neural network
Lee et al. Augmented reality game system design for stroke rehabilitation application
Gao et al. Research on the usability of hand motor function training based on VR system
Dang et al. An interactive care system based on a depth image and EEG for aged patients with dementia
González et al. Vision based interface: an alternative tool for children with cerebral palsy
Muñoz et al. Dynamic systems theory in human movement exploring coordination patterns by angle-angle diagrams using Kinect

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20171219