CN105022471A - Device and method for carrying out gesture recognition based on pressure sensor array - Google Patents
Device and method for carrying out gesture recognition based on pressure sensor array Download PDFInfo
- Publication number
- CN105022471A CN105022471A CN201410162487.3A CN201410162487A CN105022471A CN 105022471 A CN105022471 A CN 105022471A CN 201410162487 A CN201410162487 A CN 201410162487A CN 105022471 A CN105022471 A CN 105022471A
- Authority
- CN
- China
- Prior art keywords
- signal
- tendon
- user
- refers
- pressure
- 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
Links
Abstract
A device for carrying out gesture recognition based on a pressure sensor array, comprising a muscle tendon pressure signal acquisition unit, a signal conversion unit, a data processing and gesture action identification unit and the like, and being characterized in that the pressure sensor array of the muscle tendon pressure signal acquisition unit is mounted on an inner side, close to skin of a user, of a wearing mechanism. And a method for carrying out the gesture recognition based on the pressure sensor array, comprising: step 1, acquiring a signal caused by detected muscle tendon pressure of a wrist of the user; step 2, extracting characteristic data for characterizing hand gestures or actions of the user of detected muscle tendon from the acquired signal; and step 3, analyzing the extracted characteristic data to determine the hand gestures or actions of the user. The gesture recognition method and device, provided by the invention, have the advantages of being simple in principle and high gesture recognition accuracy. Meanwhile, the utilization of the device is not limited by conditions; and the device can be combined with wearable equipment including intelligent watches, bracelets and bangles, so that the device is more easily popularized and utilized.
Description
Technical field
The present invention relates to a kind of apparatus and method of gesture identification, be specifically related to a kind of gesture identifying device based on array of pressure sensors and method, belong to field of human-computer interaction.
Background technology
The research of Gesture Recognition is one of the emphasis and focus of human-computer interaction technology research field.The input method of Gesture Recognition mainly contains: based on the gesture identification of data glove, the gesture identification of view-based access control model and the gesture identification based on human action electromyographic signal (SEMG).
The gesture identification method of view-based access control model is simple and easy to do, and equipment cost is cheap, and contactless gesture motion acquisition mode makes mutual naturality and comfortableness be greatly improved.But due to the uncertainty of vision, make it depend on the visual angle of cameras view, have shortcomings such as background, environment conversion bad adaptability.Gesture identification based on electromyographic signal is not subject to the impact of external environment change of background, and calculated amount is less, has better real-time.But due to the complicacy of muscle group structure, the impact such as individual difference, electrode position of electromyographic signal, add its classification difficulty.
Lianxiang (Beijing) Co., Ltd. proposes in August, 2012 patent of invention (number of patent application is 201210308642.9) that a key name is " pick-up unit, detection method and electronic equipment ", this invention proposes a kind of monitoring device be worn on user's arm, is mounted with array of pressure sensors inside it.The step that this invention proposes hand positions identification is:
Detect the current pressure that caused by user's arm by this device and produce the testing result of current pressure;
The posture of user's arm or hand is determined according to pressure sensing result;
Posture according to the arm judged or hand generates steering order.
This invention has following limitation:
1, the pressure that user's arm causes broadly is measured, comprise arm muscles group, pressure that bone causes, and tonometric emphasis is not arranged in the key position causing each position motion of user's hand, the position, tendon place of hand exercise is namely controlled by user's wrist.Thus cause the poor efficiency and high cost measured with user's hand exercise related pressure, also make to comprise too much useless interfere information in the data measured and cause the reduction of gesture identification accuracy rate;
2, the object of this invention is only the hand positions judging user by measuring user's arm pressure, and the principle of invention is not expanded to the user's hand motion recognition aspect representing more rich information content, has larger limitation.
Summary of the invention
For overcoming the limitation that existing gesture identification method exists, the present invention proposes a kind of gesture identification method based on array of pressure sensors.
The gesture motion of people's complexity has been come by the bone at the tendon traction each position of hand, is therefore distributed with abundant tendon group at positions such as people's palm, wrists.Tendon by means of only people's wrist portion just has flexor carpi radialis muscle tendon, long abductor muscle of thumb and musculus extensor brevis pollicis tendon, the long and short extensor tendon of carpi radialis, extensor pollicis longus muscle tendon etc.The gesture motion of the people overwhelming majority is all relevant with the motion by wrist portion tendon group.From the complexity of hand each position motion state and the number of corresponding tendon, user's wrist tendon number corresponding with thumb movement is maximum, corresponding user's wrist and thumb part the most responsive to the motion of tendon.
At the array of pressure sensors that user's tendon of wrist is settled by position, wrist portion shallow table tendon when user carries out gesture motion can be monitored and act on by skin the signal that cell pressure causes.Through obtaining the analysis of signal to sensor, the characteristic parameter information of user's finger, palm, Wrist-sport can be extracted.Can judge that the lax stressful situation of each tendon and tendon are in motion conditions that is transversely subepidermal or longitudinal direction according to characteristic parameter, by can reach the object identifying hand gestures and gesture motion to the analysis of above situation, the gesture motion of identification can be mapped to the input instruction of various electronic equipment further.
Accompanying drawing illustrates:
Fig. 1 is people's wrist main tendon profile figure;
Fig. 2 is that the present invention proposes tendon of wrist pressure-detecting device schematic appearance ();
Fig. 3 is that the present invention proposes tendon of wrist pressure-detecting device schematic appearance (two):
Fig. 4 is each unit connection diagram of gesture motion recognition device;
Fig. 5 represents the schematic diagram of gesture identification step;
Fig. 6 controls thumb to stretch tendon generation pressure schematic diagram corresponding to thumb movements;
Fig. 7 is that tendon produces pressure feature schematic diagram over time;
Fig. 8 is that tendon is relative to its entopic side-play amount feature schematic diagram;
Fig. 9 is the schematic diagram of user's effective action identification;
Figure 10 is the schematic diagram of the corresponding relation that user's gesture motion and numerical character input.
Fig. 1 is people's wrist main tendon profile figure, and the wrist of people mainly contains the organizational compositions such as bone 1, tendon 2, skin 3, referring now to Fig. 1, corresponding relation concise and to the point between each tendon and hand each position action is described:
Wrist joint (articulatio radiocarpea) is bending to be shunk cause primarily of radial flexor tendon 101, musculus flexor carpi ulnaris tendon 102, Tendon palmaris longus 103; Carpal stretching, extension and two lateral movements cause primarily of the contraction of Tendon of long radial extensor carpal muscle 104, musculus extensor carpi radilis brevis tendon 105, extensor carpi ulnaris muscle tendon 106.
Bending being caused by the bending of flexor superficialis tendon 107 of metacarpophalangeal joints (joint that palm is connected with finger), and the stretching, extension of metacarpophalangeal joints is controlled by extensor tendon 108.
The bending of 2-5 finger is controlled by flexor superficialis tendon 107, and its stretching, extension is controlled by extensor tendon 108, and wherein the stretching, extension of little finger of toe is controlled by extensor digiti minimi muscle tendon 109.
The bending of thumb is controlled by flexor pollicis longus muscle tendon 110, and the stretching routine of thumb is controlled by long abductor muscle of thumb tendon 111, musculus extensor brevis pollicis tendon 112, long extensor muscle of thumb tendon 113 3 tendons.
Fig. 2, Fig. 3 are the schematic appearance that the present invention proposes gesture motion recognition device.From the appearance, this gesture motion recognition device wears mechanism 5 by integrated processing unit 4, recognition device, array of pressure sensors 6, extendable members 7 form.Wear in mechanism 5 and be mounted with extensible link 7, extensible link 7 has the retractility in certain limit, makes to wear the wrist position that user is convenient to be worn in mechanism 5, and makes to wear the skin that mechanism's inside surface is close to user's wrist.Array of pressure sensors 6 is placed in the tendon of wrist wearing mechanism's inner surface and passes through position.Integrated processing unit 4 is placed in be worn mechanism 5 and is electrically connected with array of pressure sensors.
Fig. 4 is the annexation schematic diagram of each unit of action recognition device, and the present invention proposes gesture motion recognition device and is made up of tendon pressure signal collecting unit 8, signal conversion unit 9, storage unit 10, data processing and gesture motion recognition unit 11, wireless transmission unit 12, man-machine interaction unit 13 6 major part.
Array of pressure sensors collects the pressure signal that user's wrist is detected tendon generation, this signal is derived by traffic pilot, digital signal is converted to by A/D change-over circuit, data processing and gesture motion recognition unit 11 is flowed to by telecommunication circuit, data processing and gesture motion recognition unit 11 are by the Treatment Analysis to data, identify the gesture motion of user, then the gesture motion of identification is utilized to perform operational order or the input text information of the mapping of this gesture motion, or send steering order or input text information by wireless transport module 12 to miscellaneous equipment.By human-computer interaction module 13, user can select the pattern of gesture identification or select the equipment of gesture identifying device manipulation.By storage element 10, can the characteristic parameter information that user's specification input action extracts be stored.
The present invention proposes a kind of device carrying out gesture identification based on array of pressure sensors, comprise tendon pressure signal collecting unit 8, signal conversion unit 9, data processing and gesture motion recognition unit 11, storage unit 10, settle recognition device wear mechanism 5, it is characterized in that the array of pressure sensors 6 of tendon pressure signal collecting unit 8 is placed in and wear mechanism and to be close to the users the inner side of skin;
After user wears gesture identifying device, be placed in the array of pressure sensors 6 worn inside mechanism 5, the position being distributed in user's wrist portion or passing through close to wrist tendon relevant to hand exercise (tendon group).
Preferably, the mechanism 5 that wears belonging to gesture identifying device distributes in the ring texture being close to user's wrist portion skin.
Preferably, the mechanism 5 that wears belonging to gesture identifying device has extensible link 7.
Preferably, gesture identifying device can be integrated in the electronic equipments such as intelligent watch, bracelet, bracelet.
Preferably, the array of pressure sensors 6 of gesture identifying device is divided into row or multiple row.
Preferably, the circuit board wearing mechanism 5 of gesture identifying device adopts flexible PCB.
Preferably, gesture identifying device has wireless transport module 12.
The present invention proposes a kind of method of hand gestures, action recognition, user wear recognition device make various gesture motion time, realized the identification (with reference to Fig. 5) of hand gestures, action by following steps:
A1, data acquisition: collect the signal that each detected tendon pressure of user's wrist causes;
A2, feature extraction: the characteristic extracting detected tendon characterizing consumer hand positions or action from the signal obtained;
For the feature extracted in signal, should accomplish that the feature difference that different gesture motion classification extraction obtains is obvious as far as possible, be convenient to classification, the main target of feature extraction step that Here it is;
3, gesture motion identification: by the analysis of extracting characteristic, confirm hand gestures or the action of user.
Above method relate to wrist be detected tendon produce detection signal can be represented by certain electrical quantities (as electric current, voltage etc.).
The characteristic of characterizing consumer hand positions is that each detected tendon pressure of wrist causes signal for a position distribution feature for array of pressure sensors, and this feature is used to the posture feature judging user's hand.
Fig. 6 controls thumb to stretch the tendon generation pressure schematic diagram corresponding with thumb movements.Now by this figure, extensor pollicis longus muscle tendon, musculus extensor brevis pollicis tendon 16 Pressure Distribution for array of pressure sensors 14 and the corresponding relation of user's hand gestures are described.When Fig. 6 A represents that user's thumb attitude 19 is in the flattened state being close to forefinger, the extensor pollicis longus muscle tendon relevant to thumb stretching routine, the distribution of musculus extensor brevis pollicis tendon 16 under skin 15, and above tendon 16 acts on the position distribution feature 17 of pressure along sensor array 14 of array of pressure sensors 14.When Fig. 6 B represents that user's thumb attitude 19 is in the state outwards stretched, corresponding tendon distribution and pressure distribution state.As can be seen from the contrast of two width figure, under thumb is in the state shown in Fig. 6 A, extensor pollicis longus muscle tendon, musculus extensor brevis pollicis tendon 16 are comparatively lax in skin 15 times states, and its pressure for array of pressure sensors 14 is less, and pressure is comparatively mild along the distribution of sensor array 14 position; Under thumb is in the state shown in Fig. 6 B, extensor pollicis longus muscle tendon, musculus extensor brevis pollicis tendon 16 are comparatively nervous in skin 15 times states, its pressure acted on for array of pressure sensors 14 is comparatively large, and pressure exists comparatively significantly signal drop along the feature 17 of sensor array 14 position distribution.The Pressure Distribution 18 that dotted line represents is that user inputs the position distribution feature of the corresponding tendon pressure of the above standard operation of thumb along sensor array 14 in advance.
The characteristic of another kind of characterizing consumer hand gesture motion is that each detected tendon pressure of user's wrist causes signal for the variation characteristic of time, and this feature is used to the motion characteristic judging user's hand.
Produce pressure feature schematic diagram over time with reference to tendon shown in Fig. 7, illustrates that being detected tendon causes pressure signal for the variation characteristic of time.When the hand of user completes certain action, the tendon that is associated with this action produces pressure signal and presents feature over time, and there is corresponding relation between this variation characteristic and action of user, this corresponding relation can be applicable to the identification to hand motion.As shown in Figure 7, the action 19 of user's thumb is by during to forefinger position from the position of outwards stretching, the extensor pollicis longus muscle tendon relevant to thumb stretching, musculus extensor brevis pollicis tendon 16 act on the pressure of array of pressure sensors 14, in time in change from high to low shown in solid line figure 21.When dashed graph 22 presses standard operation under representing input thumb, the pressure presented, for temporal change characteristic, by the com-parison and analysis to tendon measured data and standard operation data, can judge the action of user's hand.
A kind of signal characteristic of extraction is also had to cause signal for array of pressure sensors position relative to its entopic side-play amount for user is detected tendon pressure, namely characterize the displacement that tendon does wrist transverse movement, and this side-play amount is relative to the variation characteristic of time.This side-play amount feature can be used for the auxiliary judgment of user's hand positions, also can be used for the judgement of user's gesture motion.
With reference to tendon shown in Fig. 8 relative to its entopic side-play amount feature schematic diagram, the corresponding relation between tendon side-play amount and the action of user's hand positions is described.When user carries out certain certain gestures action, the tendon be associated with this gesture motion will offset normal position and do moving relative to wrist transverse direction, thus makes this tendon produce the skew of pressure along the position distribution appearance correspondence of sensor array.As shown in Figure 8, when user's thumb does the action 19 of teeter to the inside, control the extensor pollicis longus muscle tendon of thumb sideway movement, musculus extensor brevis pollicis tendon 16 transverse shifting to the inside accordingly, the pressure that extensor pollicis longus muscle tendon, musculus extensor brevis pollicis tendon 16 are produced occurs offseting accordingly in the distribution of sensor array location.In Fig. 8 solid line figure 23 represent thumb extensor tendon skew after pressure distribution, dashed graph 24 represent thumb stretch normal extension state (thumb do not have teeter when) under tendon when normal position, pressure is along the position distribution of sensor array, the side-play amount of thumb extensor tendon represents with Δ X.There is certain corresponding relation to the offsets of relevant tendon in some gesture motion of such user, by the measurement to the specific tendon side-play amount of user, can play supplementary recognition reaction to the identification of user's gesture motion.
By extracting the analysis of characteristic to each detected tendon, judge posture or the action of user's hand, then can according to the application program of current system state or execution, according to the user's hand positions judged or action, send the operational order corresponding with this hand positions or action by wireless transport module to manipulation electronic equipment or input corresponding information.
For avoiding the hand of user when carrying out the irrelevant action of other and gesture identification, all kinds of control information of erroneous input, can be set to action recognition device activate and locking two states.Steering order or corresponding information that (going out) all kinds of gesture motion represents can be inputted in active state, and do not input (going out) steering order or corresponding information in the locked state.By button press, the touch-screen that slides, make the switching that the modes such as specific unblock gesture motion realize the activation of action recognition device and locking two states.
Because gesture identifying device wearer is due to the difference of age, sex, the thickness of its wrist and the distributing position of each tendon are not quite similar.In order to make the identification of gesture motion have higher accurately determining, need to determine that each detected tendon of user's wrist causes signal to be listed in the position in normality situation relative to sensor array.Therefore above gesture motion recognition methods also comprise each detected tendon cause signal relative to sensor array be listed in normality situation upper/lower positions measure step:
1, user according to system suggestion makes the gesture that its normal orientation measures agreement;
2, obtaining each detected tendon pressure of user's wrist causes signal along the position distribution feature of sensor array;
3, according to the position distribution feature of pressure signal, determine that each detected tendon causes the distributing position of signal in normality situation;
4, store each detected tendon at storage element and cause the position distribution feature of signal in normality situation.
As a kind of gesture of tendon its normal orientation determination step agreement, palm forwards stretches and sets level by user, and five fingers all stretch exceptionally straight, record the position of each tendon of user's wrist as each tendon its normal orientation under this posture.
For ease of comparing the characterizing consumer hand positions extracted or motion characteristic data analysis, gesture motion recognition methods also comprises standard hand motion typing step:
1, user according to system suggestion, takes turns doing various hand positions or action;
2, the signal that each detected tendon of user's wrist causes is collected;
3, from the signal obtained, extract the characteristic of corresponding tendon characterizing consumer hand positions or action;
4, this characteristic is stored by the storage element of system.
The gesture identification method that the present invention proposes and device, have principle simple, and measurement means is direct, and the interference suffered by measuring process is less, and the accuracy rate of gesture identification is high.The restrictive condition suffered by use of this device is few simultaneously, can combine with wearable devices such as intelligent watch, bracelet, bracelets, easier in utilization and extention.
Embodiment:
Embodiment 1:
When user's hand keeps certain posture, array of pressure sensors can obtain the signal that detected tendon pressure causes, and by analyzing generation signal, can obtain the signal value that each detected tendon is corresponding.By the analysis to each tendon respective signal value, judge the hand gestures of user.
Propose a kind of recognition methods of user's hand gestures in the present embodiment, when the hand of user keeps certain posture, realized the identification of gesture by following steps:
1, obtain by pressure-detecting device the pressure signal that each detected tendon causes;
2, the position distribution of each detected tendon is determined according to the detection signal gathered;
3, signal value corresponding to each detected tendon and the displacement along the relative its normal orientation of wrist is extracted;
4, by each detected tendon respective signal value and along wrist relative its normal orientation displacement quantitative analysis, the hand gestures of user is judged;
5, according to the hand positions identified, the manipulation instruction corresponding with this hand positions is sent to by controlling equipment.
More than carry out hand gestures and know method for distinguishing, when the pressure signal value that each detected tendon of extraction is corresponding, need the time determining that signal value extracts.Generally when the hand positions of user keeps geo-stationary, when the signal namely gathered is relatively steady, pressure signal value corresponding to each detected tendon should be extracted.Time by following steps determination signal value extracts:
1, obtain by pressure-detecting device the pressure signal that each detected tendon causes;
2, in this signals collecting scope sometime forward, the characteristic parameter of this time period characterization signal relative time rate of change is extracted;
If 3 characteristic parameters are less than the threshold value set, then from this signal gathered, extract the signal value of each detected tendon and judge, otherwise waiting for that next signal inputs.
The present embodiment provides the recognition methods of user's gesture, by each detected tendon corresponding pressure signal value of extraction is compared with the pressure signal value that the corresponding tendon of storage element stored user standard hand positions causes the hand positions judging user.Judge that the hand positions of user is identical with standard gestures when the parameter value of sign two kinds of signal differences is less than setting threshold value, otherwise wait for input next time.
Embodiment 2:
User's tendon of wrist pressure-detecting device that the present invention proposes, being detected the array of pressure sensors of tendon, detecting the pressure characteristic that this tendon acts on sensor by being distributed in user's wrist.The measurement signal value that in sensor array, each sensor obtains forms the distribution plan of pressure signal value about sensor array location relative to sensor array location, as shown in 17 in Fig. 6,
By by the pressure signal value of one or more tendon relevant to user's hand gestures about sensor array location distribution plan, with pressure signal value the comparing about sensor array location distribution plan of storage element storage standard action, the hand gestures of user can be judged.
Propose a kind of recognition methods of user's hand gestures in the present embodiment, when the hand of user keeps certain posture, adopt following steps to realize the identification of hand positions:
1, the signal that user's wrist pressure causes is gathered;
2, from the signal gathered, the distribution curve of signal value about sensor array location is extracted;
3, by extracting signal value about the analysis of sensor array location distribution curve, the hand gestures of user is judged;
4, according to the hand positions identified, the manipulation instruction corresponding with this hand positions is sent to by controlling equipment.
When carrying out the judgement of user's hand positions, the method pressure signal value of extraction compared about the signal value distribution curve of sensor array location distribution curve and the action of system storage standard can be adopted, the comparison of curve can adopt mathematically about the method that similarity of curves compares further, namely when the judgement metric of two similarity of curves is in a certain scope, judge that the action of two curve representatives is identical, otherwise continue waiting signal input.
More than carry out hand gestures and know method for distinguishing, when extracting the distribution curve of pressure signal value, needing the time determining that signal value curve extracts equally, the defining method identical with embodiment 1 can be adopted.
Specific embodiment 3:
The gesture motion of user is the consecutive variations of user's hand gestures relative time, and compared with the hand gestures of user's static state, the gesture motion of user comprises more abundant content, thus can represent more manipulation information.From the angle identified, the gesture motion of user comprises the quantity of information of multidimensional, is thus more easy to improve the accuracy identified.
The present embodiment propose a kind of method of carrying out hand motion recognition, user wear recognition device make various gesture motion time, realized the identification of hand motion by following steps:
1, data acquisition: collect user and be detected the signal that tendon pressure causes;
2, active segment monitoring: by the analysis to collection signal feature, determine the time point of characterizing consumer action continuous signal section starting point and terminal;
3, feature extraction: extract each detected tendon pressure from the active signal section that obtains and cause signal for the variation characteristic of time;
4, action recognition: by causing signal for the analysis of temporal change characteristic to each detected tendon, identifies the action that user performs;
5, according to the hand motion identified, the manipulation instruction corresponding with this hand hand motion is sent to by controlling equipment.
This method relates to the detection gathering continuous signal active segment, needs the starting point and the terminal that judge user action, namely determines the time point of characterizing consumer action continuous signal active segment starting point and terminal.Generally hand positions before and after user action should be kept the moment of geo-stationary, namely gather the time point that the relatively stable time before and after hand exercise signal segment is defined as continuous signal section starting point and terminal.Time point by following steps determination active segment starting point, terminal:
1, obtain by pressure-detecting device the pressure signal that each detected tendon causes;
2, in this signals collecting scope sometime forward, the characteristic parameter of this time period characterization signal relative time rate of change is extracted;
If 3 characteristic parameters are less than the threshold value set, then this time point gathered are defined as the time point of beginning or end, otherwise wait for that next signal inputs.
This method relates to each detected tendon and causes signal for the variation characteristic of time, comprises each tendon and causes signal value to cause signal along the relative its normal orientation displacement of wrist feature over time relative to the variation characteristic of time and tendon.
Each detected tendon causes measuring-signal for the analysis of temporal change characteristic, by obtaining relatively the completing of variation characteristic with in advance storage standard action corresponding variation characteristic data of signal for the time.
By the analysis to collection signal, each detected tendon of user can be obtained and cause signal value relative to the change curve of time.This curve can be can be used as a kind of extraction feature of characterizing consumer gesture motion, and utilize this feature to carry out the identification of gesture motion.The i.e. identification of user's gesture motion, by causing signal value to compare for the change curve of time and standard operation homologous thread by obtaining each detected tendon pressure.To the comparison of pressure signal value change curve, the method that similarity of curves can be adopted to compare is carried out.
According to the judgement to user's gesture motion, according to the corresponding relation in advance between setting gesture motion and manipulation instruction, to gesture identifying device self input manipulation instruction, or instruction or input information can be manipulated to other electronic equipment as PC terminal, intelligent watch, smart mobile phone, digital camera etc. send by the wireless module of gesture identifying device.Thus the inputting interface avoiding some electronic equipment is narrow and small, user inputs the drawback of difficulty.This gesture identifying device is convenient to user and is carried with, and has the advantage of user operation convenience and ease for use.Below provide the scheme of corresponding relation between operational order and input information under the various gesture motion of user and various pattern.
(1) user's gesture motion and the conventional corresponding relation manipulating instruction are as following table.
(2) under Graph Control pattern, the gesture motion controlled for three-dimensional picture sees the following form with the relation of corresponding instruction:
If this gesture motion recognition system can combine with gravity sensing device, operate more visual in image for realization with the three-dimensional picture of complexity, hand motion can be combined with the spinning movement of arm, form combined type control action.
(3) as shown in Figure 10, utilize gesture motion identification provided by the invention and control method, can realize the information input of numeral, text, the corresponding relation that user's gesture motion and numerical character input is as following table:
Wherein English alphabet inputs according to its input mode at mobile phone standard numeric keyboard.
Embodiment 4:
Utilize the signal that each tendon of pressure sensor monitoring user wrist obtains, not only by inputting all kinds of operational order to the identification of user's gesture motion, may be used for the real-time control to screen graph object equally, as the real-time control of the gesture motion of user realization to computer, TV screen cursor can be utilized.
The present embodiment proposes a kind of method of based on gesture identification, screen object being carried out to control in real time, is completed time under system is in real-time control mode by following steps:
1, data acquisition: collect the signal that each detected tendon action causes;
2, feature extraction: the characteristic extracting energy characterizing consumer hand exercise from the signal obtained;
3, the determination of side-play amount: by this collection signal is extracted characteristic and last time collection signal extracts the comparative analysis of characteristic, judge the continuous motion feature of user's hand, and confirmation is along the mobile numerical value of x-axis, y-axis;
4, manipulation instruction is exported: move numerical value according to confirmation x-axis, y-axis, manipulated equipment is sent and manipulates instruction accordingly.
The characteristic that the present invention proposes characterizing consumer hand exercise and extracts comprises signal value feature that detected tendon pressure causes and signal along the side-play amount of sensor array distributing position relative to its its normal orientation.
A kind of gesture motion of carrying out screen object real-time control: namely palm or fist do motion up and down around wrist joint, determined the mobile numerical value Δ x of screen control object transverse direction along the change in displacement of wrist transverse direction by signal location in the collection signal of twice, front and back, gather Tendon palmaris longus for twice by front and back, the mobile numerical value Δ y that manipulation object longitudinally controls is determined in the comparative analysis of extensor tendon signal extraction eigenwert, and control with mobile numerical value Δ x, the real time kinematics of Δ y to Drawing Objects such as screen cursors.
In enforcement manipulation process, need at any time to the method that the validity of real-time control controls: namely: thumb is close to forefinger when the side of tiger's jaw, palm or fist are regarded as effective control action by the action that wrist joint does, the instruction of input manipulation in real time; And the tiger's jaw of user is opened, when thumb separates with forefinger, the action that palm does is invalid control action, in this case without input manipulation instruction.The reset of palm action can be realized by invalid control action.
Can perform with mouse-click equally in the process implementing manipulation, double-click similar action, click forefinger by thumb and realize click function near the side of tiger's jaw; Double-click forefinger by thumb and realize double-click function near the side of tiger's jaw.
Said method of the present invention can realize in hardware, firmware, or be embodied as software or computer code (can be stored in recording medium such as CD ROM, RAM, floppy disk, hard disk or magneto-optic disk), when by computing machine, processor or hardware access and executive software or computer code, this software or computer code realize disposal route described herein.Term used herein " unit " comprises hardware (being such as configured with the microprocessor of machine executable code).
The present invention proposes the variation of data inputting method and embodiment put forward the methods sequence of steps, the merging of step or deconsolidation process, limits and does not enumerate, all should think and be included within protection scope of the present invention because of length.
Above-described is only the preferred embodiment of the present invention, the invention is not restricted to the scheme that above embodiment or embodiment are listed.Be appreciated that other improvement that those skilled in the art directly derive without departing from the basic idea of the present invention or associate and change all should be thought to be included within protection scope of the present invention.
Claims (31)
1. one kind is carried out the device of gesture identification based on array of pressure sensors, comprise tendon pressure signal collecting unit (8), signal conversion unit (9), data processing and gesture motion recognition unit (11), storage unit (10), settle recognition device wear mechanism (5), it is characterized in that the array of pressure sensors (6) of tendon pressure signal collecting unit (8) is placed in and wear mechanism and to be close to the users the inner side of skin;
After user wears gesture identifying device, be placed in the array of pressure sensors (6) wearing mechanism (5) inner side, the position being distributed in user's wrist portion or passing through close to wrist tendon relevant to hand exercise or tendon group.
2. the mechanism (5) that wears as claimed in claim 1 belonging to gesture identifying device distributes in the ring texture being close to user's wrist portion skin.
3. the mechanism (5) that wears as claimed in claim 1 belonging to gesture identifying device has extensible link (7).
4. gesture identifying device can be placed in the electronic equipments such as intelligent watch, bracelet, bracelet as claimed in claim 1.
5. the array of pressure sensors (6) of gesture identifying device is divided into row or multiple row as claimed in claim 1.
6. wear mechanism (5) circuit board as claimed in claim 1 belonging to gesture identifying device and adopt flexible PCB.
7. gesture identifying device has wireless transport module (12) and man-machine interaction unit (13) as claimed in claim 1.
8. carry out a method for user's hand gestures, action recognition based on sensor array, comprise step:
A1, data acquisition: collect the signal that each detected tendon pressure of user's wrist causes;
A2, feature extraction: the characteristic extracting detected tendon characterizing consumer hand positions or action from the signal obtained;
A3, gesture motion identification: by the analysis of extracting characteristic, confirm hand gestures or the action of user.
9. method as claimed in claim 8, is characterized in that described characteristic comprises:
Detected tendon pressure cause signal for array of pressure sensors position distribution feature,
Detected tendon pressure cause signal for the time variation characteristic,
Detected tendon pressure cause signal for array of pressure sensors position relative to its entopic side-play amount and this side-play amount relative to one or more of the above feature of the variation characteristic of time.
10. method as claimed in claim 8, is characterized in that comprising the user's hand positions according to judging or action, is sent the operational order corresponding with this hand positions or action by wireless transport module or inputs corresponding information to manipulation electronic equipment.
11. methods as claimed in claim 8, is characterized in that comprising by button press, the touch-screen that slides, make the switching that the modes such as specific unblock gesture motion realize the activation of action recognition device and locking two states.
12. methods as claimed in claim 8, is characterized in that comprising the step that each detected tendon causes signal to measure relative to sensor array its normal orientation:
User according to system suggestion makes the gesture that its normal orientation measures agreement;
Obtaining each detected tendon pressure of user's wrist causes signal along the position distribution feature of sensor array;
According to the position distribution feature of pressure signal, determine that each detected tendon causes the distributing position of signal in normality situation;
Store each detected tendon at storage element and cause the position distribution feature of signal in normality situation.
13. methods as claimed in claim 8, is characterized in that comprising standard hand motion typing step:
User according to system suggestion, takes turns doing various hand positions or action;
Collect the signal that each detected tendon of user's wrist causes;
The characteristic of corresponding tendon characterizing consumer hand positions or action is extracted from the signal obtained;
This characteristic is stored by the storage element of system.
The recognition methods of 14. 1 kinds of user's hand gestures, comprises the following steps:
The pressure signal that each detected tendon causes is obtained by pressure-detecting device;
The position distribution of each detected tendon is determined according to the detection signal gathered;
Extract signal value corresponding to each detected tendon and the displacement along the relative its normal orientation of wrist;
By to each detected tendon respective signal value and along wrist relative its normal orientation displacement quantitative analysis, judge the hand gestures of user;
According to the hand positions identified, send the manipulation instruction corresponding with this hand positions to by controlling equipment.
15. methods as claimed in claim 14, is characterized in that the determining step comprising signal value extraction time point:
The pressure signal that each detected tendon causes is obtained by pressure-detecting device;
In this signals collecting scope sometime forward, extract the characteristic parameter of this time period characterization signal relative time rate of change;
If characteristic parameter is less than the threshold value set, then from this signal gathered, extracts each detected tendon signal value and judge, otherwise waiting for that next signal inputs.
16. methods as claimed in claim 14, is characterized in that the judgement of described user's hand gestures, by causing relatively having come of pressure signal value to each detected tendon corresponding pressure signal value with the corresponding tendon of storage element stored user standard hand positions.
The recognition methods of 17. 1 kinds of user's hand gestures, realizes the identification of gesture by following steps:
Gather the signal that user's wrist pressure causes;
The distribution curve of signal value about sensor array location is extracted from the signal gathered;
By to extracting signal value about the analysis of sensor array location distribution curve, judge the hand gestures of user;
According to the hand positions identified, send the manipulation instruction corresponding with this hand positions to by controlling equipment.
18. methods as claimed in claim 17, it is characterized in that the analysis of described curve, adopt the method pressure signal value of extraction compared about sensor array location distribution curve and system storage standard actuating signal Distribution value curve, adopting mathematically about the method that similarity of curves compares relatively further of curve.
The recognition methods of 19. 1 kinds of user's hand motions, realizes the identification of gesture motion by following steps:
Data acquisition: collect user and be detected the signal that tendon pressure causes;
Active segment is monitored: by the analysis to collection signal feature, determine the time point of user action continuous signal section starting point and terminal;
Feature extraction: extract each detected tendon pressure from the active signal section that obtains and cause signal for the variation characteristic of time;
Action recognition: by causing signal for the analysis of temporal change characteristic to each detected tendon, identifies the action that user performs;
According to the hand motion identified, send the manipulation instruction corresponding with this hand hand motion to by controlling equipment.
20. methods as claimed in claim 19, is characterized in that comprising the step of the time point determining active segment starting point, terminal:
The pressure signal that each detected tendon causes is obtained by pressure-detecting device;
In this signals collecting scope sometime forward, extract the characteristic parameter of this time period characterization signal relative time rate of change;
If characteristic parameter is less than the threshold value set, then this time point gathered is defined as the time point of beginning or end, otherwise waits for that next signal inputs.
21. methods as claimed in claim 19, it is characterized in that described each detected tendon causes signal for the variation characteristic of time, comprise each tendon and cause signal value to cause signal along the relative its normal orientation displacement of wrist feature over time relative to the variation characteristic of time and tendon.
22. methods as claimed in claim 19, is characterized in that the described judgement to user's hand motion, by obtaining relatively the completing of variation characteristic with in advance storage standard action corresponding variation characteristic data of signal for the time.
23. methods as claimed in claim 19, it is characterized in that the described judgement to user's hand motion, by causing signal value to compare for the change curve of time and standard operation homologous thread by obtaining each detected tendon pressure, the comparison of curve, the method adopting similarity of curves to compare further completes.
24. comprise following corresponding relation between gesture motion and manipulation instruction as claimed in claim 19
。
25. gesture motion and Graph Control instruction comprise and there is following corresponding relation as claimed in claim 19
。
26. comprise following corresponding relation as claimed in claim 19 between gesture motion and inputting digital character, wherein English alphabet inputs according to its input mode at mobile phone standard numeric keyboard
。
27. 1 kinds of methods of carrying out controlling in real time to screen object based on gesture identification, are completed by following steps:
Data acquisition: collect the signal that each detected tendon action causes;
Feature extraction: the characteristic extracting energy characterizing consumer hand exercise from the signal obtained;
The determination of side-play amount: by this collection signal extract characteristic and last time collection signal extract the comparative analysis of characteristic, judge the continuous motion feature of user's hand, and confirm the mobile numerical value of x-axis, y-axis;
Export manipulation instruction: move numerical value according to confirmation x-axis, y-axis, manipulated equipment is sent and manipulates instruction accordingly.
28. methods as claimed in claim 27, is characterized in that described characteristic comprises signal value feature that detected tendon pressure causes and signal along the side-play amount of sensor array distributing position relative to its its normal orientation.
29. methods as claimed in claim 27, is characterized in that the motion that described user's hand exercise does around wrist joint for palm or fist.
30. methods as claimed in claim 27, is characterized in that comprising the method controlled the validity of real-time control:
Thumb is close to forefinger when the side of tiger's jaw, and palm, fist are regarded as effective control action by the action that wrist joint does, the instruction of input manipulation in real time; And user's thumb is when separating with forefinger, the action that palm does is invalid control action, in this case without input manipulation instruction.
31. methods as claimed in claim 27, is characterized in that clicking forefinger by thumb realizes click function near the side of tiger's jaw; Double-click forefinger by thumb and realize double-click function near the side of tiger's jaw.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410162487.3A CN105022471A (en) | 2014-04-23 | 2014-04-23 | Device and method for carrying out gesture recognition based on pressure sensor array |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410162487.3A CN105022471A (en) | 2014-04-23 | 2014-04-23 | Device and method for carrying out gesture recognition based on pressure sensor array |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105022471A true CN105022471A (en) | 2015-11-04 |
Family
ID=54412501
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410162487.3A Pending CN105022471A (en) | 2014-04-23 | 2014-04-23 | Device and method for carrying out gesture recognition based on pressure sensor array |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105022471A (en) |
Cited By (44)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105511615A (en) * | 2015-12-04 | 2016-04-20 | 深圳大学 | Wearable text input system and method based on EMG |
| CN105717059A (en) * | 2016-02-17 | 2016-06-29 | 中山大学 | Caloric intake automatic measuring method and system based on spectrum analysis |
| CN106236098A (en) * | 2016-08-16 | 2016-12-21 | 京东方科技集团股份有限公司 | Wearable device, health detecting system based on wearable device and method |
| CN106354415A (en) * | 2016-10-08 | 2017-01-25 | 努比亚技术有限公司 | Terminal and method for recognizing user gesture thereof |
| CN106445127A (en) * | 2016-09-13 | 2017-02-22 | 努比亚技术有限公司 | Method and system for terminal control |
| CN106484101A (en) * | 2016-09-14 | 2017-03-08 | 深圳市金立通信设备有限公司 | A kind of terminal control method and equipment |
| CN106569610A (en) * | 2016-11-09 | 2017-04-19 | 李飞洋 | Method for realizing electronic equipment input function through thumb bending degree testing |
| CN106909228A (en) * | 2017-05-08 | 2017-06-30 | 电子科技大学 | A kind of positioning input device of utilization head twisting sensing |
| CN106951109A (en) * | 2017-03-31 | 2017-07-14 | 殷超 | A kind of method and its device for gathering hand gestures |
| CN107015645A (en) * | 2017-03-24 | 2017-08-04 | 广州幻境科技有限公司 | A kind of character input method based on gesture |
| CN107132919A (en) * | 2017-04-28 | 2017-09-05 | 广州幻境科技有限公司 | A kind of Chinese-character stroke input method based on gesture |
| CN107145236A (en) * | 2017-05-12 | 2017-09-08 | 中国科学技术大学 | A kind of gesture identification method and system based on tendon of wrist pressure correlation characteristic |
| CN107145233A (en) * | 2017-04-28 | 2017-09-08 | 广州幻境科技有限公司 | A kind of odd even stroke Chinese character input method based on gesture |
| CN107137092A (en) * | 2017-07-17 | 2017-09-08 | 中国科学院心理研究所 | A kind of operational motion gesture induces detecting system and its method |
| CN107329574A (en) * | 2017-06-30 | 2017-11-07 | 联想(北京)有限公司 | Input method and system for electronic equipment |
| CN107403178A (en) * | 2017-08-08 | 2017-11-28 | 方超 | Gesture acquisition system |
| CN107783642A (en) * | 2016-08-24 | 2018-03-09 | 中国航天员科研训练中心 | A kind of wrist gesture identification equipment |
| CN108536291A (en) * | 2018-03-29 | 2018-09-14 | 努比亚技术有限公司 | A kind of application operating method, wearable device and storage medium |
| CN108775681A (en) * | 2018-08-22 | 2018-11-09 | 广东美的制冷设备有限公司 | Air conditioner and its control method, device and computer readable storage medium |
| CN108897444A (en) * | 2018-06-21 | 2018-11-27 | 中国科学技术大学 | The method and system of cursor control are realized using wearable wristband type universal serial mouse |
| CN109426330A (en) * | 2017-08-22 | 2019-03-05 | 南昌欧菲显示科技有限公司 | wearable device and operation method thereof |
| CN109635820A (en) * | 2018-08-06 | 2019-04-16 | 浙江大学 | The construction method of Parkinson's disease bradykinesia video detection model based on deep neural network |
| CN109715065A (en) * | 2016-08-15 | 2019-05-03 | 乔治亚技术研究公司 | Electronic equipment and its control method |
| CN109765996A (en) * | 2018-11-23 | 2019-05-17 | 华东师范大学 | Insensitive gesture detection system and method are deviated to wearing position based on FMG armband |
| CN109800733A (en) * | 2019-01-30 | 2019-05-24 | 中国科学技术大学 | Data processing method and device, electronic equipment |
| WO2019127593A1 (en) * | 2017-12-31 | 2019-07-04 | 李庆远 | Toe ring gesturing method |
| CN109976518A (en) * | 2019-03-12 | 2019-07-05 | 合肥工业大学 | A kind of man-machine interaction method based on PVDF sensor array |
| CN109992090A (en) * | 2017-12-29 | 2019-07-09 | 展达通讯（苏州）有限公司 | Input sensing system, input sensing device and its operating method |
| CN110083247A (en) * | 2019-04-30 | 2019-08-02 | 努比亚技术有限公司 | Control wearable device operating method, device, wearable device and storage medium |
| CN110308796A (en) * | 2019-07-08 | 2019-10-08 | 合肥工业大学 | A kind of finger movement recognition methods based on wrist PVDF sensor array |
| CN110623673A (en) * | 2019-09-29 | 2019-12-31 | 华东交通大学 | Fully-flexible intelligent wrist strap for recognizing gestures of driver |
| CN110658889A (en) * | 2019-09-23 | 2020-01-07 | 上海闻泰信息技术有限公司 | Wearable device control method, wearable device control device, wearable device control equipment and storage medium |
| CN110908515A (en) * | 2019-11-27 | 2020-03-24 | 北京航空航天大学 | Gesture recognition method and device based on wrist muscle pressure |
| US10684693B2 (en) | 2017-03-02 | 2020-06-16 | Samsung Electronics Co., Ltd. | Method for recognizing a gesture and an electronic device thereof |
| CN111376246A (en) * | 2018-12-27 | 2020-07-07 | 深圳市优必选科技有限公司 | Robot operation control method, gesture recognition device and robot |
| CN111558204A (en) * | 2020-03-25 | 2020-08-21 | 威海威高医疗系统有限公司 | Intelligent grip force auxiliary glove guider |
| CN112085972A (en) * | 2019-06-12 | 2020-12-15 | 广东小天才科技有限公司 | Learning habit monitoring and reminding method and system |
| CN112617839A (en) * | 2021-01-25 | 2021-04-09 | 杭州电子科技大学 | Sensing array and system for muscle pressure signal acquisition |
| CN112817443A (en) * | 2021-01-22 | 2021-05-18 | 歌尔科技有限公司 | Display interface control method, device and equipment based on gestures and storage medium |
| CN113197569A (en) * | 2021-04-23 | 2021-08-03 | 华中科技大学 | Human body intention recognition sensor based on friction power generation and recognition method thereof |
| WO2022067963A1 (en) * | 2020-09-29 | 2022-04-07 | 泰州翔升科技服务有限公司 | Computer control method and wearable device |
| CN114997227A (en) * | 2022-05-30 | 2022-09-02 | 中国科学院长春应用化学研究所 | Gesture recognition system based on muscle stress |
| CN116167035A (en) * | 2023-04-14 | 2023-05-26 | 深圳曼瑞德科技有限公司 | Method and system for carrying out identity recognition by collecting hand actions of intelligent watch |
| WO2024000413A1 (en) * | 2022-06-30 | 2024-01-04 | Intel Corporation | Technologies for detection of wrist posture |
-
2014
- 2014-04-23 CN CN201410162487.3A patent/CN105022471A/en active Pending
Cited By (59)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017092225A1 (en) * | 2015-12-04 | 2017-06-08 | 深圳大学 | Emg-based wearable text input system and method |
| CN105511615A (en) * | 2015-12-04 | 2016-04-20 | 深圳大学 | Wearable text input system and method based on EMG |
| CN105511615B (en) * | 2015-12-04 | 2019-03-05 | 深圳大学 | Wearable text input system and method based on EMG |
| CN105717059A (en) * | 2016-02-17 | 2016-06-29 | 中山大学 | Caloric intake automatic measuring method and system based on spectrum analysis |
| US11389084B2 (en) | 2016-08-15 | 2022-07-19 | Georgia Tech Research Corporation | Electronic device and method of controlling same |
| CN109715065A (en) * | 2016-08-15 | 2019-05-03 | 乔治亚技术研究公司 | Electronic equipment and its control method |
| CN106236098A (en) * | 2016-08-16 | 2016-12-21 | 京东方科技集团股份有限公司 | Wearable device, health detecting system based on wearable device and method |
| CN106236098B (en) * | 2016-08-16 | 2019-03-08 | 京东方科技集团股份有限公司 | Wearable device, the movement detection systems based on wearable device and method |
| CN107783642A (en) * | 2016-08-24 | 2018-03-09 | 中国航天员科研训练中心 | A kind of wrist gesture identification equipment |
| CN106445127A (en) * | 2016-09-13 | 2017-02-22 | 努比亚技术有限公司 | Method and system for terminal control |
| CN106484101A (en) * | 2016-09-14 | 2017-03-08 | 深圳市金立通信设备有限公司 | A kind of terminal control method and equipment |
| CN106354415A (en) * | 2016-10-08 | 2017-01-25 | 努比亚技术有限公司 | Terminal and method for recognizing user gesture thereof |
| CN106354415B (en) * | 2016-10-08 | 2020-05-26 | 瑞安市辉煌网络科技有限公司 | Terminal and method for recognizing user gesture |
| CN106569610A (en) * | 2016-11-09 | 2017-04-19 | 李飞洋 | Method for realizing electronic equipment input function through thumb bending degree testing |
| US10684693B2 (en) | 2017-03-02 | 2020-06-16 | Samsung Electronics Co., Ltd. | Method for recognizing a gesture and an electronic device thereof |
| CN107015645A (en) * | 2017-03-24 | 2017-08-04 | 广州幻境科技有限公司 | A kind of character input method based on gesture |
| CN106951109B (en) * | 2017-03-31 | 2020-02-14 | 殷超 | Method and device for acquiring hand gesture |
| CN106951109A (en) * | 2017-03-31 | 2017-07-14 | 殷超 | A kind of method and its device for gathering hand gestures |
| CN107145233A (en) * | 2017-04-28 | 2017-09-08 | 广州幻境科技有限公司 | A kind of odd even stroke Chinese character input method based on gesture |
| CN107132919A (en) * | 2017-04-28 | 2017-09-05 | 广州幻境科技有限公司 | A kind of Chinese-character stroke input method based on gesture |
| CN106909228B (en) * | 2017-05-08 | 2020-06-26 | 电子科技大学 | Positioning input device using head twisting induction |
| CN106909228A (en) * | 2017-05-08 | 2017-06-30 | 电子科技大学 | A kind of positioning input device of utilization head twisting sensing |
| CN107145236B (en) * | 2017-05-12 | 2020-02-07 | 中国科学技术大学 | Gesture recognition method and system based on wrist tendon pressure related characteristics |
| CN107145236A (en) * | 2017-05-12 | 2017-09-08 | 中国科学技术大学 | A kind of gesture identification method and system based on tendon of wrist pressure correlation characteristic |
| CN107329574A (en) * | 2017-06-30 | 2017-11-07 | 联想(北京)有限公司 | Input method and system for electronic equipment |
| CN107137092B (en) * | 2017-07-17 | 2024-03-08 | 中国科学院心理研究所 | Operation gesture induction detection system and method thereof |
| CN107137092A (en) * | 2017-07-17 | 2017-09-08 | 中国科学院心理研究所 | A kind of operational motion gesture induces detecting system and its method |
| CN107403178A (en) * | 2017-08-08 | 2017-11-28 | 方超 | Gesture acquisition system |
| CN109426330A (en) * | 2017-08-22 | 2019-03-05 | 南昌欧菲显示科技有限公司 | wearable device and operation method thereof |
| CN109992090A (en) * | 2017-12-29 | 2019-07-09 | 展达通讯（苏州）有限公司 | Input sensing system, input sensing device and its operating method |
| WO2019127593A1 (en) * | 2017-12-31 | 2019-07-04 | 李庆远 | Toe ring gesturing method |
| CN108536291A (en) * | 2018-03-29 | 2018-09-14 | 努比亚技术有限公司 | A kind of application operating method, wearable device and storage medium |
| CN108897444A (en) * | 2018-06-21 | 2018-11-27 | 中国科学技术大学 | The method and system of cursor control are realized using wearable wristband type universal serial mouse |
| CN109635820A (en) * | 2018-08-06 | 2019-04-16 | 浙江大学 | The construction method of Parkinson's disease bradykinesia video detection model based on deep neural network |
| CN109635820B (en) * | 2018-08-06 | 2020-07-17 | 浙江大学 | Construction method of Parkinson's disease bradykinesia video detection model based on deep neural network |
| CN108775681A (en) * | 2018-08-22 | 2018-11-09 | 广东美的制冷设备有限公司 | Air conditioner and its control method, device and computer readable storage medium |
| CN109765996A (en) * | 2018-11-23 | 2019-05-17 | 华东师范大学 | Insensitive gesture detection system and method are deviated to wearing position based on FMG armband |
| CN111376246A (en) * | 2018-12-27 | 2020-07-07 | 深圳市优必选科技有限公司 | Robot operation control method, gesture recognition device and robot |
| CN109800733A (en) * | 2019-01-30 | 2019-05-24 | 中国科学技术大学 | Data processing method and device, electronic equipment |
| CN109976518A (en) * | 2019-03-12 | 2019-07-05 | 合肥工业大学 | A kind of man-machine interaction method based on PVDF sensor array |
| CN110083247A (en) * | 2019-04-30 | 2019-08-02 | 努比亚技术有限公司 | Control wearable device operating method, device, wearable device and storage medium |
| CN112085972A (en) * | 2019-06-12 | 2020-12-15 | 广东小天才科技有限公司 | Learning habit monitoring and reminding method and system |
| CN112085972B (en) * | 2019-06-12 | 2022-08-23 | 广东小天才科技有限公司 | Learning habit monitoring and reminding method and system |
| CN110308796A (en) * | 2019-07-08 | 2019-10-08 | 合肥工业大学 | A kind of finger movement recognition methods based on wrist PVDF sensor array |
| CN110308796B (en) * | 2019-07-08 | 2022-12-02 | 合肥工业大学 | Finger motion identification method based on wrist PVDF sensor array |
| CN110658889A (en) * | 2019-09-23 | 2020-01-07 | 上海闻泰信息技术有限公司 | Wearable device control method, wearable device control device, wearable device control equipment and storage medium |
| CN110623673B (en) * | 2019-09-29 | 2022-01-28 | 华东交通大学 | Fully-flexible intelligent wrist strap for recognizing gestures of driver |
| CN110623673A (en) * | 2019-09-29 | 2019-12-31 | 华东交通大学 | Fully-flexible intelligent wrist strap for recognizing gestures of driver |
| CN110908515A (en) * | 2019-11-27 | 2020-03-24 | 北京航空航天大学 | Gesture recognition method and device based on wrist muscle pressure |
| CN111558204A (en) * | 2020-03-25 | 2020-08-21 | 威海威高医疗系统有限公司 | Intelligent grip force auxiliary glove guider |
| CN111558204B (en) * | 2020-03-25 | 2024-01-12 | 威海威高医疗系统有限公司 | Intelligent grip auxiliary glove guider |
| WO2022067963A1 (en) * | 2020-09-29 | 2022-04-07 | 泰州翔升科技服务有限公司 | Computer control method and wearable device |
| CN112817443A (en) * | 2021-01-22 | 2021-05-18 | 歌尔科技有限公司 | Display interface control method, device and equipment based on gestures and storage medium |
| CN112617839A (en) * | 2021-01-25 | 2021-04-09 | 杭州电子科技大学 | Sensing array and system for muscle pressure signal acquisition |
| CN113197569A (en) * | 2021-04-23 | 2021-08-03 | 华中科技大学 | Human body intention recognition sensor based on friction power generation and recognition method thereof |
| CN114997227A (en) * | 2022-05-30 | 2022-09-02 | 中国科学院长春应用化学研究所 | Gesture recognition system based on muscle stress |
| WO2024000413A1 (en) * | 2022-06-30 | 2024-01-04 | Intel Corporation | Technologies for detection of wrist posture |
| CN116167035A (en) * | 2023-04-14 | 2023-05-26 | 深圳曼瑞德科技有限公司 | Method and system for carrying out identity recognition by collecting hand actions of intelligent watch |
| CN116167035B (en) * | 2023-04-14 | 2023-06-27 | 深圳曼瑞德科技有限公司 | Method and system for carrying out identity recognition by collecting hand actions of intelligent watch |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105022471A (en) | Device and method for carrying out gesture recognition based on pressure sensor array | |
| McIntosh et al. | EMPress: Practical hand gesture classification with wrist-mounted EMG and pressure sensing | |
| Zhang et al. | Recognizing hand gestures with pressure-sensor-based motion sensing | |
| CN107480697B (en) | Myoelectric gesture recognition method and system | |
| Cheng et al. | Visualization of activated muscle area based on sEMG | |
| US20150109202A1 (en) | Systems, articles, and methods for gesture identification in wearable electromyography devices | |
| CN104267813A (en) | Method for wristband and bracelet type products to realize input or selection through ten kinds of gestures | |
| CN102402289B (en) | Mouse recognition method for gesture based on machine vision | |
| US10521018B2 (en) | Human body-based interaction method and interaction apparatus | |
| Huang et al. | An EMG-based handwriting recognition through dynamic time warping | |
| CN102184011A (en) | Human-computer interaction equipment | |
| Wan et al. | A new subtle hand gestures recognition algorithm based on EMG and FSR | |
| CN113849068B (en) | Understanding and interaction method and system for multi-modal information fusion of gestures | |
| CN104536574A (en) | Glove type input device and input method thereof | |
| CN104966011A (en) | Method for non-collaborative judgment and operating authorization restriction for mobile terminal child user | |
| Khodabandelou et al. | Attention-based gated recurrent unit for gesture recognition | |
| CN110618754A (en) | Surface electromyogram signal-based gesture recognition method and gesture recognition armband | |
| Jiang et al. | Development of a real-time hand gesture recognition wristband based on sEMG and IMU sensing | |
| CN106886741A (en) | A kind of gesture identification method of base finger identification | |
| CN110866468A (en) | Gesture recognition system and method based on passive RFID | |
| Lee et al. | Taiwan sign language (TSL) recognition based on 3D data and neural networks | |
| CN107450672B (en) | Wrist type intelligent device with high recognition rate | |
| CN101630193A (en) | Hand induction equipment | |
| CN109283998A (en) | Three-dimensional capacitive wearable human-computer interaction device and method | |
| CN106020442A (en) | Sensing method for intelligent sensing glove |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20151104 |