CN207037600U - A kind of Sign Language Recognition translates armlet - Google Patents
A kind of Sign Language Recognition translates armlet Download PDFInfo
- Publication number
- CN207037600U CN207037600U CN201720816796.7U CN201720816796U CN207037600U CN 207037600 U CN207037600 U CN 207037600U CN 201720816796 U CN201720816796 U CN 201720816796U CN 207037600 U CN207037600 U CN 207037600U
- Authority
- CN
- China
- Prior art keywords
- sign language
- armlet
- processor
- language recognition
- module
- 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.)
- Active
Links
Landscapes
- User Interface Of Digital Computer (AREA)
Abstract
The utility model provides a kind of Sign Language Recognition translation armlet, including nonrigid connector and multiple housings, multiple housings are connected by nonrigid connector form armlet body respectively, and described nonrigid connector can carry out elastic telescopic with so that armlet body is suitable wears in different arms;Surface myoelectric sensor is correspondingly arranged on the multiple housing respectively, inertial sensor is set at least one housing, processor is set at least one housing, each surface myoelectric sensor, inertial sensor electrically connect with processor respectively;It is provided for showing the display module of sign language interpreter information on the outer wall of at least one housing.The nonrigid connector of armlet of the present utility model is made by rubber, can be adjusted according to the forearm thickness of wearer elastic;Moreover, armlet can be by the text importing after identification on armlet screen, aphasic or communicatee can directly obtain, convenient direct.
Description
Technical field
The intelligent translation technical field of sign language is the utility model is related to, in particular it relates to which a kind of Sign Language Recognition translates armlet.
Background technology
In daily life, hearing-impaired people can only be exchanged with sign language with other people.Exchanged in lives such as shopping, traffic
During run into it is many not convenient.
Existing Sign Language Recognition can be divided into:Non- wearable hand language recognition device and wearable hand language recognition device.
For non-wearable hand language recognition device, mainly by gathering sign language action message by visual apparatus such as cameras,
By image procossing, gesture is identified.This mode needs to carry out the collection of sign language image by the camera of specialty, and whole
Camera is needed to keep sign language image to obtain against hearing-impaired people in sign language communication process, therefore, which is not only to equipment requirement
Higher, operation is relatively complicated, and is unfavorable for simple direct communication.
For wearable hand language recognition device, using the information collecting device being worn on object, corresponding actions are obtained
Information, so as to identify gesture.What application was more at present is a kind of Sign Language Recognition gloves, and a series of sensors are set in the gloves,
The translation of sign language is realized in action by gathering finger.From the point of view of practical application, Sign Language Recognition gloves costly, are not easy mostly
Popularization, and usage experience is poor.
In addition, existing Sign Language Recognition equipment is required for supporting intelligent terminal, by by the data of collection at intelligent end
End equipment carries out Treatment Analysis, and aphasic needs to hold in exchange terminal device and to obtain the translation knot on intelligent terminal
Fruit, usage experience are poor.
In view of this, it is special to propose the utility model.
Utility model content
The purpose of this utility model is to provide a kind of Sign Language Recognition translation armlet, and in order to achieve the above object, this practicality is new
Type adopts the following technical scheme that:
A kind of Sign Language Recognition translates armlet, including nonrigid connector and multiple housings, and multiple housings are connected by elasticity respectively
Junctor, which connects, forms armlet body, and described nonrigid connector can carry out elastic telescopic with so that armlet body is suitable wears in not
Same arm;
Surface myoelectric sensor is correspondingly arranged on the multiple housing respectively, inertia sensing is set at least one housing
Device, processor is set at least one housing, each surface myoelectric sensor, inertial sensor are electrically connected with processor respectively
Connect;
It is provided for showing the display module of sign language interpreter information on the outer wall of at least one housing.
Further, including differential-mode amplifier, the signal output part access differential mode of described surface myoelectric sensor amplify
Device carries out signal amplification.
Further, including A/D converter, the input of A/D converter and the signal output part of differential-mode amplifier connect,
The electric signal that surface myoelectric sensor gathers is converted to data signal by A/D converter;The output end connection of the A/D converter
Processor.
Further, the input of the signal output part connection A/D converter of described inertial sensor, A/D converter
The electric signal that inertial sensor gathers is converted into data signal.
Further, including data communication module, described processor are carried out with intelligent terminal by communication module
The signal that surface myoelectric sensor, inertial sensor gather is transferred to intelligent end by communication connection, processor by communication signal
End equipment, intelligent terminal, as algorithm basis, are aided with noise reduction process algorithm and feature carry using convolutional neural networks technology
Algorithm is taken, carries out sign language interpreter, and the word after translation is returned into processor with real on display module by communication module
When show.
Further, including battery and the charging inlet that is electrically connected with battery.
Further, Cao Zuoanjian is set positioned at display module bottom on the housing.
Further, including shock module, described shock module electrically connect with processor, and processor receives sign language and turned over
Control shock module vibration reminding user to translate after translation word to complete.
Further, including voice conversion module and voice playing module, described voice conversion module, speech play mould
Block electrically connects with processor respectively, and processor is sent to voice conversion module, voice modulus of conversion after receiving sign language interpreter word
Sign language interpreter text conversion is voice and played out by voice playing module by block.
Further, described housing includes at least eight, and described nonrigid connector is rubber strip, each two housing it
Between pass through at least one rubber band connection.
The Sign Language Recognition armlet of the present invention has the characteristics of following:
1st, Sign Language Recognition translation armlet includes, and a pair of (two) armlets, distribution is worn on left and right forearm, opens armlet power supply
After preheat 30s, while both hands do sign language action, and armlet identifies sign language after execution, and sign language is connected into Chinese includes shielding
On curtain, and sentence is pronounced to read by loudspeaker.
2nd, under speech recognition mode, the sentence of speaking in communication process can be identified, and by the text after identification
Word is shown on armlet screen, and armlet sends vibrations and represents that identification is completed, and aphasic or communicatee can see on screen
Sentence.
3rd, both hands arm armlet can be connected to mobile phone by wireless network focus, have corresponding software can be to arm on mobile phone
Ring is configured and software upgrading.It can also will be used for habitually sentence and be synchronized to software with action message, as conventional
Vocabulary deposit.
4th, the sentence of identification is failed for armlet, data transfer is sent to computing to cell phone software, then by network on armlet
On server, motion characteristic analysis is carried out by server, returns to most probable word result to mobile phone and armlet.
5th, armlet can also it is independent one use, by identify single armed signal to action be identified.
6th, armlet is charged by micro-usb interfaces, built-in lithium battery.Can a standby week, continuous use can
Use 10 hours.
7th, the nonrigid connector 1 of armlet is made by rubber, can be adjusted according to the forearm thickness of wearer elastic.
8th, the data of armlet collection are divided into two major classes, inertial sensor and surface myoelectric sensor:Inertial sensor can be with
The physical action of forearm macroscopic view is obtained, including:Translation, rotation;Surface myoelectric sensor gathers forearm muscle variable signal, can be with
Speculate the bending for obtaining comprehensive finger and the state upheld;With reference to two kinds of signals, the sign language of determination can be finally given.
9th, identification process gathers 10,000 by deep learning network, each sign language action by the adult of all ages and classes, sex
More than secondary, to improve discrimination.
Brief description of the drawings
Fig. 1 the utility model Sign Language Recognition translates the dimensional structure diagram of armlet;
Fig. 2 the utility model Sign Language Recognition translates the fundamental diagram of armlet.
Embodiment
A kind of Sign Language Recognition translation armlet of the present utility model is described in detail below in conjunction with the accompanying drawings:
As shown in figure 1, a kind of Sign Language Recognition translation armlet of the present utility model, including nonrigid connector 1 and multiple housings
3, multiple housings 3 are connected by nonrigid connector 1 form armlet body respectively, and described nonrigid connector 1 can carry out bullet
Property it is flexible to cause armlet body is suitable to wear in different arms;
Surface myoelectric sensor 2 is correspondingly arranged on the multiple housing 3 respectively, sets inertia to pass at least one housing 3
Sensor, processor is set at least one housing, each surface myoelectric sensor, inertial sensor are electrically connected with processor respectively
Connect;
It is provided for showing the display module 5 of sign language interpreter information on the outer wall of at least one housing 3.
In the expression of sign language, forearm muscle will shrink and diastole.Therefore, can be by detecting the movable information of muscle
So as to obtain the state of human arm and finger motion, and muscular movement can produce electromyographic signal, therefore can be small by measuring
Arm EMG signal obtains the feature of arm and finger motion.Electromyographic signal has two kinds of acquisition modes, pin electrode electromyographic signal and table
Facial muscle electric signal, surface electromyogram signal (sEMG) have the characteristics of non-invasive and measurement facilitates, are widely used.
The Sign Language Recognition translation armlet of the present embodiment gathers aphasic by surface myoelectric sensor 2 and is carrying out sign language exchange
When muscular movement information.The matrix of the electrode plates of the present embodiment surface myoelectric sensor 2 is made with copper, and electroplate is double
Polar form, a contrast reference electrode, i.e. indifferent electrode among two electrodes be present, for reducing noise, improve to common mode
The rejection ability of signal.Noise on electric wire is reduced by differential amplification method.
EMG signal is detected by two copper silver-coated electrodes, and two input signals make the difference, and identical " common mode " component can be gone
Remove, only different " differential mode " components is exaggerated.Any noise away from test point will be shown as " common mode " signal in test point;
And the signal for detecting near surface shows as difference, will be exaggerated.So the power line noise of relative distance farther out will be eliminated
Totally, and the EMG signal of relative close will be exaggerated.Its accuracy is weighed by common-mode rejection ratio (CMRR).
When EMG information is organized to propagate in (volume conductor) in vivo, increase with transmission range and decay rapidly.So electrode
The most strong muscle bellies of EMU should be placed on, to reduce the EMG of proximity of muscle interference (crosstalk).Choosing is improved using less electrode
Flexibility is selected, but brings larger skin contact resistance.Sensor is INA128 chips, for measuring telecommunications caused by muscle
Number.Because muscle has very big impedance, the power supply of equivalent to one, the both ends internal resistance very big (several kilohms) of input in itself.This core
Piece frequency acquisition is very high, can reach 200Hz, can dynamically gather the motion conditions of human muscle in real time.
The inertial sensor of the Sign Language Recognition translation armlet of the present embodiment is nine axle inertial sensors.
The processor of the Sign Language Recognition translation armlet of the present embodiment is ARM process chips, can complete simple multichannel letter
Number filter function, signal acquisition amplification after be used for remove random noise.Simultaneously muscle is produced during arm motion to tremble
It is dynamic, cause the high dither of signal, it is also desirable to eliminate by wave filter.
Further, Cao Zuoanjian 4 is set positioned at the bottom of display module 5 on the housing 3.
Preferably, the housing 3 described in the present embodiment includes at least eight, and described nonrigid connector 1 is rubber strip, and every two
Pass through at least one rubber band connection between individual housing 3.The nonrigid connector 1 of the Sign Language Recognition translation armlet of the present embodiment is rubber
Adhesive tape, it can be adjusted according to the forearm thickness of wearer elastic.
Further, the Sign Language Recognition translation armlet of the present embodiment includes differential-mode amplifier, described surface myoelectric sensing
The signal output part access differential-mode amplifier of device carries out signal amplification.
Further, the Sign Language Recognition translation armlet of the present embodiment includes A/D converter, the input of A/D converter and
The electric signal that surface myoelectric sensor gathers is converted to digital letter by the signal output part connection of differential-mode amplifier, A/D converter
Number;The output end connection processor of the A/D converter.
Further, the signal output part connection A/ of the inertial sensor described in the Sign Language Recognition translation armlet of the present embodiment
The electric signal that inertial sensor gathers is converted to data signal by the input of D converters, A/D converter.
Further, the Sign Language Recognition translation armlet of the present embodiment includes data communication module, described processor and intelligence
Energy terminal device carries out communication connection, the letter that processor gathers surface myoelectric sensor, inertial sensor by communication module
Number intelligent terminal is transferred to by communication signal, intelligent terminal is using convolutional neural networks technology as algorithm base
Plinth, it is aided with noise reduction process algorithm and feature extraction algorithm, carries out sign language interpreter, and the word after translation is returned by communication module
Processor is transmitted to the real-time display on display module.
Further, the Sign Language Recognition translation armlet of the present embodiment includes battery and the charging inlet electrically connected with battery.
Further, the Sign Language Recognition translation armlet of the present embodiment includes shock module, described shock module and processing
Device electrically connects, and processor controls shock module vibration reminding user to translate completion after receiving sign language interpreter word.
Further, the Sign Language Recognition translation armlet of the present embodiment includes voice conversion module and voice playing module, institute
The voice conversion module stated, voice playing module are electrically connected with processor respectively, and processor is sent out after receiving sign language interpreter word
Voice conversion module is given, sign language interpreter text conversion is voice and broadcast by voice playing module by voice conversion module
Put.
The Sign Language Recognition armlet of the present embodiment has the characteristics of following:
1st, Sign Language Recognition translation armlet includes, and a pair of (two) armlets, distribution is worn on left and right forearm, opens armlet power supply
After preheat 30s, while both hands do sign language action, and armlet identifies sign language after execution, and sign language is connected into Chinese includes shielding
On curtain, and sentence is pronounced to read by loudspeaker.
2nd, under speech recognition mode, the sentence of speaking in communication process can be identified, and by the text after identification
Word is shown on armlet screen, and armlet sends vibrations and represents that identification is completed, and aphasic or communicatee can see on screen
Sentence.
3rd, both hands arm armlet can be connected to mobile phone by wireless network focus, have corresponding software can be to arm on mobile phone
Ring is configured and software upgrading.It can also will be used for habitually sentence and be synchronized to software with action message, as conventional
Vocabulary deposit.
4th, the sentence of identification is failed for armlet, data transfer is sent to computing to cell phone software, then by network on armlet
On server, motion characteristic analysis is carried out by server, returns to most probable word result to mobile phone and armlet.
5th, armlet can also it is independent one use, by identify single armed signal to action be identified.
6th, armlet is charged by micro-usb interfaces, built-in lithium battery.Can a standby week, continuous use can
Use 10 hours.
7th, the nonrigid connector of armlet is made by rubber, can be adjusted according to the forearm thickness of wearer elastic.
8th, the data of armlet collection are divided into two major classes, inertial sensor and surface myoelectric sensor:Inertial sensor can be with
The physical action of forearm macroscopic view is obtained, including:Translation, rotation;Surface myoelectric sensor gathers forearm muscle variable signal, can be with
Speculate the bending for obtaining comprehensive finger and the state upheld;With reference to two kinds of signals, the sign language of determination can be finally given.
9th, identification process gathers 10,000 by deep learning network, each sign language action by the adult of all ages and classes, sex
More than secondary, to improve discrimination.
Described above is only preferred embodiment of the present utility model, not the utility model is made any formal
Limitation, although the utility model is disclosed above with preferred embodiment, but is not limited to the utility model, any to be familiar with
The technical staff of this patent is not being departed from the range of technical solutions of the utility model, when the technology contents work using above-mentioned prompting
Go out a little change or be modified to the equivalent embodiment of equivalent variations, as long as being the content without departing from technical solutions of the utility model,
Any simple modification, equivalent change and modification made according to the technical essence of the utility model to above example, still belongs to
In the range of the utility model.
Claims (10)
1. a kind of Sign Language Recognition translates armlet, it is characterised in that including nonrigid connector and multiple housings, multiple housings lead to respectively
Cross nonrigid connector and connect composition armlet body, described nonrigid connector can carry out elastic telescopic to cause armlet body
It is suitable to wear in different arms;
Surface myoelectric sensor is correspondingly arranged on the multiple housing respectively, inertial sensor is set at least one housing, extremely
Processor is set on a few housing, each surface myoelectric sensor, inertial sensor electrically connect with processor respectively;
It is provided for showing the display module of sign language interpreter information on the outer wall of at least one housing.
2. a kind of Sign Language Recognition translation armlet according to claim 1, it is characterised in that described including differential-mode amplifier
Surface myoelectric sensor signal output part access differential-mode amplifier carry out signal amplification.
3. a kind of Sign Language Recognition translation armlet according to claim 2, it is characterised in that including A/D converter, A/D turns
The input of parallel operation and the signal output part of differential-mode amplifier connect, the telecommunications that A/D converter gathers surface myoelectric sensor
Number be converted to data signal;The output end connection processor of the A/D converter.
A kind of 4. Sign Language Recognition translation armlet according to claim 3, it is characterised in that the letter of described inertial sensor
The electric signal that inertial sensor gathers is converted to digital letter by the input of number output end connection A/D converter, A/D converter
Number.
5. a kind of Sign Language Recognition translation armlet according to claim 1-4 any one, it is characterised in that lead to including data
Module is interrogated, described processor carries out communication by communication module with intelligent terminal and is connected, and processor passes surface myoelectric
Sensor, the signal of inertial sensor collection are transferred to intelligent terminal by communication signal, and intelligent terminal carries out sign language
Translation, and the word after translation is returned into processor with the real-time display on display module by communication module.
A kind of 6. Sign Language Recognition translation armlet according to claim 1-4 any one, it is characterised in that including battery and
The charging inlet electrically connected with battery.
7. a kind of Sign Language Recognition translation armlet according to claim 1-4 any one, it is characterised in that on the housing
Positioned at display module bottom, Cao Zuoanjian is set.
8. a kind of Sign Language Recognition translation armlet according to claim 1-4 any one, it is characterised in that including shaking mould
Block, described shock module electrically connect with processor, and processor controls shock module vibrations to carry after receiving sign language interpreter word
The user that wakes up, which translates, to complete.
9. a kind of Sign Language Recognition translation armlet according to claim 1-4 any one, it is characterised in that turn including voice
Mold changing block and voice playing module, described voice conversion module, voice playing module electrically connect with processor respectively, processor
Receive and voice conversion module be sent to after sign language interpreter word, voice conversion module by sign language interpreter text conversion be voice simultaneously
Played out by voice playing module.
A kind of 10. Sign Language Recognition translation armlet according to claim 1-4 any one, it is characterised in that described shell
Body includes at least eight, and described nonrigid connector is rubber strip, passes through at least one rubber band connection between each two housing.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201720816796.7U CN207037600U (en) | 2017-07-06 | 2017-07-06 | A kind of Sign Language Recognition translates armlet |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201720816796.7U CN207037600U (en) | 2017-07-06 | 2017-07-06 | A kind of Sign Language Recognition translates armlet |
Publications (1)
Publication Number | Publication Date |
---|---|
CN207037600U true CN207037600U (en) | 2018-02-23 |
Family
ID=61465989
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201720816796.7U Active CN207037600U (en) | 2017-07-06 | 2017-07-06 | A kind of Sign Language Recognition translates armlet |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN207037600U (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109473016A (en) * | 2019-01-15 | 2019-03-15 | 刘净植 | It is a kind of novel to listen barrier, deaf and dumb crowd's auxiliary security and information interchange device and method |
CN109508088A (en) * | 2018-10-23 | 2019-03-22 | 诺百爱(杭州)科技有限责任公司 | One kind is based on electromyography signal Sign Language Recognition translation armlet and sign Language Recognition Method |
WO2021057248A1 (en) * | 2019-09-24 | 2021-04-01 | 北京海益同展信息科技有限公司 | Armband structure |
-
2017
- 2017-07-06 CN CN201720816796.7U patent/CN207037600U/en active Active
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109508088A (en) * | 2018-10-23 | 2019-03-22 | 诺百爱(杭州)科技有限责任公司 | One kind is based on electromyography signal Sign Language Recognition translation armlet and sign Language Recognition Method |
CN109473016A (en) * | 2019-01-15 | 2019-03-15 | 刘净植 | It is a kind of novel to listen barrier, deaf and dumb crowd's auxiliary security and information interchange device and method |
WO2021057248A1 (en) * | 2019-09-24 | 2021-04-01 | 北京海益同展信息科技有限公司 | Armband structure |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN208799217U (en) | A kind of portable cardiac detection device and cardioelectric monitor system based on electronics epidermis | |
CN109743656B (en) | Intelligent motion earphone based on electroencephalogram idea and implementation method and system thereof | |
CN207037600U (en) | A kind of Sign Language Recognition translates armlet | |
CN207008541U (en) | A kind of control armlet of the VR/AR interactions helmet | |
CN205158728U (en) | Sign language translating system | |
CN108968952A (en) | A kind of brain myoelectricity and Inertia information synchronous acquisition device | |
CN106236079A (en) | Electric and the sleep monitor eyeshield of eye electricity compound detection and sleep monitor method for brain | |
CN202223388U (en) | Wearable brain-controlled intelligent prosthesis | |
WO2019228174A1 (en) | Facial expression recognition-based smart vr eyewear | |
CN104352248A (en) | Heart sound, lung sound and fetal heart sound acquisition and health management system and method | |
CN105832073A (en) | Intelligent interactive emotional care bolster robot system | |
CN205849452U (en) | A kind of checking with EMG method clothes | |
CN105816168A (en) | Electrocardiogram detection device | |
CN103750836A (en) | Wearable myoelectricity instrument | |
CN103169471A (en) | Portable electroencephalogram detection system | |
CN105943025A (en) | Earphone type electrocardiogram monitor | |
CN203619544U (en) | SMD sensor electrode and physiological parameter detecting system | |
CN113871028A (en) | Interactive rehabilitation system based on myoelectric intelligent wearing | |
CN206239412U (en) | A kind of wearable sleep monitor eyeshade for the electric compound detection of brain electricity and eye | |
CN205072865U (en) | Wearing formula electroencephalogram acquired ware | |
CN206710887U (en) | A kind of idea dolly | |
CN205692125U (en) | Gesture discriminating device | |
CN204581273U (en) | A kind of hand held ECG Signal Sampling System for identity validation | |
CN107802263A (en) | A kind of wearable sound feedback system based on Electroencephalo signal | |
CN207590676U (en) | A kind of portable cardiac detection device based on electronics epidermis |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
GR01 | Patent grant | ||
GR01 | Patent grant |