CN109325411A - Dynamic sign Language Recognition Method based on Kinect - Google Patents
Dynamic sign Language Recognition Method based on Kinect Download PDFInfo
- Publication number
- CN109325411A CN109325411A CN201810938994.XA CN201810938994A CN109325411A CN 109325411 A CN109325411 A CN 109325411A CN 201810938994 A CN201810938994 A CN 201810938994A CN 109325411 A CN109325411 A CN 109325411A
- Authority
- CN
- China
- Prior art keywords
- kinect
- sign language
- language recognition
- dynamic
- recognition method
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
- G06V40/28—Recognition of hand or arm movements, e.g. recognition of deaf sign language
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Biology (AREA)
- General Health & Medical Sciences (AREA)
- Psychiatry (AREA)
- Social Psychology (AREA)
- Human Computer Interaction (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
The dynamic sign Language Recognition Method based on Kinect that the invention discloses a kind of, this method comprises: step 1, obtains human skeleton data using Kinect and relevant depth information obtains hand position；Step 2, hand is detected and is tracked using Kinect；Step 3, hand-characteristic is extracted, and is pre-processed；Step 4, gesture feature vector parameter is extracted, and carries out the dynamic Sign Language Recognition based on HMM and SVM.The dynamic sign Language Recognition Method based on Kinect overcomes that resolution in the prior art is low, problem complicated for operation, realizes resolution height, effect easy to operate.
Description
Technical field
The present invention relates to the application processing of Kinect, and in particular, to the dynamic sign Language Recognition Method based on Kinect.
Background technique
With the fast development of computer technology in recent years, human-computer interaction becomes the important of computer technology research field
Component part.This technology mainly includes the identification of movement to the mankind, language, the identification of expression and sign language.Wherein sign language is known
The a part not understood as Body Languages, there is an extremely important effect, on the one hand it be between deaf-mute and deaf-mute and
Main exchange way between normal person, the important channel of another aspect human-computer interaction.
Traditional sign Language Recognition Method mainly has the methods of template matching, neural network, HMM, but these methods all exist
Certain deficiency needs to obtain the observation number in sampling feature vectors such as in training HMM model, and for different
Its depth image of sign language changes greatly, it is not easy to determine observation number.And it is frequent for the method for traditional Sign Language Recognition
It is single use, this makes the precision of Sign Language Recognition and user experience not high.
Have at present by using data glove and obtain the feature and deliberate action of human hands, although by this method
Available higher recognition accuracy, but data glove has using complicated and expensive deficiency for now
Point.
Summary of the invention
The dynamic sign Language Recognition Method based on Kinect that the object of the present invention is to provide a kind of, should the dynamic based on Kinect
Sign Language Recognition Method overcomes that resolution in the prior art is low, problem complicated for operation, realizes resolution height, easy to operate
Effect.
To achieve the goals above, the present invention provides a kind of the dynamic sign Language Recognition Method based on Kinect, this method
Include:
Step 1, human skeleton data are obtained using Kinect and relevant depth information obtains hand position；
Step 2, hand is detected and is tracked using Kinect；
Step 3, hand-characteristic is extracted, and is pre-processed；
Step 4, gesture feature vector parameter is extracted, and carries out the dynamic Sign Language Recognition based on HMM and SVM.
Preferably, this method further include:
The dynamic trajectory of human hands is handled, and the action sequence of human hands is split；
In the unsuccessful situation of dynamic Sign Language Recognition based on HMM and SVM, optimal possible a variety of sign languages are obtained, benefit
Five kinds of sign languages of action sequence and this of human hands are compared and analyzed with Pearson came algorithm, identify sign language.
Preferably, the method handled the dynamic trajectory of human hands includes:
Track normalized and based on Kalman filtering track correction.
Preferably, the method dynamic trajectory of human hands handled further include:
Action sequence is split, by the same size of the action sequence of the length scale of action sequence to template.
Preferably, the quantity of a variety of sign languages is 5 kinds.
According to the above technical scheme, the more traditional method of dynamic sign Language Recognition Method of the invention have equipment cost compared with
Low, easy to operate, accuracy of identification is higher, the preferable feature of user experience.More traditional dynamic sign Language Recognition Method, the present invention
Using the method for HMM model and SVM models coupling, the accuracy of identification is improved；Equipment used in the present invention is
Kinect has the characteristics that cost is relatively low and easy to operate compared to other relevant devices.
Other features and advantages of the present invention will the following detailed description will be given in the detailed implementation section.
Detailed description of the invention
The drawings are intended to provide a further understanding of the invention, and constitutes part of specification, with following tool
Body embodiment is used to explain the present invention together, but is not construed as limiting the invention.In the accompanying drawings:
Fig. 1 is the flow chart for illustrating a kind of dynamic sign Language Recognition Method based on Kinect of the invention.
Specific embodiment
Below in conjunction with attached drawing, detailed description of the preferred embodiments.It should be understood that this place is retouched
The specific embodiment stated is merely to illustrate and explain the present invention, and is not intended to restrict the invention.
The present invention provides a kind of dynamic sign Language Recognition Method based on Kinect, this method comprises:
Step 1, human skeleton data are obtained using Kinect and relevant depth information obtains hand position；
Step 2, hand is detected and is tracked using Kinect；
Step 3, hand-characteristic is extracted, and is pre-processed；
Step 4, gesture feature vector parameter is extracted, and carries out the dynamic Sign Language Recognition based on HMM and SVM.
According to the above technical scheme, the more traditional method of dynamic sign Language Recognition Method of the invention have equipment cost compared with
Low, easy to operate, accuracy of identification is higher, the preferable feature of user experience.More traditional dynamic sign Language Recognition Method, the present invention
Using the method for HMM model and SVM models coupling, the accuracy of identification is improved；Equipment used in the present invention is
Kinect has the characteristics that cost is relatively low and easy to operate compared to other relevant devices.
In the preferred embodiment of the present invention, this method can also include:
The dynamic trajectory of human hands is handled, and the action sequence of human hands is split；
In the unsuccessful situation of dynamic Sign Language Recognition based on HMM and SVM, optimal possible a variety of sign languages are obtained, benefit
Five kinds of sign languages of action sequence and this of human hands are compared and analyzed with Pearson came algorithm, identify sign language.
The action sequence for decomposing user is additionally used in invention and leads to the action sequence of the action sequence of user and template
The method for crossing the comparison of Pearson came algorithm, to further improve accuracy and user experience.
In the preferred embodiment of the present invention, the method handled the dynamic trajectory of human hands can wrap
It includes:
Track normalized and based on Kalman filtering track correction.
In the present invention, the mode of Kalman filtering has been used to correct the Kinect data acquired, further
Improve data reliability, to improve the accuracy of dynamic Sign Language Recognition.
In the preferred embodiment of the present invention, the method handled to the dynamic trajectory of human hands can be with
Include:
Action sequence is split, by the same size of the action sequence of the length scale of action sequence to template.
In the preferred embodiment of the present invention, the quantity of a variety of sign languages is 5 kinds.
The present invention is used for human-computer interaction.For exchanging between deaf-mute and between deaf-mute and normal person.For deaf
Mute or normal person's training sign language.
It is described the prefered embodiments of the present invention in detail above in conjunction with attached drawing, still, the present invention is not limited to above-mentioned realities
The detail in mode is applied, within the scope of the technical concept of the present invention, a variety of letters can be carried out to technical solution of the present invention
Monotropic type, these simple variants all belong to the scope of protection of the present invention.
It is further to note that specific technical features described in the above specific embodiments, in not lance
In the case where shield, can be combined in any appropriate way, in order to avoid unnecessary repetition, the present invention to it is various can
No further explanation will be given for the combination of energy.
In addition, various embodiments of the present invention can be combined randomly, as long as it is without prejudice to originally
The thought of invention, it should also be regarded as the disclosure of the present invention.
Claims (5)
1. a kind of dynamic sign Language Recognition Method based on Kinect, which is characterized in that this method comprises:
Step 1, human skeleton data are obtained using Kinect and relevant depth information obtains hand position；
Step 2, hand is detected and is tracked using Kinect；
Step 3, hand-characteristic is extracted, and is pre-processed；
Step 4, gesture feature vector parameter is extracted, and carries out the dynamic Sign Language Recognition based on HMM and SVM.
2. the dynamic sign Language Recognition Method according to claim 1 based on Kinect, which is characterized in that this method is also wrapped
It includes:
The dynamic trajectory of human hands is handled, and the action sequence of human hands is split；
In the unsuccessful situation of dynamic Sign Language Recognition based on HMM and SVM, optimal possible a variety of sign languages are obtained, skin is utilized
You compare and analyze five kinds of sign languages of action sequence and this of human hands inferior algorithm, identify sign language.
3. the dynamic sign Language Recognition Method according to claim 2 based on Kinect, which is characterized in that human hands
The method that dynamic trajectory is handled includes:
Track normalized and based on Kalman filtering track correction.
4. the dynamic sign Language Recognition Method according to claim 2 based on Kinect, which is characterized in that human hands
The method that dynamic trajectory is handled further include:
Action sequence is split, by the same size of the action sequence of the length scale of action sequence to template.
5. the dynamic sign Language Recognition Method according to claim 2 based on Kinect, which is characterized in that a variety of sign languages
Quantity be 5 kinds.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810938994.XA CN109325411A (en) | 2018-08-17 | 2018-08-17 | Dynamic sign Language Recognition Method based on Kinect |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810938994.XA CN109325411A (en) | 2018-08-17 | 2018-08-17 | Dynamic sign Language Recognition Method based on Kinect |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109325411A true CN109325411A (en) | 2019-02-12 |
Family
ID=65264307
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810938994.XA Pending CN109325411A (en) | 2018-08-17 | 2018-08-17 | Dynamic sign Language Recognition Method based on Kinect |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109325411A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110096987A (en) * | 2019-04-24 | 2019-08-06 | 东北大学 | A kind of sign language action identification method based on two-way 3DCNN model |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105807926A (en) * | 2016-03-08 | 2016-07-27 | 中山大学 | Unmanned aerial vehicle man-machine interaction method based on three-dimensional continuous gesture recognition |
| CN107092349A (en) * | 2017-03-20 | 2017-08-25 | 重庆邮电大学 | A kind of sign Language Recognition and method based on RealSense |
| CN107358171A (en) * | 2017-06-22 | 2017-11-17 | 华中师范大学 | A kind of gesture identification method based on COS distance and dynamic time warping |
-
2018
- 2018-08-17 CN CN201810938994.XA patent/CN109325411A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105807926A (en) * | 2016-03-08 | 2016-07-27 | 中山大学 | Unmanned aerial vehicle man-machine interaction method based on three-dimensional continuous gesture recognition |
| CN107092349A (en) * | 2017-03-20 | 2017-08-25 | 重庆邮电大学 | A kind of sign Language Recognition and method based on RealSense |
| CN107358171A (en) * | 2017-06-22 | 2017-11-17 | 华中师范大学 | A kind of gesture identification method based on COS distance and dynamic time warping |
Non-Patent Citations (2)
| Title |
|---|
| JIANG, HANNI 等: "Adaptive key frame extraction from RGB-D for hand gesture recognition", 《PROCEEDINGS OF SPIE》 * |
| 郭晓利 等: "基于Kinect深度信息的动态手势识别", 《东北电力大学学报》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110096987A (en) * | 2019-04-24 | 2019-08-06 | 东北大学 | A kind of sign language action identification method based on two-way 3DCNN model |
| CN110096987B (en) * | 2019-04-24 | 2022-04-12 | 东北大学 | Dual-path 3DCNN model-based mute action recognition method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10860839B2 (en) | Method and apparatus for generating training data for human face recognition, device and computer storage medium | |
| US10109219B2 (en) | System and method for automated sign language recognition | |
| Yang et al. | Robust sign language recognition by combining manual and non-manual features based on conditional random field and support vector machine | |
| US20160042228A1 (en) | Systems and methods for recognition and translation of gestures | |
| CN111027403B (en) | Gesture estimation method, device, equipment and computer readable storage medium | |
| Elpeltagy et al. | Multi‐modality‐based Arabic sign language recognition | |
| Ding et al. | HMM with improved feature extraction-based feature parameters for identity recognition of gesture command operators by using a sensed Kinect-data stream | |
| CN109543644B (en) | Multi-modal gesture recognition method | |
| CN107993651B (en) | Voice recognition method and device, electronic equipment and storage medium | |
| Ding et al. | Feature design scheme for Kinect-based DTW human gesture recognition | |
| CN112198966B (en) | Stroke identification method and system based on FMCW radar system | |
| Miah et al. | Rotation, Translation and Scale Invariant Sign Word Recognition Using Deep Learning. | |
| Khurana et al. | Static hand gestures recognition system using shape based features | |
| CN107909003B (en) | gesture recognition method for large vocabulary | |
| CN103489000A (en) | Achieving method of human movement recognition training system | |
| CN103714331A (en) | Facial expression feature extraction method based on point distribution model | |
| Sharma et al. | Trbaggboost: An ensemble-based transfer learning method applied to Indian Sign Language recognition | |
| Geetha et al. | Dynamic gesture recognition of Indian sign language considering local motion of hand using spatial location of Key Maximum Curvature Points | |
| Zhang et al. | Multimodal fusion framework based on statistical attention and contrastive attention for sign language recognition | |
| Farooq et al. | A comparison of hardware based approaches for sign language gesture recognition systems | |
| CN109325411A (en) | Dynamic sign Language Recognition Method based on Kinect | |
| CN112115715A (en) | Natural language text processing method and device, storage medium and electronic equipment | |
| Fraiwan et al. | A Kinect-based system for Arabic sign language to speech translation | |
| Oliva et al. | Filipino sign language recognition for beginners using kinect | |
| Dignan et al. | An AI-based approach for improved sign language recognition using multiple videos |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190212 |
|
| RJ01 | Rejection of invention patent application after publication |