CN106934351A - Gesture identification method, device and electronic equipment - Google Patents
Gesture identification method, device and electronic equipment Download PDFInfo
- Publication number
- CN106934351A CN106934351A CN201710105905.9A CN201710105905A CN106934351A CN 106934351 A CN106934351 A CN 106934351A CN 201710105905 A CN201710105905 A CN 201710105905A CN 106934351 A CN106934351 A CN 106934351A
- Authority
- CN
- China
- Prior art keywords
- gesture
- distortion
- panoramic picture
- correction
- panorama
- 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.)
- Granted
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
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/107—Static hand or arm
- G06V40/113—Recognition of static hand signs
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/24—Aligning, centring, orientation detection or correction of the image
- G06V10/247—Aligning, centring, orientation detection or correction of the image by affine transforms, e.g. correction due to perspective effects; Quadrilaterals, e.g. trapezoids
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Multimedia (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Human Computer Interaction (AREA)
- Image Processing (AREA)
- Studio Devices (AREA)
Abstract
The embodiment of the invention discloses a kind of gesture identification method, device and electronic installation, it is related to technical field of data processing, can solve the problem that the inefficient problem of gesture identification in the prior art.The gesture identification method of the embodiment of the present invention includes：Obtain the panoramic picture comprising gesture operation that single camera shoots from panorama component；Using deformity correction algorithm corresponding with the panorama component, distortion correction operation is carried out to the panoramic picture, obtain correction chart picture；Piecemeal operation is performed to the correction chart picture, block image is obtained；The gesture feature value of the block image is extracted, and the order of the gesture operation is determined based on the gesture feature value.Additionally, the embodiment of the invention also discloses gesture identifying device and electronic equipment.By the scheme of the embodiment of the present invention, the efficiency of gesture identification can be effectively improved.
Description
Technical field
The present invention relates to technical field of data processing, more particularly to the data processing based on gesture identification.
Background technology
With the development of computer and hardware technology, the operational capability of computer is more and more stronger, and hardware cost is also increasingly
Low, Intelligent hardware is gradually applied in the production and living of people, and this is accomplished by user and effectively can carry out letter with Intelligent hardware
Breath interaction, transmits command information, and the manipulation of the gesture of view-based access control model is undoubtedly a good selection.
Gesture identification is produced based on computer vision and image recognition technology.Gesture identification is first-class usually using shooting
Image capture device, the image to gesture is acquired, and calibration matching, modeling is carried out by algorithm process, so as to generate correlation
Two-dimentional or three-dimensional hand information, then the information such as position, attitudes vibration by hand-characteristic point carries out the calculating of hand exercise,
Hand coordinate and vector are obtained, and then gesture is tracked.
In computer vision field, the camera of traditional gesture identification mostly uses common plane camera.These camera institutes
The scene of shooting, can only be unidirectional, have the shortcomings that visual angle is small, there is dead angle.If with this camera do gesture identification and
Control, at the back of Intelligent hardware camera, just has very big vision dead zone.In this region, the gesture of people cannot be by intelligence
Energy hardware is captured, so as to gesture and control Intelligent hardware cannot be recognized effectively.
In view of this, gesture identification is carried out using 360 degree of panorama cameras, has obtained increasingly being widely applied.Using 360
After degree panorama camera, Intelligent hardware can just capture the gesture of people in the range of 360 degree, such that it is able to without dead angle ground recipient
Gesture instruction.
Inventor realize it is of the invention during find, using multiple cameras 360 degree panorama cameras collection panorama
Photo can cause intelligent hardware devices complex structure, equipment cost to rise；Gathered using 360 degree of panorama cameras of single camera
To panoramic view usually there will be certain radial distortion and distortion, with distortion and distortion panoramic view in, originally
Gesture Recognition Algorithm can fail.In addition, the size of panoramic picture is generally large, and smart machine processes a frame panoramic picture
The time for generally needing consumption more, this have impact on recognition speed of the smart machine for gesture operation to a certain extent.
The content of the invention
In view of this, a kind of gesture identification method, device and electronic equipment are the embodiment of the invention provides, it is at least part of
Solve problems of the prior art.
In a first aspect, a kind of gesture identification method is the embodiment of the invention provides, including：
Obtain the panoramic picture comprising gesture operation that single camera shoots from panorama component；
Using deformity correction algorithm corresponding with the panorama component, distortion correction operation is carried out to the panoramic picture, obtained
Correction chart picture；
Piecemeal operation is performed to the correction chart picture, block image is obtained；
The gesture feature value of the block image is extracted, and the order of the gesture operation is determined based on the gesture feature value.
As a kind of specific implementation of the embodiment of the present invention, calculated using deformity correction corresponding with the panorama component
Method, distortion correction operation is carried out to the panoramic picture, including：
Obtain calibration model corresponding with the panorama component；
Parameter predigesting operation is carried out to the calibration model, calibration model is simplified；
Based on the simplification calibration model, deformity correction algorithm is formed, and based on the deformity correction algorithm to the panoramic picture
Carry out distortion correction operation.
As a kind of specific implementation of the embodiment of the present invention, the panoramic picture is carried out based on the deformity correction algorithm
Distortion correction is operated, including：
The center of distortion point of the panoramic picture is searched, the center of distortion point is set to the two-dimensional coordinate system of the panoramic picture
Origin；
Determine lateral coordinates x and longitudinal coordinate y that the pixel p on the panoramic picture is fastened in the two-dimensional coordinate；
Based on the deformity correction algorithm, coordinate transform is carried out to lateral coordinates x and longitudinal coordinate y, obtain new horizontal stroke
To coordinate X and new longitudinal coordinate Y；
Coordinates of the pixel p on the panoramic picture is transformed to (X, Y) by (x, y).
As a kind of specific implementation of the embodiment of the present invention, based on the deformity correction algorithm, to lateral coordinates x and
Longitudinal coordinate y carries out coordinate transform, obtains new lateral coordinates X and new longitudinal coordinate Y, including：
Obtain the radial distance r between pixel p and the two-dimensional coordinate system origin；
Obtain the first correction coefficient k of the simplification calibration model1And the second correction coefficient k2；
Based on radial distance r, first correction coefficient k1And second correction coefficient k2, dot product is performed to coordinate (x, y)
Computing, obtains the new coordinate (X, Y) of pixel p.
As a kind of specific implementation of the embodiment of the present invention, the gesture feature value of the block image is extracted, and be based on
The gesture feature value determines the order of the gesture operation, including：
The gesture feature value is carried out into similarity-rough set with the parameter in default gesture identification model, gesture similarity is obtained
Value；And
Based on the gesture Similarity value, the order of the gesture operation is determined.
As a kind of specific implementation of the embodiment of the present invention, based on the gesture Similarity value, the gesture operation is determined
Order, including：
When the gesture Similarity value presets Similarity value less than first and the block image is more than pre-set dimension, continue
Cutting operation is performed to the block image.
As a kind of specific implementation of the embodiment of the present invention, based on the gesture Similarity value, the gesture operation is determined
Order, including：
When the gesture Similarity value is more than the second default Similarity value, will gesture command corresponding with the gesture feature value
It is defined as the order of the gesture operation.
As a kind of specific implementation of the embodiment of the present invention, based on the gesture Similarity value, the gesture operation is determined
Order, including：
When the closed-loop interval that the gesture Similarity value is constituted between the first default Similarity value and the second default Similarity value
When, the gesture operation is defined as order operation undetermined；And
Storage is undetermined to the order to operate related gesture feature value.
Used as a kind of specific implementation of the embodiment of the present invention, method also includes：
Statistics is undetermined to the order to operate related gesture feature value；
The specific instruction of order operation undetermined is determined based on the gesture feature value；And
The gesture identification parameter related to the specific instruction is updated in the default gesture identification model.
Second aspect, the embodiment of the present invention additionally provides a kind of gesture identifying device, including：
Acquisition module, for obtaining the panoramic picture comprising gesture operation that single camera shoots from panorama component；
Correction module, for utilizing deformity correction algorithm corresponding with the panorama component, line distortion is entered to the panoramic picture
Correct operation, obtains correction chart picture；
Piecemeal module, for performing piecemeal operation to the correction chart picture, obtains block image；
Determining module, the gesture feature value for extracting the block image, and the gesture is determined based on the gesture feature value
The order of operation.
Used as a kind of specific implementation of the embodiment of the present invention, correction module is additionally operable to：
Obtain calibration model corresponding with the panorama component；
Parameter predigesting operation is carried out to the calibration model, calibration model is simplified；
Based on the simplification calibration model, deformity correction algorithm is formed, and based on the deformity correction algorithm to the panoramic picture
Carry out distortion correction operation.
Used as a kind of specific implementation of the embodiment of the present invention, the correction module is additionally operable to：
The center of distortion point of the panoramic picture is searched, the center of distortion point is set to the two-dimensional coordinate system of the panoramic picture
Origin；
Determine lateral coordinates x and longitudinal coordinate y that the pixel p on the panoramic picture is fastened in the two-dimensional coordinate；
Based on the deformity correction algorithm, coordinate transform is carried out to lateral coordinates x and longitudinal coordinate y, obtain new horizontal stroke
To coordinate X and new longitudinal coordinate Y；
Coordinates of the pixel p on the panoramic picture is transformed to (X, Y) by (x, y).
Used as a kind of specific implementation of the embodiment of the present invention, correction module is additionally operable to：
Obtain the radial distance r between pixel p and the two-dimensional coordinate system origin；
Obtain the first correction coefficient k of the simplification calibration model1And the second correction coefficient k2；
Based on radial distance r, first correction coefficient k1And second correction coefficient k2, dot product is performed to coordinate (x, y)
Computing, obtains the new coordinate (X, Y) of pixel p.
Used as a kind of specific implementation of the embodiment of the present invention, determining module is additionally operable to：
The gesture feature value is carried out into similarity-rough set with the parameter in default gesture identification model, gesture similarity is obtained
Value；And
Based on the gesture Similarity value, the order of the gesture operation is determined.
Used as a kind of specific implementation of the embodiment of the present invention, determining module is additionally operable to：
When the gesture Similarity value presets Similarity value less than first and the block image is more than pre-set dimension, continue
Cutting operation is performed to the block image.
Used as a kind of specific implementation of the embodiment of the present invention, determining module is additionally operable to：
When the gesture Similarity value is more than the second default Similarity value, will gesture command corresponding with the gesture feature value
It is defined as the order of the gesture operation.
Used as a kind of specific implementation of the embodiment of the present invention, determining module is additionally operable to：
When the closed-loop interval that the gesture Similarity value is constituted between the first default Similarity value and the second default Similarity value
When, the gesture operation is defined as order operation undetermined；And
Storage is undetermined to the order to operate related gesture feature value.
Used as a kind of specific implementation of the embodiment of the present invention, determining module is additionally operable to：
Statistics is undetermined to the order to operate related gesture feature value；
The specific instruction of order operation undetermined is determined based on the gesture feature value；And
The gesture identification parameter related to the specific instruction is updated in the default gesture identification model.
The third aspect, the embodiment of the present invention additionally provides a kind of electronic equipment, and the electronic equipment includes：
At least one processor；And,
The memory being connected with least one processor communication；Wherein,
The memory storage has can be by the instruction of at least one computing device, and the instruction is by least one processor
Perform, so that at least one processor is able to carry out the hand that any implementation of foregoing first aspect or first aspect is somebody's turn to do
Gesture recognition methods.
Fourth aspect, the embodiment of the present invention additionally provides a kind of non-transient computer readable storage medium storing program for executing, the non-transient meter
Calculation machine readable storage medium storing program for executing stores computer instruction, and the computer instruction is used to make the computer perform aforementioned first aspect or the
Gesture identification method in any implementation of one side.
5th aspect, the embodiment of the present invention additionally provides a kind of computer program product, and the computer program product includes
The calculation procedure on non-transient computer readable storage medium storing program for executing is stored, the computer program includes programmed instruction, when the program
When instruction is computer-executed, the gesture in making the computer perform any implementation of aforementioned first aspect or first aspect
Recognition methods.
Gesture identification method provided in an embodiment of the present invention, device, electronic equipment, non-transient computer readable storage medium storing program for executing
And computer program, the panoramic picture comprising gesture operation that single camera shoots from panorama component can be obtained, utilize
Deformity correction algorithm corresponding with the panorama component, distortion correction operation is carried out to the panoramic picture, correction chart picture is obtained, to this
Correction chart picture performs piecemeal operation, obtains block image, extracts the gesture feature value of the block image, and based on the gesture feature
Value determines the order of the gesture operation.So, panoramic picture is obtained using single camera, reduces the cost of equipment；Existing
Have on Gesture Recognition Algorithm, by training and the characteristic value of dynamic adjustment gesture identification so that under the conditions of radial distortion, gesture
Discrimination is improved；Piecemeal treatment is carried out by panoramic picture, each two field picture of camera shooting is further increased
Gesture identification speed.
Brief description of the drawings
Technical scheme in order to illustrate more clearly the embodiments of the present invention, below will be attached to what is used needed for embodiment
Figure is briefly described, it should be apparent that, drawings in the following description are only some embodiments of the present invention, for this area
For those of ordinary skill, on the premise of not paying creative work, other accompanying drawings can also be obtained according to these accompanying drawings.
Fig. 1 a are a kind of structural representation of panorama camera device provided in an embodiment of the present invention；
Fig. 1 b are the structural representation of another panorama camera device provided in an embodiment of the present invention
Fig. 2 is a kind of schematic flow sheet of gesture identification method provided in an embodiment of the present invention；
Fig. 3 is the schematic flow sheet of another gesture identification method provided in an embodiment of the present invention；
Fig. 4 is the schematic flow sheet of another gesture identification method provided in an embodiment of the present invention；
Fig. 5 is the schematic flow sheet of another gesture identification method provided in an embodiment of the present invention；
Fig. 6 is the schematic flow sheet of another gesture identification method provided in an embodiment of the present invention；
Fig. 7 is a kind of schematic block diagram of gesture identifying device provided in an embodiment of the present invention；
Fig. 8 is the structural representation of a kind of electronic equipment provided in an embodiment of the present invention.
Specific embodiment
The embodiment of the present invention is described in detail below in conjunction with the accompanying drawings.
It will be appreciated that described embodiment is only a part of embodiment of the invention, rather than whole embodiments.Base
Embodiment in the present invention, those of ordinary skill in the art obtained under the premise of creative work is not made it is all its
Its embodiment, belongs to the scope of protection of the invention.
Fig. 1 a and Fig. 1 b are respectively the structural representation of the panorama camera device of embodiments of the invention offer, panoramic shooting
Device can be the panorama camera or panoramic camera that can shoot 360 degree of panoramic pictures or 360 degree of panoramic videos.Panoramic shooting
Device can be independent realization on a large scale it is interior without dead angle monitoring, be capable of no visual blind spot monitoring user gesture operation, accomplish
Seamless monitoring.One panorama camera device can replace many common video cameras, accomplish seamless under conditions of low cost
Monitoring, therefore it is widely used in every field, including prison, government bodies, bank, social safety, public place, text
Change place etc..
Referring to Fig. 1 a, panorama camera device can include camera 101, reflection subassembly 102a, reflection subassembly 102b, top cover
104 and base 105, wherein reflection subassembly 102a and reflection subassembly 102b collectively constitute panorama component.Reflection subassembly 102a and
102b is made up of the material acted on light transmitting, and incident light 103a is by after emitting module 102a transmittings, forming reflected light
103b, launching light 103b take the photograph by after emitting module 102a secondary reflections, forming launching light 103c, final launching light 103c entrance
In first 101, panoramic picture is formed.
In addition to the implementation method that Fig. 1 a are provided, Fig. 1 b additionally provide another panorama camera device 10, the panorama
Camera head 10 includes：Camera 101, refracted component 106 and base 105.Refracted component 106 constitutes panorama camera device
Panorama component.Incident light 103a takes the photograph by after the refraction of refracted component 106, forming refraction light 103d, the 103d entrance of refraction light
In first 101, panoramic picture is formed.
In addition to the panorama camera device 10 that Fig. 1 a and Fig. 1 b are provided, panorama camera device 10 can also be other types
Use single camera obtain panoramic picture camera.
Panorama camera device in the present embodiment, the panorama group valency used cooperatively using single camera and camera, can
Panoramic picture is obtained in the case where cost is effectively reduced.It is and many simultaneously as using single camera capturing panoramic view image
Panoramic picture is spliced into after individual shooting figure collection image to compare, reduce the complexity of image procossing.
As the concrete application of above-mentioned panorama camera device, a kind of gesture identification method is the embodiment of the invention provides, joined
See Fig. 2, the method is comprised the following steps：
S201, obtains the panoramic picture comprising gesture operation that single camera shoots from panorama component.
In gesture identification field, the camera for gesture identification mostly uses common plane camera.These cameras are clapped
The scene taken the photograph, can only be unidirectional, have the shortcomings that visual angle is small, there is dead angle.If doing gesture identification and control with this camera
System, at the back of Intelligent hardware, just has very big dead angle.In this region, the gesture of people cannot be caught by Intelligent hardware
Arrive, so that None- identified and control.After 360 degree of panorama cameras, Intelligent hardware just can capture people in the range of 360 degree
Gesture, such that it is able to without dead angle ground recipient instruction.
Specifically, after camera 101 has gathered panoramic picture, panorama camera device or communicated to connect with panorama camera device
Other electronic equipments, can obtain camera 101 collection panoramic picture.Panoramic picture can be that camera is clapped by timing
Picture frame is extracted in video that is that the mode taken the photograph is obtained, or being shot from camera to obtain.
S202, using deformity correction algorithm corresponding with the panorama component, distortion correction is carried out to the panoramic picture
Operation, obtains correction chart picture.
The collection of panoramic picture is carried out using panorama component, due to camber reflection or the presence of refraction, causes to obtain panorama
Can there is distortion in image.Accordingly, it would be desirable to the reflection or refraction situation to panorama component carry out mathematical modeling, based on the Mathematical Modeling
Panorama component is reflected or the distortion figure of refraction generation is corrected, form correction chart picture.
By taking radial distortion as an example, radial distortion be typically due between picture point and its ideal image position occur it is interior
To or export-oriented movement and the ranging offset that causes.Direction according to distortion is different, can be divided into positive radial distortion variable and bear
Radial distortion variable.Positive radial distortion amount makes pixel be moved away from the direction of image centerline, causes pincushion distortion.Phase
Instead, negative radial distortion amount makes pixel want to be moved close to the direction of image center, will so cause barrel distortion.Thus
Refine radial distortion Mathematical Modeling be：
δx=x (k1r2+k2r4+k3r6+…)
δy=y (k1r2+k2r4+k3r6+…)
Wherein, δx、δyRepresent the coordinate (x, y) of pixel p in horizontal and vertical side-play amount, k1、k2、k3It is distortion model
Coefficient.
S203, piecemeal operation is performed to the correction chart picture, obtains block image.
Because panoramic picture stock size is all than larger, if recognizing the mode of gesture according to an image overall, it will
Cause processing speed slower.For the video image of captured in real-time, such as the video image of 30FPS directly carries out image procossing,
The processing pressure of processor will be caused to become big, or even the real-time processing of video image can not be completed, lead to not complete gesture
Identification.Piecemeal is carried out to image, the polylith image after piecemeal can carry out parallel processing with GPU, can accelerate gesture identification
Speed.
Due to the influence for distorting, if the global recognition gesture in fault image, global total distortion can be than larger, other
Region can have very big interference to gesture area.If carrying out image block, every piece of distortion of image relatively can be smaller, other
Region is just small to gesture area interference, it is possible to improve the probability of gesture identification.
, it is necessary to carry out dividing processing to panoramic picture in practical operation, treatment is identified parallel.Block image it is big
Small, algorithm can be using diminishing step by step, and the mode for staggeredly positioning is carried out.For example, can be according to present image 1/2 side long and wide
Formula segmentation go down, until the image after segmentation it is long × it is wide be less than 64 × 64 pixels just stop split.When image segmentation is carried out, can
Staggeredly to position so that staff is possible in the central area of image.
S204, extracts the gesture feature value of the block image, and determine the gesture behaviour based on the gesture feature value
The order of work.
It is existing in the panoramic view with distortion and distortion because panoramic view has certain distortion and distortion
Gesture Recognition Algorithm can fail, and the speed of one two field picture for the treatment of also can be slower.
Therefore, on existing Gesture Recognition Algorithm, the characteristic value of training and dynamic adjustment gesture identification so that in radial distortion
Under the conditions of, gesture identification rate is improved, and the speed of one two field picture for the treatment of also can be quickly.
Specifically, existing gesture algorithm is identified both for the gesture of closely (＜ 2m).And control intelligence
Hardware in larger distance, it is necessary to do gesture identification, for example, interior is within 5 meters, outdoor is within 10 meters.So needing
Artificial optimization is carried out for longer-distance gesture.
Gesture identification is carried out for feature based value, it usually needs carry out model training in advance, by artificial mark and machine
The mode of device study, it is determined that the corresponding operational order of final characteristic value.
Used as an example, for the training of gesture identification in Radial Distortion Image, job step is divided into：
(1) the plane picture storehouse of collection belt gesture；
(2) gesture position in the picture is indicated with existing Gesture Recognition Algorithm；
(3) with the method for radial distortion, original image is carried out radial distortion.And calculate the hand that previous step sign is obtained
The new position of gesture；
(4) characteristic value that previous step indicates the image of gesture position is calculated with existing Gesture Recognition Algorithm；
(5) parameter of existing Gesture Recognition Algorithm is adjusted, the characteristic value for making it previous step be calculated is regarded as
Gesture.
By the method in the present embodiment, by using single camera capturing panoramic view image, the same of equipment cost is reduced
When avoid the appearance of vision dead zone；Mode based on piecemeal treatment is split to panoramic picture, reduces processor treatment
The pressure of panoramic picture；Gesture identification is carried out using the Gesture Recognition Algorithm after adjustment, the degree of accuracy of gesture identification is improve.
According to another embodiment of the present invention, referring to Fig. 3, using deformity correction algorithm corresponding with panorama component, to described
Panoramic picture carries out distortion correction operation, may include steps of：
S301, obtains calibration model corresponding with the panorama component.
By taking radial distortion as an example, radial distortion is the distortion along lens radius directional spreding, and its producing cause is that light exists
Place away from lens centre is more serious than the skew produced by paracentral place, this distortion table in common camera lens
Existing more obvious, radial distortion mainly includes two kinds of barrel distortion and pincushion distortion.
Radial distortion is symmetrical distortion, and the distortion at center is 0, produced along radial direction by optical centre (center of distortion) and by
It is cumulative big, the Mathematical Modeling that distortion model is used for description fault image and source images corresponding relation：
(X, Y) T=(x, y) T (1+k1r2+k2r4+ ...+higher order term)
r2=x2+y2
Wherein, x, y represent the relative coordinate of pixel in image to center of distortion, X, and Y represents after distortion pixel in image
The corresponding coordinates of point p, k1、k2It is the coefficient of distortion model.
S302, parameter predigesting operation is carried out to the calibration model, is simplified calibration model.
Calculated to simplify, generally only consider single order and second order term, computation model is reduced to：
(X, Y)T=(x, y)T(1+k1r2+k2r4)
Simplify by model, on the premise of Correctness of model is ensured, greatly reduce the complexity of computing,
Improve operation efficiency.
S303, based on the simplified calibration model, forms deformity correction algorithm, and based on the deformity correction algorithm to institute
Stating panoramic picture carries out distortion correction operation.
Specifically, referring to Fig. 4, step S303 can be realized in the following way：
S401, searches the center of distortion point of the panoramic picture, and the center of distortion point is set into the panoramic picture
Two-dimensional coordinate origin.
For specific panorama shooting device, because the shape of panorama component is fixed, its center of distortion point
It is usually fixed a little, after panoramic picture is obtained, as long as the coordinate points for generally searching fixation can be found in distorting
Heart point.Meanwhile, with the center of distortion as the origin of coordinates, set two-dimensional coordinate system.
S402, determines lateral coordinates x and longitudinal direction seat that the pixel p on the panoramic picture is fastened in the two-dimensional coordinate
Mark y；
Specifically, pixel distances and orientation of the current pixel point p apart from center of distortion, such as pixel p can be obtained
Positioned at the upper left side of center of distortion point, the horizontal and vertical pixel distance of its range coordinate origin is respectively 400 and 500, then may be used
With by the lateral coordinates x and longitudinal coordinate y of current pixel point p be defined as -400 and 500, the i.e. coordinate of pixel p for (- 400,
500)。
S403, based on the deformity correction algorithm, coordinate transform is carried out to the lateral coordinates x and the longitudinal coordinate y,
Obtain new lateral coordinates X and new longitudinal coordinate Y；
Coordinate transform can be carried out to the coordinate of pixel p using formula (X, Y) T=(x, y) T (1+k1r2+k2r4), entered
And obtain the new coordinate of pixel p.
S404, coordinates of the pixel p on the panoramic picture is transformed to (X, Y) by (x, y).
By aforesaid operations, effective treatment can be carried out to lopsided image, be subsequently effectively to carry out gesture identification to carry
Basis is supplied.
Referring to Fig. 5, the embodiment of the present invention additionally provides a kind of specific gesture identification mode, comprises the following steps：
S501, similarity-rough set is carried out by the gesture feature value with the parameter in default gesture identification model, obtains in one's hands
Gesture Similarity value.
Because the randomness of the gesture of people, need many people's multimodes in advance to same gesture to do static statistics, carry
The discrimination of gesture high.For each gesture identification, its result of calculation is the Similarity value represented using probability.
Gesture operation can be identified using various gestures recognizer, such as image matching algorithm, hidden Ma Erke
Husband's model, Meanshift algorithms etc..It is compared with the parameter in identification model by by gesture feature value, can be obtained in one's hands
Gesture Similarity value.
Whether S502, judge gesture Similarity value less than the first default Similarity value.
After step S501 obtains gesture Similarity value, multiple similarity judgment thresholds can be set, such as, and setting first
Default Similarity value is 40%, the probability for less than 40%, it is believed that be no gesture.It is of course also possible to set as needed
First default Similarity value is other numerical value.
Whether S503, judge block image more than pre-set dimension.
Because the piecemeal to panoramic picture is that substep is carried out, now need to judge the size of block image whether more than pre-
If size (such as 200 × 200).
S504, continues to perform cutting operation to the block image.
In order to improve the discrimination of block image, the image for block image more than pre-set dimension can be continued executing with
Cutting operation treatment.
S505, whether gesture Similarity value is more than the second default Similarity value.
In addition to judging similarity whether less than the first preset value, can also further judge that gesture Similarity value is
It is no to be more than the second default Similarity value (for example, 60%).
S506, gesture command corresponding with the gesture feature value is defined as the order of the gesture operation.
Characteristic value is obtained more than the second default Similarity value for gesture similarity, assert that the gesture command determines, directly existed
Corresponding order is searched in default characteristic value-command mapping table, the order that will be found is defined as the order of gesture operation.
In addition to the foregoing steps, can also by certain probability interval be set as indeterminacy section (such as 40%~
60%), for this interval gesture operation, following steps are performed：
S507, order operation undetermined is defined as by the gesture operation.
S508, the storage gesture feature value related to the order operation undetermined.
For uncertain gesture identification, dynamic statistics are carried out using Bayesian Classification Arithmetic, dynamic adjusts identification parameter,
So that being improved for the gesture identification rate of fixed user.
Through the above way, corresponding operation can be performed for different types of gesture operation, improves gesture identification
The degree of accuracy.
For uncertain gesture identification, dynamic statistics are carried out using Bayesian Classification Arithmetic, dynamic adjusts identification parameter,
Based on it is assumed hereinafter that：
(1) randomness of the gesture of user so that when user does gesture, is not that gesture all can standards and norms each time.
(2) during Intelligent hardware is operated, when user has done a nonstandard gesture and Intelligent hardware is not responded to,
User can be appreciated that the lack of standardization of oneself, and then then do a gesture for specification.
(3) Intelligent hardware will associate same type of lack of standardization and specification gesture, be calculated with Bayes's classification
Method carries out dynamic statistics, dynamic adjustment identification parameter.
(4) when the gesture record for same type lack of standardization is more, an individual character for same type gesture will be formed
Gesture so that the gesture identification rate for fixed user is improved.
Therefore, referring to Fig. 6, gesture identification method provided in an embodiment of the present invention can also comprise the following steps：
S601, the statistics gesture feature value related to the order operation undetermined.
Specifically, the gesture feature value of the order operation correlation undetermined in preset time period can be counted, or judge life
Make whether the number of the related gesture feature value of operation undetermined reaches predetermined threshold value, when the gesture feature for ordering operation undetermined related
When value reaches predetermined threshold value, the gesture feature value to ordering operation undetermined related is counted.
S602, the specific instruction of the order operation undetermined is determined based on the gesture feature value.
For uncertain gesture, dynamic statistics, dynamic adjustment identification parameter are carried out using Bayesian Classification Arithmetic so that
Gesture identification rate for fixed user is improved.So as to enable users to operate nature, and gesture command meets oneself
Action feature.
S603 carries out more the gesture identification parameter related to the specific instruction in the default gesture identification model
Newly.
Through the above way, the recognition accuracy of gesture command can further be improved.
It is corresponding with above-mentioned gesture identification method, referring to Fig. 7, the embodiment of the invention also discloses a kind of gesture identifying device
70, including：
Acquisition module 701, for obtaining the panorama sketch comprising gesture operation that single camera shoots from panorama component
Picture.
In gesture identification field, the camera for gesture identification mostly uses common plane camera.These cameras are clapped
The scene taken the photograph, can only be unidirectional, have the shortcomings that visual angle is small, there is dead angle.If doing gesture identification and control with this camera
System, at the back of Intelligent hardware, just has very big dead angle.In this region, the gesture of people cannot be caught by Intelligent hardware
Arrive, so that None- identified and control.After 360 degree of panorama cameras, Intelligent hardware just can capture people in the range of 360 degree
Gesture, such that it is able to without dead angle ground recipient instruction.
Specifically, after camera 101 has gathered panoramic picture, panorama camera device or communicated to connect with panorama camera device
Other electronic equipments, can obtain camera 101 collection panoramic picture.Panoramic picture can be that camera is clapped by timing
Picture frame is extracted in video that is that the mode taken the photograph is obtained, or being shot from camera to obtain.
Correction module 702, for utilizing deformity correction algorithm corresponding with the panorama component, enters to the panoramic picture
Line distortion correct operation, obtains correction chart picture.
The collection of panoramic picture is carried out using panorama component, due to camber reflection or the presence of refraction, causes to obtain panorama
Can there is distortion in image.Accordingly, it would be desirable to the reflection or refraction situation to panorama component carry out mathematical modeling, based on the Mathematical Modeling
Panorama component is reflected or the distortion figure of refraction generation is corrected, form correction chart picture.
By taking radial distortion as an example, radial distortion be typically due between picture point and its ideal image position occur it is interior
To or export-oriented movement and the ranging offset that causes.Direction according to distortion is different, can be divided into positive radial distortion variable and bear
Radial distortion variable.Positive radial distortion amount makes pixel be moved away from the direction of image centerline, causes pincushion distortion.Phase
Instead, negative radial distortion amount makes pixel want to be moved close to the direction of image center, will so cause barrel distortion.Thus
Refine radial distortion Mathematical Modeling be：
δx=x (k1r2+k2r4+k3r6+…)
δy=y (k1r2+k2r4+k3r6+…)
Wherein, δx、δyRepresent the coordinate (x, y) of pixel p in horizontal and vertical side-play amount, k1、k2、k3It is distortion model
Coefficient.
Piecemeal module 703, for performing piecemeal operation to the correction chart picture, obtains block image；
Because panoramic picture stock size is all than larger, if recognizing the mode of gesture according to an image overall, it will
Cause processing speed slower.For the video image of captured in real-time, such as the video image of 30FPS directly carries out image procossing,
The processing pressure of processor will be caused to become big, or even the real-time processing of video image can not be completed, lead to not complete gesture
Identification.Piecemeal is carried out to image, the polylith image after piecemeal can carry out parallel processing with GPU, can accelerate gesture identification
Speed.
Due to the influence for distorting, if the global recognition gesture in fault image, global total distortion can be than larger, other
Region can have very big interference to gesture area.If carrying out image block, every piece of distortion of image relatively can be smaller, other
Region is just small to gesture area interference, it is possible to improve the probability of gesture identification.
, it is necessary to carry out dividing processing to panoramic picture in practical operation, treatment is identified parallel.Block image it is big
Small, algorithm can be using diminishing step by step, and the mode for staggeredly positioning is carried out.For example, can be according to present image 1/2 side long and wide
Formula segmentation go down, until the image after segmentation it is long × it is wide be less than 64 × 64 pixels just stop split.When image segmentation is carried out, can
Staggeredly to position so that staff is possible in the central area of image.
Determining module 704, the gesture feature value for extracting the block image, and determined based on the gesture feature value
The order of the gesture operation.
It is existing in the panoramic view with distortion and distortion because panoramic view has certain distortion and distortion
Gesture Recognition Algorithm can fail, and the speed of one two field picture for the treatment of also can be slower.
Therefore, on existing Gesture Recognition Algorithm, the characteristic value of training and dynamic adjustment gesture identification so that in radial distortion
Under the conditions of, gesture identification rate is improved, and the speed of one two field picture for the treatment of also can be quickly.
Specifically, existing gesture algorithm is identified both for the gesture of closely (＜ 2m).And control intelligence
Hardware in larger distance, it is necessary to do gesture identification, for example, interior is within 5 meters, outdoor is within 10 meters.So needing
Artificial optimization is carried out for longer-distance gesture.
Gesture identification is carried out for feature based value, it usually needs carry out model training in advance, by artificial mark and machine
The mode of device study, it is determined that the corresponding operational order of final characteristic value.
Used as an example, for the training of gesture identification in Radial Distortion Image, job step is divided into：
(1) the plane picture storehouse of collection belt gesture；
(2) gesture position in the picture is indicated with existing Gesture Recognition Algorithm；
(3) with the method for radial distortion, original image is carried out radial distortion.And calculate the hand that previous step sign is obtained
The new position of gesture；
(4) characteristic value that previous step indicates the image of gesture position is calculated with existing Gesture Recognition Algorithm；
(5) parameter of existing Gesture Recognition Algorithm is adjusted, the characteristic value for making it previous step be calculated is regarded as
Gesture.
Device in the present embodiment, by using single camera capturing panoramic view image, keeps away while reducing equipment cost
The appearance of vision dead zone is exempted from；Mode based on piecemeal treatment is split to panoramic picture, reduces processor treatment panorama
The pressure of image；Gesture identification is carried out using the Gesture Recognition Algorithm after adjustment, the degree of accuracy of gesture identification is improve.
The corresponding embodiment of the other functions of gesture identifying device provided in an embodiment of the present invention 70 and gesture identification method
Or implementation method is corresponding, will not be repeated here.
Referring to Fig. 8, the embodiment of the present invention additionally provides a kind of electronic equipment 80, and electronic equipment 80 can include：At least one
Individual processor 801, memory 802, input/output interface 803, radio circuit 804, voicefrequency circuit 805, camera assembly 806 and complete
Scape component 807.Wherein, radio circuit 804 receives signal by antenna 8041；Voicefrequency circuit 805 respectively with loudspeaker 8051 and wheat
Gram wind 8052 is connected；Camera assembly 806 is used to obtain the panoramic rays of the offer of panorama component 807, forms panoramic picture or panorama
Video, camera assembly 806 can be the camera 101 that shows in Fig. 1 a and Fig. 1 b, or other kinds of with shooting
The equipment of function；Panoramic picture or panoramic video storage are in memory 802.At least one processor 801 and memory
802 communication connections, the memory 802 is stored with can be by the instruction of at least one computing device, and the instruction is by institute
State at least one processor 801 to perform, so that at least one processor is able to carry out foregoing any gesture identification method
Embodiment.
The electronic equipment exists as a single image recognition apparatus, it is also possible to match somebody with somebody as one of other equipment
Part, instructs for providing gesture identification to other equipment.For example, the electronic equipment can exist in a variety of forms, including
But it is not limited to：
(1) mobile communication equipment：The characteristics of this kind equipment is that possess mobile communication function, and to provide speech, data
It is main target to communicate.This Terminal Type includes：Smart mobile phone (such as iPhone), multimedia handset, feature mobile phone, and it is low
End mobile phone etc..
(2) super mobile personal computer equipment：This kind equipment belongs to the category of personal computer, there is calculating and treatment work(
Can, typically also possess mobile Internet access characteristic.This Terminal Type includes：PDA, MID and UMPC equipment etc., such as iPad.
(3) portable entertainment device：This kind equipment can show and play content of multimedia.The kind equipment includes：Audio,
Video player (such as iPod), handheld device, e-book, and intelligent toy and portable car-mounted navigation equipment.
(4) particular server：The equipment for providing the service of calculating, the composition of server includes processor, hard disk, internal memory, is
System bus etc., server is similar with general computer architecture, but due to needing to provide highly reliable service, therefore in treatment
The requirement of the aspects such as ability, stability, reliability, security, scalability, manageability is higher.
(5) with gesture identification function unmanned plane, robot or similar products.
(6) other have the electronic equipment of gesture identification function.
It should be noted that herein, such as first and second or the like relational terms are used merely to a reality
Body or operation make a distinction with another entity or operation, and not necessarily require or imply these entities or deposited between operating
In any this actual relation or order.And, term " including ", "comprising" or its any other variant be intended to
Nonexcludability is included, so that process, method, article or equipment including a series of key elements not only will including those
Element, but also other key elements including being not expressly set out, or also include being this process, method, article or equipment
Intrinsic key element.In the absence of more restrictions, the key element limited by sentence "including a ...", it is not excluded that
Also there is other identical element in process, method, article or equipment including the key element.
Each embodiment in this specification is described by the way of correlation, identical similar portion between each embodiment
Divide mutually referring to what each embodiment was stressed is the difference with other embodiment.
For especially for device embodiment, because it is substantially similar to embodiment of the method, so the comparing of description is simple
Single, the relevent part can refer to the partial explaination of embodiments of method.
Represent in flow charts or logic and/or step described otherwise above herein, for example, being considered use
In the order list of the executable instruction for realizing logic function, in may be embodied in any computer-readable medium, for
Instruction execution system, device or equipment (such as computer based system, including the system of processor or other can be held from instruction
The system of row system, device or equipment instruction fetch and execute instruction) use, or with reference to these instruction execution systems, device or set
It is standby and use.For the purpose of this specification, " computer-readable medium " can any can be included, store, communicate, propagate or pass
The dress that defeated program is used for instruction execution system, device or equipment or with reference to these instruction execution systems, device or equipment
Put.The more specifically example (non-exhaustive list) of computer-readable medium includes following：With the electricity that one or more are connected up
Connecting portion (electronic installation), portable computer diskette box (magnetic device), random access memory (RAM), read-only storage
(ROM), erasable edit read-only storage (EPROM or flash memory), fiber device, and portable optic disk is read-only deposits
Reservoir (CDROM).In addition, computer-readable medium can even is that the paper that can thereon print described program or other are suitable
Medium, because optical scanner for example can be carried out by paper or other media, then enters edlin, interpretation or if necessary with it
His suitable method is processed electronically to obtain described program, is then stored in computer storage.
It should be appreciated that each several part of the invention can be realized with hardware, software, firmware or combinations thereof.
In the above-described embodiment, multiple steps or method can in memory and by suitable instruction be performed with storage
The software or firmware that system is performed are realized.If for example, being realized with hardware, with another embodiment, can use
Any one of following technology well known in the art or their combination are realized：With for realizing logic work(to data-signal
The discrete logic of the logic gates of energy, the application specific integrated circuit with suitable combinational logic gate circuit, programmable gate
Array (PGA), field programmable gate array (FPGA) etc..
The above, specific embodiment only of the invention, but protection scope of the present invention is not limited thereto, and it is any
Those familiar with the art the invention discloses technical scope in, the change or replacement that can be readily occurred in, all should
It is included within the scope of the present invention.Therefore, protection scope of the present invention should be defined by scope of the claims.
Claims (10)
1. a kind of gesture identification method, it is characterised in that including：
Obtain the panoramic picture comprising gesture operation that single camera shoots from panorama component；
Using deformity correction algorithm corresponding with the panorama component, distortion correction operation is carried out to the panoramic picture, obtained
Correction chart picture；
Piecemeal operation is performed to the correction chart picture, block image is obtained；
The gesture feature value of the block image is extracted, and the order of the gesture operation is determined based on the gesture feature value.
2. gesture identification method according to claim 1, it is characterised in that described using corresponding with the panorama component
Deformity correction algorithm, distortion correction operation is carried out to the panoramic picture, including：
Obtain calibration model corresponding with the panorama component；
Parameter predigesting operation is carried out to the calibration model, calibration model is simplified；
Based on the simplified calibration model, deformity correction algorithm is formed, and based on the deformity correction algorithm to the panorama sketch
As carrying out distortion correction operation.
3. gesture identification method according to claim 2, it is characterised in that it is described based on the deformity correction algorithm to institute
Stating panoramic picture carries out distortion correction operation, including：
The center of distortion point of the panoramic picture is searched, the center of distortion point is set to the two-dimensional coordinate of the panoramic picture
It is origin；
Determine lateral coordinates x and longitudinal coordinate y that the pixel p on the panoramic picture is fastened in the two-dimensional coordinate；
Based on the deformity correction algorithm, coordinate transform is carried out to the lateral coordinates x and the longitudinal coordinate y, obtain new
Lateral coordinates X and new longitudinal coordinate Y；
Coordinates of the pixel p on the panoramic picture is transformed to (X, Y) by (x, y).
4. gesture identification method according to claim 3, it is characterised in that described based on the deformity correction algorithm, it is right
The lateral coordinates x and the longitudinal coordinate y carry out coordinate transform, obtain new lateral coordinates X and new longitudinal coordinate Y, bag
Include：
Obtain the radial distance r between the pixel p and the two-dimensional coordinate system origin；
Obtain the first correction coefficient k of the simplified calibration model1And the second correction coefficient k2；
Based on the radial distance r, the first correction coefficient k1And the second correction coefficient k2, point is performed to coordinate (x, y)
Product computing, obtains the new coordinate (X, Y) of the pixel p.
5. gesture identification method according to claim 1, it is characterised in that the gesture of the extraction block image is special
Value indicative, and the order of the gesture operation is determined based on the gesture feature value, including：
The gesture feature value is carried out into similarity-rough set with the parameter in default gesture identification model, gesture similarity is obtained
Value；And
Based on the gesture Similarity value, the order of the gesture operation is determined.
6. gesture identification method according to claim 5, it is characterised in that described based on the gesture Similarity value, really
The order of the fixed gesture operation, including：
When the gesture Similarity value presets Similarity value less than first and the block image is more than pre-set dimension, continue
Cutting operation is performed to the block image.
7. gesture identification method according to claim 5, it is characterised in that described based on the gesture Similarity value, really
The order of the fixed gesture operation, including：
When the gesture Similarity value is more than the second default Similarity value, will gesture command corresponding with the gesture feature value
It is defined as the order of the gesture operation.
8. gesture identification method according to claim 5, it is characterised in that described based on the gesture Similarity value, really
The order of the fixed gesture operation, including：
When the closed-loop interval that the gesture Similarity value is constituted between the first default Similarity value and the second default Similarity value,
The gesture operation is defined as order operation undetermined；And
Store the gesture feature value related to the order operation undetermined.
9. a kind of gesture identifying device, it is characterised in that including：
Acquisition module, for obtaining the panoramic picture comprising gesture operation that single camera shoots from panorama component；
Correction module, for utilizing deformity correction algorithm corresponding with the panorama component, line distortion is entered to the panoramic picture
Correct operation, obtains correction chart picture；
Piecemeal module, for performing piecemeal operation to the correction chart picture, obtains block image；
Determining module, the gesture feature value for extracting the block image, and the hand is determined based on the gesture feature value
The order of gesture operation.
10. a kind of electronic equipment, it is characterised in that the electronic equipment includes：
At least one processor；And,
The memory being connected with least one processor communication；Wherein,
The memory storage has can be by the instruction of at least one computing device, and the instruction is by described at least one
Reason device is performed, so that at least one processor is able to carry out the gesture identification method described in foregoing any claim 1-9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710105905.9A CN106934351B (en) | 2017-02-23 | 2017-02-23 | Gesture recognition method and device and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710105905.9A CN106934351B (en) | 2017-02-23 | 2017-02-23 | Gesture recognition method and device and electronic equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106934351A true CN106934351A (en) | 2017-07-07 |
| CN106934351B CN106934351B (en) | 2020-12-29 |
Family
ID=59423147
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710105905.9A Active CN106934351B (en) | 2017-02-23 | 2017-02-23 | Gesture recognition method and device and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106934351B (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107212971A (en) * | 2017-07-11 | 2017-09-29 | 孟现华 | A kind of method for carrying aphasis patient |
| CN107260412A (en) * | 2017-07-11 | 2017-10-20 | 孟现华 | Folding intellectuality stretcher |
| CN108344442A (en) * | 2017-12-30 | 2018-07-31 | 广州本元信息科技有限公司 | A kind of object state detection recognition method, storage medium and system |
| CN108490607A (en) * | 2018-02-24 | 2018-09-04 | 江苏斯当特动漫设备制造有限公司 | A kind of holographic virtual implementing helmet based on cultural tour service |
| CN111950328A (en) * | 2019-05-15 | 2020-11-17 | 阿里巴巴集团控股有限公司 | Method and device for determining object class in picture |
| CN112329529A (en) * | 2020-09-29 | 2021-02-05 | 远光软件股份有限公司 | Automatic calibration and checking system and method for sample storage cabinet |
| CN112819725A (en) * | 2021-02-05 | 2021-05-18 | 广东电网有限责任公司广州供电局 | Rapid image correction method for radial distortion |
| CN113536879A (en) * | 2020-04-21 | 2021-10-22 | 和硕联合科技股份有限公司 | Image recognition method and device thereof, artificial intelligence model training method and device thereof |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102523395A (en) * | 2011-11-15 | 2012-06-27 | 中国科学院深圳先进技术研究院 | Television system having multi-point touch function, touch positioning identification method and system thereof |
| CN103118189A (en) * | 2013-01-25 | 2013-05-22 | 广东欧珀移动通信有限公司 | Post-call gesture operation method and post-call gesture operation device for mobile phone |
| CN103247031A (en) * | 2013-04-19 | 2013-08-14 | 华为技术有限公司 | Method, terminal and system for correcting aberrant image |
| CN105046249A (en) * | 2015-09-07 | 2015-11-11 | 哈尔滨市一舍科技有限公司 | Human-computer interaction method |
| CN105957015A (en) * | 2016-06-15 | 2016-09-21 | 武汉理工大学 | Thread bucket interior wall image 360 DEG panorama mosaicing method and system |
-
2017
- 2017-02-23 CN CN201710105905.9A patent/CN106934351B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102523395A (en) * | 2011-11-15 | 2012-06-27 | 中国科学院深圳先进技术研究院 | Television system having multi-point touch function, touch positioning identification method and system thereof |
| CN103118189A (en) * | 2013-01-25 | 2013-05-22 | 广东欧珀移动通信有限公司 | Post-call gesture operation method and post-call gesture operation device for mobile phone |
| CN103247031A (en) * | 2013-04-19 | 2013-08-14 | 华为技术有限公司 | Method, terminal and system for correcting aberrant image |
| CN105046249A (en) * | 2015-09-07 | 2015-11-11 | 哈尔滨市一舍科技有限公司 | Human-computer interaction method |
| CN105957015A (en) * | 2016-06-15 | 2016-09-21 | 武汉理工大学 | Thread bucket interior wall image 360 DEG panorama mosaicing method and system |
Non-Patent Citations (1)
| Title |
|---|
| JUHO KANNALA等: "A Generic Camera Model and Calibration Method for Conventional，Wide-Angle, and Fish-Eye Lenses", 《IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE》 * |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107212971A (en) * | 2017-07-11 | 2017-09-29 | 孟现华 | A kind of method for carrying aphasis patient |
| CN107260412A (en) * | 2017-07-11 | 2017-10-20 | 孟现华 | Folding intellectuality stretcher |
| CN107212971B (en) * | 2017-07-11 | 2018-09-21 | 薛红 | A kind of folding intelligence stretcher |
| CN107260412B (en) * | 2017-07-11 | 2019-01-04 | 雷磊 | Folding intelligence stretcher |
| CN108344442A (en) * | 2017-12-30 | 2018-07-31 | 广州本元信息科技有限公司 | A kind of object state detection recognition method, storage medium and system |
| CN108490607A (en) * | 2018-02-24 | 2018-09-04 | 江苏斯当特动漫设备制造有限公司 | A kind of holographic virtual implementing helmet based on cultural tour service |
| CN111950328A (en) * | 2019-05-15 | 2020-11-17 | 阿里巴巴集团控股有限公司 | Method and device for determining object class in picture |
| WO2020228593A1 (en) * | 2019-05-15 | 2020-11-19 | 阿里巴巴集团控股有限公司 | Method and apparatus for determining categories of target objects in picture |
| CN113536879A (en) * | 2020-04-21 | 2021-10-22 | 和硕联合科技股份有限公司 | Image recognition method and device thereof, artificial intelligence model training method and device thereof |
| CN112329529A (en) * | 2020-09-29 | 2021-02-05 | 远光软件股份有限公司 | Automatic calibration and checking system and method for sample storage cabinet |
| CN112329529B (en) * | 2020-09-29 | 2023-08-18 | 远光软件股份有限公司 | Automatic calibration checking device and method for sample storage cabinet |
| CN112819725A (en) * | 2021-02-05 | 2021-05-18 | 广东电网有限责任公司广州供电局 | Rapid image correction method for radial distortion |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106934351B (en) | 2020-12-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106934351A (en) | Gesture identification method, device and electronic equipment | |
| US10198823B1 (en) | Segmentation of object image data from background image data | |
| US9965865B1 (en) | Image data segmentation using depth data | |
| CN104317391B (en) | A kind of three-dimensional palm gesture recognition exchange method and system based on stereoscopic vision | |
| WO2023280065A1 (en) | Image reconstruction method and apparatus for cross-modal communication system | |
| CN108764048A (en) | Face critical point detection method and device | |
| US20200104568A1 (en) | Method and Device for Face Image Processing, Storage Medium, and Electronic Device | |
| CN110458805A (en) | Plane detection method, computing device and circuit system | |
| CN111160291B (en) | Human eye detection method based on depth information and CNN | |
| AU2014368997A1 (en) | System and method for identifying faces in unconstrained media | |
| US11868523B2 (en) | Eye gaze classification | |
| KR20220024736A (en) | Visual positioning method and related device | |
| JP2014032623A (en) | Image processor | |
| Zou et al. | Microarray camera image segmentation with Faster-RCNN | |
| EP3975047B1 (en) | Method for determining validness of facial feature, and electronic device | |
| CN114882222A (en) | Improved YOLOv5 target detection model construction method and tea tender shoot identification and picking point positioning method | |
| CN103150575A (en) | Real-time three-dimensional unmarked human body gesture recognition method and system | |
| Li et al. | FC-SLAM: Federated learning enhanced distributed visual-LiDAR SLAM in cloud robotic system | |
| CN112700568B (en) | Identity authentication method, equipment and computer readable storage medium | |
| Fan et al. | Accurate 3d single object tracker with local-to-global feature refinement | |
| Zhang et al. | Two-stage domain adaptation for infrared ship target segmentation | |
| CN112446845A (en) | Map construction method, map construction device, SLAM system, and storage medium | |
| Li et al. | Monocular 3-D Object Detection Based on Depth-Guided Local Convolution for Smart Payment in D2D Systems | |
| Fang et al. | Dynamic gesture recognition based on multimodal fusion model | |
| CN114565777A (en) | Data processing method and device |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |