CN111860346A - Dynamic gesture recognition method and device, electronic equipment and storage medium - Google Patents
Dynamic gesture recognition method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN111860346A CN111860346A CN202010713910.XA CN202010713910A CN111860346A CN 111860346 A CN111860346 A CN 111860346A CN 202010713910 A CN202010713910 A CN 202010713910A CN 111860346 A CN111860346 A CN 111860346A
- Authority
- CN
- China
- Prior art keywords
- target
- image
- gesture recognition
- hand
- images
- 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
- 238000000034 method Methods 0.000 title claims abstract description 82
- 238000003062 neural network model Methods 0.000 claims description 34
- 210000001364 upper extremity Anatomy 0.000 claims description 24
- 238000004590 computer program Methods 0.000 claims description 12
- 230000003993 interaction Effects 0.000 abstract description 14
- 238000001514 detection method Methods 0.000 abstract description 13
- 230000002829 reductive effect Effects 0.000 abstract description 4
- 230000008569 process Effects 0.000 description 20
- 230000003068 static effect Effects 0.000 description 8
- 238000010586 diagram Methods 0.000 description 7
- 230000009471 action Effects 0.000 description 6
- 230000006870 function Effects 0.000 description 6
- 230000006399 behavior Effects 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000036961 partial effect Effects 0.000 description 2
- 230000001427 coherent effect Effects 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000000670 limiting effect Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
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
Abstract
The application provides a dynamic gesture recognition method, a device, an electronic device and a storage medium, wherein the dynamic gesture recognition method comprises the following steps: acquiring a real-time shot image; determining a first target gesture recognition image from the real-time shot image; determining a plurality of second target gesture recognition images from the real-time shot images according to the first target gesture recognition image and a preset interval; combining the first target gesture recognition image and the plurality of second target gesture recognition images into a target dynamic gesture video; and identifying the target dynamic gesture video according to a preset dynamic gesture identification network model to obtain a dynamic gesture identification result. According to the dynamic gesture recognition method, the dynamic gesture recognition device, the electronic equipment and the storage medium, the gesture of the user is recognized through the dynamic gesture video, so that the false detection of the gesture of the user can be greatly reduced, the detection accuracy of gesture recognition is improved, and the limitation on the behavior of the user in the man-machine interaction scene is further reduced.
Description
Technical Field
The present application relates to the field of gesture recognition technologies, and in particular, to a dynamic gesture recognition method and apparatus, an electronic device, and a storage medium.
Background
With the continuous development and progress of scientific technology, gesture recognition technology has been gradually applied to the scene of human-computer interaction to realize the control of electronic equipment through gestures.
At present, in a human-computer interaction scene, a gesture recognition mode based on a static image is mostly adopted for gesture recognition, however, the gesture recognition mode based on the static image can only distinguish whether a gesture occurs in the static image, when a user carelessly places a gesture picture which is consistent with or similar to a certain operation gesture, the gesture recognition mode based on the static image can easily cause gesture false detection, and the gesture recognition limits the behavior of the user in the human-computer interaction scene.
Disclosure of Invention
An object of the embodiments of the present application is to provide a dynamic gesture recognition method, apparatus, electronic device, and storage medium, which recognize a gesture of a user through a dynamic gesture video, so as to greatly reduce false detection of the gesture of the user, improve detection accuracy of gesture recognition, and further reduce limitation on a behavior of the user in a human-computer interaction scene.
In a first aspect, an embodiment of the present application provides a dynamic gesture recognition method, including:
acquiring a real-time shot image;
determining a first target gesture recognition image from the real-time shot image;
determining a plurality of second target gesture recognition images from the real-time shot images according to the first target gesture recognition image and a preset interval;
combining the first target gesture recognition image and the plurality of second target gesture recognition images into a target dynamic gesture video;
and identifying the target dynamic gesture video according to a preset dynamic gesture identification network model to obtain a dynamic gesture identification result.
In the implementation process, the dynamic gesture recognition method of the embodiment of the application combines the first target gesture recognition image and the plurality of second target gesture recognition images into the target dynamic gesture video, and recognizes the target dynamic gesture video through the preset dynamic gesture recognition network model to obtain the dynamic gesture recognition result.
Further, before the acquiring the live-shot image, the method further includes:
acquiring an initial shooting image;
and adjusting the shooting parameters of the shooting component according to the target shooting object in the initial shooting image so as to enable the target shooting object to be at the main body position of the shooting picture.
In the implementation process, the method can adjust the shooting parameters of the shooting component through initially shooting the target shooting object in the image so as to enable the target shooting object to be located at the main body position of the shooting picture, and is particularly suitable for long-distance gesture recognition.
Further, the determining a first target gesture recognition image from the live-shot image comprises:
intercepting a first local shot image containing a target shot object from the real-time shot image according to the area of the target shot object in the real-time shot image;
carrying out image interception on the first local shot image according to a preset neural network model to obtain a first target hand image;
determining the first target hand image as a first target gesture recognition image.
In the implementation process, the method takes the first target hand image in the first local shot image containing the target shot object as the first target gesture recognition image, so that the first target gesture recognition image is mainly the hand image of the target shot object, the proportion of the hand image of the target shot object is increased, the hand characteristics of the target shot object are more obvious, the determination of a plurality of second target gesture recognition images and the gesture recognition of a target dynamic gesture video can be facilitated, and the detection accuracy of the gesture recognition is improved.
Further, the image capturing the first local shot image according to a preset neural network model to obtain a first target hand image includes:
analyzing and obtaining the hand position, the hand size, the hand picture ratio and the upper limb trunk information of the target shooting object in the first local shooting image through a preset neural network model;
and carrying out image interception on the first local shot image by using the hand position, the hand size, the hand picture ratio and the upper limb trunk information of the target shot object through the preset neural network model to obtain a first target hand image.
In the implementation process, the method combines the hand position, the hand size, the hand picture ratio and the upper limb trunk information of the target shooting object, carries out image interception on the first local shooting image containing the target shooting object to obtain a first target hand image, and can enable the intercepted first target hand image to be more complete, accurate and reliable.
Further, the determining a plurality of second target gesture recognition images from the real-time shot image according to the first target gesture recognition image and a preset interval includes:
determining a plurality of target shooting images from the real-time shooting images according to the first target gesture recognition image and a preset interval;
correspondingly intercepting a plurality of second local shot images containing the target shot object from each target shot image according to the area of the target shot object in each target shot image;
image interception is carried out on each second local shot image according to the preset neural network model, and a plurality of second target hand images are obtained;
determining a plurality of the second target hand images as a plurality of second target gesture recognition images.
In the implementation process, the method takes a second target hand image in a second local shot image containing a target shot object as a second target gesture recognition image, the second target gesture recognition image is mainly the hand image of the target shooting object, the proportion of the hand image of the target shooting object is increased, the hand characteristics of the target shooting object are more obvious, meanwhile, the first target gesture recognition image and the second target gesture recognition image determine the hand image of the target shooting object as the target gesture recognition image in the same way, so that the consistency of the first target gesture recognition image and the second target gesture recognition image is better ensured, the effect of the combined target dynamic gesture video is better, therefore, the gesture recognition of the dynamic gesture video can be facilitated, and the detection accuracy of the gesture recognition is further improved.
Further, the image capturing is performed on each second local shot image according to the preset neural network model to obtain a plurality of second target hand images, and the method includes:
analyzing and obtaining the hand position, the hand size, the hand picture ratio and the upper limb trunk information of the target shooting object in each second local shooting image through the preset neural network model;
and correspondingly carrying out image interception on each second local shot image according to the hand position, the hand size, the hand picture ratio and the upper limb trunk information of the target shot object in each second local shot image through the preset neural network model to obtain a plurality of second target hand images.
In the implementation process, the method combines the hand position, the hand size, the hand picture ratio and the upper limb trunk information of the target shooting object, carries out image interception on the second local shooting image containing the target shooting object to obtain a second target hand image, and can enable the intercepted second target hand image to be more complete, accurate and reliable.
Further, after the determining the first target hand image as a first target gesture recognition image, before determining a plurality of second target gesture recognition images from the real-time captured image according to the first target gesture recognition image and a preset interval, the method further includes:
judging whether an operation gesture exists in the first target gesture recognition image;
and if so, executing the step of determining a plurality of second target gesture recognition images from the real-time shot images according to the first target gesture recognition image and a preset interval.
In the implementation process, the method can determine the plurality of second target gesture recognition images from the real-time shot images when the operation gesture exists in the first target gesture recognition image, so that the gesture of a user is prevented from being dynamically recognized when the operation gesture does not exist in the first target gesture recognition image, unnecessary gesture dynamic recognition is reduced, and the gesture recognition efficiency can be improved.
In a second aspect, an embodiment of the present application provides a dynamic gesture recognition apparatus, including:
the acquisition module is used for acquiring a real-time shooting image;
the first determining module is used for determining a first target gesture recognition image from the real-time shot image;
the second determining module is used for determining a plurality of second target gesture recognition images from the real-time shot images according to the first target gesture recognition image and a preset interval;
the combination module is used for combining the first target gesture recognition image and the second target gesture recognition images into a target dynamic gesture video;
and the gesture recognition module is used for recognizing the target dynamic gesture video according to a preset dynamic gesture recognition network model to obtain a dynamic gesture recognition result.
In the implementation process, the dynamic gesture recognition device of the embodiment of the application combines the first target gesture recognition image and the plurality of second target gesture recognition images into the target dynamic gesture video, and recognizes the target dynamic gesture video through the preset dynamic gesture recognition network model to obtain the dynamic gesture recognition result.
In a third aspect, an embodiment of the present application provides an electronic device, including a memory and a processor, where the memory is used to store a computer program, and the processor runs the computer program to make the electronic device execute the above dynamic gesture recognition method.
In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, which stores a computer program, and the computer program, when executed by a processor, implements the above dynamic gesture recognition method.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are required to be used in the embodiments of the present application will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and that those skilled in the art can also obtain other related drawings based on the drawings without inventive efforts.
Fig. 1 is a first flowchart of a dynamic gesture recognition method according to an embodiment of the present disclosure;
fig. 2 is a second flowchart of a dynamic gesture recognition method according to an embodiment of the present application;
fig. 3 is a schematic flowchart of step S140 according to a first embodiment of the present application;
fig. 4 is a block diagram of a dynamic gesture recognition apparatus according to a second embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be described below with reference to the drawings in the embodiments of the present application.
It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, it need not be further defined and explained in subsequent figures. Meanwhile, in the description of the present application, the terms "first", "second", and the like are used only for distinguishing the description, and are not to be construed as indicating or implying relative importance.
At present, in a human-computer interaction scene, a gesture recognition mode based on a static image is mostly adopted for gesture recognition, however, the gesture recognition mode based on the static image can only distinguish whether a gesture occurs in the static image, when a user carelessly places a gesture picture which is consistent with or similar to a certain operation gesture, the gesture recognition mode based on the static image can easily cause gesture false detection, and the gesture recognition limits the behavior of the user in the human-computer interaction scene.
In view of the above problems in the prior art, the present application provides a dynamic gesture recognition method, apparatus, electronic device and storage medium, which recognize a gesture of a user through a dynamic gesture video, so as to greatly reduce false detection of the gesture of the user, improve detection accuracy of gesture recognition, and further reduce limitation on a behavior of the user in a human-computer interaction scene.
Example one
Referring to fig. 1, fig. 1 is a first flowchart of a dynamic gesture recognition method according to an embodiment of the present disclosure. The dynamic gesture recognition method described below in the embodiments of the present application can be applied to a human-computer interaction device, which may be an unmanned aerial vehicle, a mobile phone, or a tablet computer, and the like.
The embodiment of the application mainly uses a dynamic gesture recognition method to be applied to an unmanned aerial vehicle for relevant explanation, wherein the unmanned aerial vehicle is provided with a camera shooting component.
The dynamic gesture recognition method comprises the following steps:
step S130, a real-time shot image is acquired.
In the present embodiment, the number of live-view images is plural, and the plural live-view images are generally continuously captured images.
A plurality of real-time shooting images can be obtained through the camera shooting component carried by the unmanned aerial vehicle.
Step S140, a first target gesture recognition image is determined from the real-time captured image.
In this embodiment, the first target gesture recognition image may be one of the real-time captured images, or may be a partial image of one of the real-time captured images.
Alternatively, when the first target gesture recognition image is determined from the real-time captured images, the determination may be performed according to the hand motion of the target object in the plurality of real-time captured images, for example, the determination may be performed according to whether the hand motion of the target object in the plurality of real-time captured images is similar to a certain preset operation gesture.
And S150, determining a plurality of second target gesture recognition images from the real-time shot images according to the first target gesture recognition image and the preset interval.
In this embodiment, the preset section may be a section defined by the number of frames of the image, and assuming that the number of frames of the real-time captured image corresponding to the first target gesture recognition image is N, the preset section may be an image with the number of frames N divided from the number of frames N-t to N + t in the image with the number of frames N-t to N + t, that is, the real-time captured image with the number of frames N-t to N + t corresponding to the plurality of second target gesture recognition images.
In addition, the preset interval may also be a time domain, and when a plurality of second target gesture recognition images are determined, the plurality of second target gesture recognition images may be determined from the plurality of real-time captured images according to the capturing time of the real-time captured image corresponding to the first target gesture recognition image and the preset time domain.
Step S160, combining the first target gesture recognition image and the plurality of second target gesture recognition images into a target dynamic gesture video.
In this embodiment, the target dynamic gesture video is a video including a dynamic process of a hand motion of the target shooting object.
In general, the hand motion of a target shooting object in a target dynamic gesture video is a coherent process.
And S170, identifying the target dynamic gesture video according to a preset dynamic gesture identification network model to obtain a dynamic gesture identification result.
In this embodiment, the preset dynamic gesture recognition network model is a pre-trained model, and the preset dynamic gesture recognition network model is used for recognizing the target dynamic gesture video to obtain a dynamic gesture recognition result.
It can be understood that the dynamic gesture recognition result may correspond to a certain operation or action of the unmanned aerial vehicle, or may be an operation or action without a corresponding unmanned aerial vehicle, that is, it is stated that the hand action of the target shooting object does not belong to an operation gesture.
Optionally, the dynamic gesture recognition result may be used to trigger the turning on and off of the gesture interaction function of the drone.
According to the dynamic gesture recognition method, the first target gesture recognition image and the second target gesture recognition images are combined into the target dynamic gesture video, the target dynamic gesture video is recognized through the preset dynamic gesture recognition network model, and a dynamic gesture recognition result is obtained.
In a human-computer interaction scenario between a user and an unmanned aerial vehicle, the gesture recognition of the unmanned aerial vehicle on the user is usually remote gesture recognition, and in order to improve the detection accuracy of the remote gesture recognition, a possible implementation manner is provided in the embodiment of the present application, referring to fig. 2, fig. 2 is a second flow diagram of the dynamic gesture recognition method provided in the embodiment of the present application, and the dynamic gesture recognition method in the embodiment of the present application may further include the following steps before acquiring a real-time captured image in step S130:
step S110, acquiring an initial shooting image;
in step S120, the shooting parameters of the image pickup device are adjusted according to the target shooting object in the initial shooting image, so that the target shooting object is at the subject position of the shooting picture.
The initial captured image may be a single sheet. The target photographic subject is a user, the target photographic subject can be single or multiple, and in the embodiment of the application, the target photographic subject is mainly taken as a single object for relevant explanation.
The shooting parameters of the image pickup means may be at least one of conventional shooting parameters such as a shooting angle, a shooting focal length, a resolution of a shot screen, and the like of the image pickup means.
It is understood that the subject of the target photographic subject is in the subject position of the photographic screen, that is, the ratio of the target photographic subject in the photographic screen is larger than that in the initial photographic image.
Alternatively, the proportion of the target photographic subject in the photographic screen may be set in advance.
Alternatively, when the shooting parameters of the image pickup device are adjusted according to the target shooting object in the initial shooting image so that the target shooting object is at the subject position of the shooting picture, the shooting parameters of the image pickup device may be adjusted according to the size and distribution position of the target shooting object in the initial shooting image.
In the process, the method can adjust the shooting parameters of the camera shooting component through initially shooting the target shooting object in the image so as to enable the target shooting object to be located at the main body position of the shooting picture, and is particularly suitable for long-distance gesture recognition.
It should be noted that, when the dynamic gesture recognition method according to the embodiment of the present application is applied to a mobile phone or a tablet computer, the steps S110 and S120 do not need to be executed.
In order to increase the proportion of the hand image of the target photographic object and make the hand characteristics of the target photographic object more obvious, the embodiment of the present application provides a possible implementation manner, see fig. 3, and fig. 3 is a schematic flowchart of step S140 provided in the embodiment of the present application, and the dynamic gesture recognition method in the embodiment of the present application, step S140, determining the first target gesture recognition image from the real-time photographic image, may include the following steps:
step S141, according to the area of the target shooting object in the real-time shooting image, intercepting a first local shooting image containing the target shooting object from the real-time shooting image;
step S142, carrying out image interception on the first local shot image according to a preset neural network model to obtain a first target hand image;
in step S143, the first target hand image is determined as the first target gesture recognition image.
In conjunction with the content of the step S140, the live view image in the step S141 may be a single live view image that can be determined according to the hand motion of the target photographic subject in the plurality of live view images.
The preset neural network model is a pre-trained model, and the preset neural network model can be a pre-trained multi-scale convolutional neural network model.
It will be appreciated that the first target hand image is primarily a hand image of the target photographic subject.
In the process, the method takes the first target hand image in the first local shot image containing the target shot object as the first target gesture recognition image, so that the first target gesture recognition image is mainly the hand image of the target shot object, the proportion of the hand image of the target shot object is increased, the hand characteristics of the target shot object are more obvious, the determination of a plurality of second target gesture recognition images and the gesture recognition of a target dynamic gesture video can be facilitated, and the detection accuracy of the gesture recognition is improved.
Optionally, in step S142, performing image capturing on the first local captured image according to a preset neural network model to obtain a first target hand image, which may include:
analyzing and obtaining the hand position, the hand size, the hand picture ratio and the upper limb trunk information of the target shooting object in the first local shooting image through a preset neural network model;
and carrying out image interception on the first local shot image by using the hand position, the hand size, the hand picture ratio and the upper limb trunk information of the target shot object through a preset neural network model to obtain a first target hand image.
The upper limb trunk is the upper limb trunk of the target shooting object, and the upper limb trunk information can include the position, size and picture proportion of the upper limb trunk.
In the process, the method combines the hand position, the hand size, the hand picture ratio and the upper limb trunk information of the target shooting object, carries out image interception on the first local shooting image containing the target shooting object to obtain a first target hand image, and can enable the intercepted first target hand image to be more complete, accurate and reliable.
Based on the above process, in order to better ensure consistency between the first target gesture recognition image and the second target gesture recognition image and facilitate gesture recognition of the dynamic gesture video, a possible implementation manner is provided in the embodiment of the present application, in the dynamic gesture recognition method of the embodiment of the present application, the step S150 of determining a plurality of second target gesture recognition images from the real-time captured images according to the first target gesture recognition image and the preset interval may include the following steps:
determining a plurality of target shooting images from the real-time shooting images according to the first target gesture recognition image and a preset interval;
correspondingly intercepting a plurality of second local shot images containing the target shot object from each target shot image according to the area of the target shot object in each target shot image;
image interception is carried out on each second local shot image according to a preset neural network model to obtain a plurality of second target hand images;
determining the plurality of second target hand images as a plurality of second target gesture recognition images.
It should be noted that, for the explanation of this embodiment, the above contents are referred to correspondingly, and are not repeated herein.
In the above process, the method takes a second target hand image in a second partial photographic image containing the target photographic subject as a second target gesture recognition image, the second target gesture recognition image is mainly the hand image of the target shooting object, the proportion of the hand image of the target shooting object is increased, the hand characteristics of the target shooting object are more obvious, meanwhile, the first target gesture recognition image and the second target gesture recognition image determine the hand image of the target shooting object as the target gesture recognition image in the same way, so that the consistency of the first target gesture recognition image and the second target gesture recognition image is better ensured, the effect of the combined target dynamic gesture video is better, therefore, the gesture recognition of the dynamic gesture video can be facilitated, and the detection accuracy of the gesture recognition is further improved.
Optionally, when image capturing is performed on each second local captured image according to a preset neural network model to obtain a plurality of second target hand images, the method may include:
analyzing and obtaining the hand position, the hand size, the hand picture ratio and the upper limb trunk information of the target shooting object in each second local shooting image through a preset neural network model;
and correspondingly carrying out image interception on each second local shot image by using the hand position, the hand size, the hand picture ratio and the upper limb trunk information of the target shot object in each second local shot image through a preset neural network model to obtain a plurality of second target hand images.
In the process, the method combines the hand position, the hand size, the hand picture ratio and the upper limb trunk information of the target shooting object, carries out image interception on the second local shooting image containing the target shooting object to obtain a second target hand image, and can enable the intercepted second target hand image to be more complete, accurate and reliable.
In a scene of human-computer interaction between a user and an unmanned aerial vehicle, in order to reduce unnecessary gesture dynamic recognition, a possible implementation manner is provided in the embodiments of the present application, and the dynamic gesture recognition method in the embodiments of the present application may further include, after determining a first target hand image as a first target gesture recognition image, before determining a plurality of second target gesture recognition images from a real-time captured image according to the first target gesture recognition image and a preset interval, the following steps:
judging whether an operation gesture exists in the first target gesture recognition image;
and if so, determining a plurality of second target gesture recognition images from the real-time shot images according to the first target gesture recognition image and the preset interval.
In the process, the method can determine the plurality of second target gesture recognition images from the real-time shot images when the operation gesture exists in the first target gesture recognition image, so that the gesture of a user is prevented from being dynamically recognized when the operation gesture does not exist in the first target gesture recognition image, unnecessary gesture dynamic recognition is reduced, and the gesture recognition efficiency can be improved.
Example two
In order to implement the corresponding method of the above embodiments to achieve the corresponding functions and technical effects, a dynamic gesture recognition apparatus is provided below.
Referring to fig. 4, fig. 4 is a block diagram of a dynamic gesture recognition apparatus according to an embodiment of the present disclosure.
The dynamic gesture recognition device of the embodiment of the application comprises:
an obtaining module 210, configured to obtain a real-time captured image;
a first determining module 220, configured to determine a first target gesture recognition image from the real-time captured image;
a second determining module 230, configured to determine a plurality of second target gesture recognition images from the real-time captured image according to the first target gesture recognition image and the preset interval;
a combination module 240, configured to combine the first target gesture recognition image and the plurality of second target gesture recognition images into a target dynamic gesture video;
and the gesture recognition module 250 is configured to recognize the target dynamic gesture video according to a preset dynamic gesture recognition network model to obtain a dynamic gesture recognition result.
The dynamic gesture recognition device combines the first target gesture recognition image and the second target gesture recognition images into a target dynamic gesture video, and recognizes the target dynamic gesture video through the preset dynamic gesture recognition network model to obtain a dynamic gesture recognition result.
As an optional implementation, the obtaining module 210 may be further configured to obtain an initial captured image;
the dynamic gesture recognition device in the embodiment of the application may further include a shooting adjustment module, configured to adjust shooting parameters of the shooting component according to a target shooting object in the initial shooting image, so that the target shooting object is located at a main body position of the shooting picture.
As an optional implementation manner, the first determining module 220 may be specifically configured to:
intercepting a first local shot image containing a target shot object from the real-time shot image according to the area of the target shot object in the real-time shot image;
image interception is carried out on the first local shot image according to a preset neural network model to obtain a first target hand image;
the first target hand image is determined to be a first target gesture recognition image.
Optionally, when the first determining module 220 performs image capturing on the first local captured image according to a preset neural network model to obtain the first target hand image, it may:
analyzing and obtaining the hand position, the hand size, the hand picture ratio and the upper limb trunk information of the target shooting object in the first local shooting image through a preset neural network model;
and carrying out image interception on the first local shot image by using the hand position, the hand size, the hand picture ratio and the upper limb trunk information of the target shot object through a preset neural network model to obtain a first target hand image.
As an optional implementation manner, the second determining module 230 may specifically be configured to:
determining a plurality of target shooting images from the real-time shooting images according to the first target gesture recognition image and a preset interval;
correspondingly intercepting a plurality of second local shot images containing the target shot object from each target shot image according to the area of the target shot object in each target shot image;
image interception is carried out on each second local shot image according to a preset neural network model to obtain a plurality of second target hand images;
determining the plurality of second target hand images as a plurality of second target gesture recognition images.
Optionally, when the second determining module 230 performs image capturing on each second local captured image according to a preset neural network model to obtain a plurality of second target hand images, it may:
analyzing and obtaining the hand position, the hand size, the hand picture ratio and the upper limb trunk information of the target shooting object in each second local shooting image through a preset neural network model;
and correspondingly carrying out image interception on each second local shot image by using the hand position, the hand size, the hand picture ratio and the upper limb trunk information of the target shot object in each second local shot image through a preset neural network model to obtain a plurality of second target hand images.
As an optional implementation manner, the dynamic gesture recognition apparatus according to the embodiment of the application may further include a determining module, configured to determine whether an operation gesture exists in the first target gesture recognition image.
The dynamic gesture recognition apparatus can implement the dynamic gesture recognition method of the first embodiment. The alternatives in the first embodiment are also applicable to the present embodiment, and are not described in detail here.
The rest of the embodiments of the present application may refer to the contents of the first embodiment, and in this embodiment, details are not repeated.
EXAMPLE III
An embodiment of the present application provides an electronic device, which includes a memory and a processor, where the memory is used to store a computer program, and the processor runs the computer program to make the electronic device execute the above dynamic gesture recognition method.
Optionally, the electronic device may be a human-computer interaction device, and the human-computer interaction device may be a drone, a mobile phone, or a tablet computer.
In addition, an embodiment of the present application further provides a computer-readable storage medium, which stores a computer program, and when the computer program is executed by a processor, the method for recognizing a dynamic gesture as described above is implemented.
In the embodiments provided in the present application, it should be understood that the disclosed apparatus and method can be implemented in other ways. The apparatus embodiments described above are merely illustrative, and for example, the flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
In addition, functional modules in the embodiments of the present application may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.
The functions, if implemented in the form of software functional modules and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application or portions thereof that substantially contribute to the prior art may be embodied in the form of a software product stored in a storage medium and including instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
The above description is only an example of the present application and is not intended to limit the scope of the present application, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, improvement and the like made within the spirit and principle of the present application shall be included in the protection scope of the present application. It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, it need not be further defined and explained in subsequent figures.
The above description is only for the specific embodiments of the present application, but the scope of the present application is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present application, and shall be covered by the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
Claims (10)
1. A dynamic gesture recognition method, comprising:
acquiring a real-time shot image;
determining a first target gesture recognition image from the real-time shot image;
determining a plurality of second target gesture recognition images from the real-time shot images according to the first target gesture recognition image and a preset interval;
combining the first target gesture recognition image and the plurality of second target gesture recognition images into a target dynamic gesture video;
and identifying the target dynamic gesture video according to a preset dynamic gesture identification network model to obtain a dynamic gesture identification result.
2. The dynamic gesture recognition method according to claim 1, wherein prior to said acquiring a live shot image, the method further comprises:
acquiring an initial shooting image;
and adjusting the shooting parameters of the shooting component according to the target shooting object in the initial shooting image so as to enable the target shooting object to be at the main body position of the shooting picture.
3. The dynamic gesture recognition method according to claim 1, wherein the determining a first target gesture recognition image from the live captured image comprises:
intercepting a first local shot image containing a target shot object from the real-time shot image according to the area of the target shot object in the real-time shot image;
carrying out image interception on the first local shot image according to a preset neural network model to obtain a first target hand image;
determining the first target hand image as a first target gesture recognition image.
4. The dynamic gesture recognition method according to claim 3, wherein the image capturing the first local captured image according to a preset neural network model to obtain a first target hand image comprises:
analyzing and obtaining the hand position, the hand size, the hand picture ratio and the upper limb trunk information of the target shooting object in the first local shooting image through a preset neural network model;
and carrying out image interception on the first local shot image by using the hand position, the hand size, the hand picture ratio and the upper limb trunk information of the target shot object through the preset neural network model to obtain a first target hand image.
5. The method according to claim 4, wherein the determining a plurality of second target gesture recognition images from the real-time captured images according to the first target gesture recognition image and a preset interval comprises:
determining a plurality of target shooting images from the real-time shooting images according to the first target gesture recognition image and a preset interval;
correspondingly intercepting a plurality of second local shot images containing the target shot object from each target shot image according to the area of the target shot object in each target shot image;
image interception is carried out on each second local shot image according to the preset neural network model, and a plurality of second target hand images are obtained;
determining a plurality of the second target hand images as a plurality of second target gesture recognition images.
6. The dynamic gesture recognition method according to claim 5, wherein the image capturing each second local captured image according to the preset neural network model to obtain a plurality of second target hand images comprises:
analyzing and obtaining the hand position, the hand size, the hand picture ratio and the upper limb trunk information of the target shooting object in each second local shooting image through the preset neural network model;
and correspondingly carrying out image interception on each second local shot image according to the hand position, the hand size, the hand picture ratio and the upper limb trunk information of the target shot object in each second local shot image through the preset neural network model to obtain a plurality of second target hand images.
7. The dynamic gesture recognition method of claim 3, wherein after the determining the first target hand image as a first target gesture recognition image, before the determining a plurality of second target gesture recognition images from the live capture image according to the first target gesture recognition image and a preset interval, the method further comprises:
judging whether an operation gesture exists in the first target gesture recognition image;
and if so, executing the step of determining a plurality of second target gesture recognition images from the real-time shot images according to the first target gesture recognition image and a preset interval.
8. A dynamic gesture recognition apparatus, comprising:
the acquisition module is used for acquiring a real-time shooting image;
the first determining module is used for determining a first target gesture recognition image from the real-time shot image;
the second determining module is used for determining a plurality of second target gesture recognition images from the real-time shot images according to the first target gesture recognition image and a preset interval;
the combination module is used for combining the first target gesture recognition image and the second target gesture recognition images into a target dynamic gesture video;
and the gesture recognition module is used for recognizing the target dynamic gesture video according to a preset dynamic gesture recognition network model to obtain a dynamic gesture recognition result.
9. An electronic device comprising a memory for storing a computer program and a processor for executing the computer program to cause the electronic device to perform the dynamic gesture recognition method according to any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that it stores a computer program which, when executed by a processor, implements a dynamic gesture recognition method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010713910.XA CN111860346A (en) | 2020-07-22 | 2020-07-22 | Dynamic gesture recognition method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010713910.XA CN111860346A (en) | 2020-07-22 | 2020-07-22 | Dynamic gesture recognition method and device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111860346A true CN111860346A (en) | 2020-10-30 |
Family
ID=72949679
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010713910.XA Pending CN111860346A (en) | 2020-07-22 | 2020-07-22 | Dynamic gesture recognition method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111860346A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112308041A (en) * | 2020-11-26 | 2021-02-02 | 中国人民解放军陆军工程大学 | Unmanned platform gesture control method based on vision |
| CN112527107A (en) * | 2020-11-30 | 2021-03-19 | 京东方科技集团股份有限公司 | Gesture recognition method and device, electronic equipment and storage medium |
| CN113197403A (en) * | 2021-05-14 | 2021-08-03 | 广东华联云谷科技研究院有限公司 | Method capable of preventing virus infection and smart bracelet |
| CN113642413A (en) * | 2021-07-16 | 2021-11-12 | 新线科技有限公司 | Control method, apparatus, device and medium |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130278504A1 (en) * | 2011-11-01 | 2013-10-24 | Xiaofeng Tong | Dynamic gesture based short-range human-machine interaction |
| US20160078289A1 (en) * | 2014-09-16 | 2016-03-17 | Foundation for Research and Technology - Hellas (FORTH) (acting through its Institute of Computer | Gesture Recognition Apparatuses, Methods and Systems for Human-Machine Interaction |
| WO2018033154A1 (en) * | 2016-08-19 | 2018-02-22 | 北京市商汤科技开发有限公司 | Gesture control method, device, and electronic apparatus |
| CN109409277A (en) * | 2018-10-18 | 2019-03-01 | 北京旷视科技有限公司 | Gesture identification method, device, intelligent terminal and computer storage medium |
| CN109886070A (en) * | 2018-12-24 | 2019-06-14 | 珠海格力电器股份有限公司 | A kind of apparatus control method, device, storage medium and equipment |
| CN109960980A (en) * | 2017-12-22 | 2019-07-02 | 北京市商汤科技开发有限公司 | Dynamic gesture identification method and device |
| CN110688914A (en) * | 2019-09-09 | 2020-01-14 | 苏州臻迪智能科技有限公司 | Gesture recognition method, intelligent device, storage medium and electronic device |
| US20200057886A1 (en) * | 2018-08-14 | 2020-02-20 | Boe Technology Group Co., Ltd. | Gesture recognition method and apparatus, electronic device, and computer-readable storage medium |
| CN111158467A (en) * | 2019-12-12 | 2020-05-15 | 青岛小鸟看看科技有限公司 | Gesture interaction method and terminal |
-
2020
- 2020-07-22 CN CN202010713910.XA patent/CN111860346A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130278504A1 (en) * | 2011-11-01 | 2013-10-24 | Xiaofeng Tong | Dynamic gesture based short-range human-machine interaction |
| US20160078289A1 (en) * | 2014-09-16 | 2016-03-17 | Foundation for Research and Technology - Hellas (FORTH) (acting through its Institute of Computer | Gesture Recognition Apparatuses, Methods and Systems for Human-Machine Interaction |
| WO2018033154A1 (en) * | 2016-08-19 | 2018-02-22 | 北京市商汤科技开发有限公司 | Gesture control method, device, and electronic apparatus |
| CN109960980A (en) * | 2017-12-22 | 2019-07-02 | 北京市商汤科技开发有限公司 | Dynamic gesture identification method and device |
| US20200057886A1 (en) * | 2018-08-14 | 2020-02-20 | Boe Technology Group Co., Ltd. | Gesture recognition method and apparatus, electronic device, and computer-readable storage medium |
| CN109409277A (en) * | 2018-10-18 | 2019-03-01 | 北京旷视科技有限公司 | Gesture identification method, device, intelligent terminal and computer storage medium |
| CN109886070A (en) * | 2018-12-24 | 2019-06-14 | 珠海格力电器股份有限公司 | A kind of apparatus control method, device, storage medium and equipment |
| CN110688914A (en) * | 2019-09-09 | 2020-01-14 | 苏州臻迪智能科技有限公司 | Gesture recognition method, intelligent device, storage medium and electronic device |
| CN111158467A (en) * | 2019-12-12 | 2020-05-15 | 青岛小鸟看看科技有限公司 | Gesture interaction method and terminal |
Non-Patent Citations (3)
| Title |
|---|
| 张圆圆;: "基于视觉的手势识别技术及其应用研究", 计算技术与自动化, no. 01 * |
| 王西颖;戴国忠;张习文;张凤军;: "基于HMM-FNN模型的复杂动态手势识别", 软件学报, no. 09 * |
| 田喜平;赵红丹;: "基于Kinect的动态孤立手势识别研究", 微电子学与计算机, no. 02 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112308041A (en) * | 2020-11-26 | 2021-02-02 | 中国人民解放军陆军工程大学 | Unmanned platform gesture control method based on vision |
| CN112527107A (en) * | 2020-11-30 | 2021-03-19 | 京东方科技集团股份有限公司 | Gesture recognition method and device, electronic equipment and storage medium |
| US11600116B2 (en) | 2020-11-30 | 2023-03-07 | Boe Technology Group Co., Ltd. | Methods and apparatuses for recognizing gesture, electronic devices and storage media |
| CN113197403A (en) * | 2021-05-14 | 2021-08-03 | 广东华联云谷科技研究院有限公司 | Method capable of preventing virus infection and smart bracelet |
| CN113197403B (en) * | 2021-05-14 | 2023-02-17 | 广州乾睿医疗科技有限公司 | Method capable of preventing virus infection and smart bracelet |
| CN113642413A (en) * | 2021-07-16 | 2021-11-12 | 新线科技有限公司 | Control method, apparatus, device and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10949952B2 (en) | Performing detail enhancement on a target in a denoised image | |
| CN109325933B (en) | Method and device for recognizing copied image | |
| CN111860346A (en) | Dynamic gesture recognition method and device, electronic equipment and storage medium | |
| US10762649B2 (en) | Methods and systems for providing selective disparity refinement | |
| US20150310259A1 (en) | Using facial data for device authentication or subject identification | |
| CN105681627B (en) | Image shooting method and electronic equipment | |
| CN111738735B (en) | Image data processing method and device and related equipment | |
| CN110688914A (en) | Gesture recognition method, intelligent device, storage medium and electronic device | |
| CN108037830B (en) | Method for realizing augmented reality | |
| CN111640165A (en) | Method and device for acquiring AR group photo image, computer equipment and storage medium | |
| CN113906437A (en) | Improved face quality of captured images | |
| CN112887615B (en) | Shooting method and device | |
| CN115620398A (en) | Target action detection method and device | |
| US11087121B2 (en) | High accuracy and volume facial recognition on mobile platforms | |
| CN114125226A (en) | Image shooting method and device, electronic equipment and readable storage medium | |
| CN112752110A (en) | Video presentation method and device, computing equipment and storage medium | |
| CN117097982B (en) | Target detection method and system | |
| CN110991307A (en) | Face recognition method, device, equipment and storage medium | |
| CN111753793B (en) | Model training method and device, face screening method and electronic equipment | |
| CN113691731B (en) | Processing method and device and electronic equipment | |
| KR102628714B1 (en) | Photography system for surpporting to picture for mobile terminal and method thereof | |
| US11003957B2 (en) | Method and apparatus for certificate identification | |
| CN110620911B (en) | Video stream processing method and device of camera and terminal equipment | |
| KR101951158B1 (en) | lighting device having a control function and method thereof | |
| CN117453635A (en) | Image deletion method, device, electronic equipment and readable storage medium |
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 |