CN109621332A - A kind of attribute determining method, device, equipment and the storage medium of body-building movement - Google Patents
A kind of attribute determining method, device, equipment and the storage medium of body-building movement Download PDFInfo
- Publication number
- CN109621332A CN109621332A CN201811638461.6A CN201811638461A CN109621332A CN 109621332 A CN109621332 A CN 109621332A CN 201811638461 A CN201811638461 A CN 201811638461A CN 109621332 A CN109621332 A CN 109621332A
- Authority
- CN
- China
- Prior art keywords
- dimensional
- building
- movement
- manikin
- practical
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- A—HUMAN NECESSITIES
- A63—SPORTS; GAMES; AMUSEMENTS
- A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
- A63B24/00—Electric or electronic controls for exercising apparatus of preceding groups; Controlling or monitoring of exercises, sportive games, training or athletic performances
- A63B24/0003—Analysing the course of a movement or motion sequences during an exercise or trainings sequence, e.g. swing for golf or tennis
- A63B24/0006—Computerised comparison for qualitative assessment of motion sequences or the course of a movement
-
- A—HUMAN NECESSITIES
- A63—SPORTS; GAMES; AMUSEMENTS
- A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
- A63B24/00—Electric or electronic controls for exercising apparatus of preceding groups; Controlling or monitoring of exercises, sportive games, training or athletic performances
- A63B24/0062—Monitoring athletic performances, e.g. for determining the work of a user on an exercise apparatus, the completed jogging or cycling distance
-
- A—HUMAN NECESSITIES
- A63—SPORTS; GAMES; AMUSEMENTS
- A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
- A63B71/00—Games or sports accessories not covered in groups A63B1/00 - A63B69/00
- A63B71/06—Indicating or scoring devices for games or players, or for other sports activities
- A63B71/0619—Displays, user interfaces and indicating devices, specially adapted for sport equipment, e.g. display mounted on treadmills
-
- A—HUMAN NECESSITIES
- A63—SPORTS; GAMES; AMUSEMENTS
- A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
- A63B24/00—Electric or electronic controls for exercising apparatus of preceding groups; Controlling or monitoring of exercises, sportive games, training or athletic performances
- A63B24/0003—Analysing the course of a movement or motion sequences during an exercise or trainings sequence, e.g. swing for golf or tennis
- A63B24/0006—Computerised comparison for qualitative assessment of motion sequences or the course of a movement
- A63B2024/0012—Comparing movements or motion sequences with a registered reference
- A63B2024/0015—Comparing movements or motion sequences with computerised simulations of movements or motion sequences, e.g. for generating an ideal template as reference to be achieved by the user
-
- A—HUMAN NECESSITIES
- A63—SPORTS; GAMES; AMUSEMENTS
- A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
- A63B2220/00—Measuring of physical parameters relating to sporting activity
- A63B2220/05—Image processing for measuring physical parameters
Abstract
The invention discloses attribute determining method, device, equipment and the storage mediums of a kind of body-building movement, this method comprises: obtaining the body-building video of user, extraction obtains the video frame in body-building video;By video frame input three-dimensional prediction model trained in advance, corresponding practical three-dimensional (3 D) manikin is obtained;According to the parameter of key point each in practical three-dimensional (3 D) manikin, the actual motion parameter of user is determined;According to actual motion parameter and standard movement parameter, the attribute that body-building acts in body-building video is determined.The present invention passes through three-dimensional prediction model trained in advance, the two-dimensional video frame image in body-building video directly can be converted into practical three-dimensional (3 D) manikin, so as to according to the parameter of each key point in practical three-dimensional (3 D) manikin, determine the standard degree of body-building movement, realize the recognition accuracy that ensure that the attribute of body-building movement while reducing use cost.
Description
Technical field
The present embodiments relate to attribute determining method, devices, equipment that physical training skill more particularly to a kind of body-building act
And storage medium.
Background technique
With the continuous promotion of people's living standard, people increasingly pay attention to the raising of physical fitness.In practical body-building
Cheng Zhong, in order to identify body-building movement whether standard, can be instructed by personal coach, but taking due to personal coach
With higher, it is unable to satisfy proprietary body-building needs.
It is passed currently, mainly being measured using the multi-shaft inertial in wearable device (for example, Intelligent bracelet, smartwatch etc.)
Sensor identifies body-building movement.But wearable device is chiefly used in the scene that running etc. needs swing arm, thus to swing arm
When unconspicuous body-building movement is acquired identification, the recognition effect which acts body-building is bad.
In visual direction, usually obtained by the artis of user in detection two dimensional image or directly using depth camera
The three-dimensional coordinate of artis identifies body-building movement, and estimates the standard degree of body-building movement.But due in two dimensional image
Artis causes the accuracy of identification to be affected due to lacking depth information;Although depth camera accuracy is higher, due to
The reasons such as its price cause the popularity rate of depth camera lower, to be unable to satisfy the demand of most of user.
Summary of the invention
In view of this, the present invention provides attribute determining method, device, equipment and the storage medium of a kind of body-building movement,
While reducing use cost, the recognition accuracy of the attribute of body-building movement ensure that.
In a first aspect, the embodiment of the invention provides a kind of attribute determining methods of body-building movement, comprising:
The body-building video of user is obtained, extraction obtains the video frame in the body-building video;
By video frame input three-dimensional prediction model trained in advance, corresponding practical three-dimensional (3 D) manikin, institute are obtained
Stating three-dimensional prediction model is that the standard three-dimensional manikin of respective action in two dimensional image and two dimensional image is inputted to depth mind
It is obtained through network training;
According to the parameter of each key point in the practical three-dimensional (3 D) manikin, the actual motion parameter of user is determined;
According to the actual motion parameter and standard movement parameter, the attribute that body-building acts in the body-building video is determined.
Further, the attribute of the body-building movement includes at least one of following: the standard degree of body-building movement, body-building movement
Type.
Further, the three-dimensional prediction model is that corresponding standard three-dimensional will be acted in two dimensional image and two dimensional image
Manikin input deep neural network training obtains, comprising:
Acquire the two dimensional image that body-building of the user at same visual angle acts;
Corresponding standard three-dimensional manikin is found according to the body-building movement in the two dimensional image;
The two dimensional image and standard three-dimensional manikin input deep neural network are trained, corresponded to
Three-dimensional prediction model.
Further, it in the three-dimensional prediction model that video frame input is trained in advance, obtains corresponding practical three-dimensional
After manikin, further includes:
Corresponding practical three-dimensional (3 D) manikin is formed into practical three-dimensional (3 D) manikin sequence according to the extraction sequence of video frame.
Further, the parameter according to each key point in the practical three-dimensional (3 D) manikin, determines the reality of user
Kinematic parameter, comprising:
Extract the reality of each key point in each practical three-dimensional (3 D) manikin in the practical three-dimensional (3 D) manikin sequence
Border three-dimensional coordinate;
Combination is ranked up to the practical three-dimensional coordinate of each key point according to the extraction sequence of video frame, is generated practical three-dimensional
Coordinate sequence;
The actual motion parameter of user is determined according to the practical three-dimensional coordinate sequence.
Further, the kinematic parameter includes at least one of following: the relative position variation of motion profile, each key point
Change with the relative angle of each key point.
Second aspect, the embodiment of the invention also provides a kind of attribute determining devices of body-building movement, comprising:
Module is obtained, for obtaining the body-building video of user, extraction obtains the video frame in the body-building video;
First determining module obtains corresponding reality for the three-dimensional prediction model that video frame input is trained in advance
Border three-dimensional (3 D) manikin, the three-dimensional prediction model are by the standard three-dimensional people of respective action in two dimensional image and two dimensional image
Body Model input deep neural network training obtains;
Second determining module determines user's for the parameter according to each key point in the practical three-dimensional (3 D) manikin
Actual motion parameter;
Third determining module, for determining the body-building video according to the actual motion parameter and standard movement parameter
The attribute of middle body-building movement.
Further, the three-dimensional prediction model is that corresponding standard three-dimensional will be acted in two dimensional image and two dimensional image
Manikin input deep neural network training obtains, and is specifically used for:
Acquire the two dimensional image that body-building of the user at same visual angle acts;
Corresponding standard three-dimensional manikin is found according to the body-building movement in the two dimensional image;
The two dimensional image and standard three-dimensional manikin input deep neural network are trained, corresponded to
Three-dimensional prediction model.
Further, described device, further includes:
Comprising modules, for obtaining corresponding reality in the three-dimensional prediction model that video frame input is trained in advance
After three-dimensional (3 D) manikin, corresponding practical three-dimensional (3 D) manikin is formed into practical 3 D human body according to the extraction sequence of video frame
Model sequence.
Further, second determining module, comprising:
Extraction unit, for extracting in the practical three-dimensional (3 D) manikin sequence in each practical three-dimensional (3 D) manikin
The practical three-dimensional coordinate of each key point;
Combination producing unit is ranked up the practical three-dimensional coordinate of each key point for the extraction sequence according to video frame
Combination generates practical three-dimensional coordinate sequence;
Determination unit, for determining the actual motion parameter of user according to the practical three-dimensional coordinate sequence.
Further, the kinematic parameter includes at least one of following: the relative position variation of motion profile, each key point
Change with the relative angle of each key point.
The third aspect, the embodiment of the invention also provides a kind of body-building movement attribute determine equipment, comprising: memory with
And one or more processors;
The memory, for storing one or more programs;
When one or more of programs are executed by one or more of processors, so that one or more of processing
Device realizes the attribute determining method of body-building movement as described in relation to the first aspect.
Fourth aspect, it is described the embodiment of the invention also provides a kind of storage medium comprising computer executable instructions
Computer executable instructions by computer processor when being executed for executing the attribute of body-building movement as described in relation to the first aspect
Determine method.
The present invention extracts by the body-building video of acquisition user and obtains the video frame in body-building video;Video frame is inputted
Three-dimensional prediction model trained in advance, obtains corresponding practical three-dimensional (3 D) manikin, three-dimensional prediction model for by two dimensional image with
And the standard three-dimensional manikin input deep neural network training of respective action obtains in two dimensional image;According to practical three-dimensional people
The parameter of each key point in body Model determines the actual motion parameter of user;According to actual motion parameter and standard movement parameter,
Determine the attribute that body-building acts in body-building video.The embodiment of the present invention by three-dimensional prediction model trained in advance, can directly by
Two-dimensional video frame image in body-building video is converted to practical three-dimensional (3 D) manikin, so as to according in practical three-dimensional (3 D) manikin
Each key point parameter, determine body-building movement attribute, realize while reducing use cost, ensure that body-building move
The recognition accuracy of the attribute of work.
Detailed description of the invention
Fig. 1 is a kind of flow chart of the attribute determining method for body-building movement that the embodiment of the present invention one provides;
Fig. 2 is a kind of flow chart of three-dimensional prediction model generating method provided in an embodiment of the present invention;
Fig. 3 is the display schematic diagram that a kind of video frame provided in an embodiment of the present invention is converted to practical three-dimensional (3 D) manikin;
Fig. 4 is a kind of display schematic diagram of key point provided in an embodiment of the present invention;
Fig. 5 is the flow chart of the attribute determining method of another body-building movement provided in an embodiment of the present invention;
Fig. 6 is the flow chart of the attribute determining method of another body-building movement provided in an embodiment of the present invention;
Fig. 7 is a kind of structural block diagram of the attribute determining device of body-building movement provided in an embodiment of the present invention;
Fig. 8 is that a kind of attribute of body-building movement provided in an embodiment of the present invention determines the structural schematic diagram of equipment.
Specific embodiment
The present invention is described in further detail with reference to the accompanying drawings and examples.It is understood that this place is retouched
The specific embodiment stated is used only for explaining the present invention rather than limiting the invention.It also should be noted that in order to just
Only the parts related to the present invention are shown in description, attached drawing rather than entire infrastructure.
Fig. 1 is a kind of flow chart of the attribute determining method of body-building movement provided in an embodiment of the present invention, in the present embodiment
The attribute determining method of the body-building movement of offer can determine that equipment executes by the attribute that body-building acts, the attribute of body-building movement
Determine that equipment can be realized by way of software and/or hardware, the body-building movement attribute determine equipment can be two or
Multiple physical entities are constituted, and are also possible to a physical entity and are constituted.The attribute that body-building acts in the present embodiment determines that equipment can
For personal computer (Personal Computer, PC).
With reference to Fig. 1, the attribute determining method of body-building movement specifically comprises the following steps:
S110, the body-building video for obtaining user, extraction obtain the video frame in body-building video.
Wherein, body-building video is the video that the different body-building comprising the same visual angle of user act.In embodiment, often
What a body-building video was made of multiple video frames, also, each video frame is two dimensional image.In actual mechanical process
In, a body-building video may include the corresponding all body-building movements of a body-building type of action, and it is dynamic to may also comprise multiple body-building
Make the corresponding all body-building movements of type.In embodiment, for the ease of being illustrated to the attribute determining method that body-building acts,
To include that the corresponding all body-building movements of a body-building type of action are illustrated in a body-building video.
Meanwhile in the practical operation of body-building video for obtaining user, the body-building video of real-time recording, Ke Yili can be obtained
Xie Wei is acted during user for body-building by the body-building that the mobile terminal configured with camera obtains user in real time, then will
Comprising body-building movement two dimensional image be saved on the mobile terminal, and upload to mobile terminal foundation have communication connection
The attribute of body-building movement determines in equipment.Certainly, the body-building video for having recorded completion can also be directly acquired, it can be understood as,
The body-building video that the user has recorded completion is directly acquired from local storage space or network.Wherein, mobile terminal can be
Smart phone, ipad, laptop etc. are configured with the terminal device of camera.
It should be noted that if when directly acquiring the body-building video for having recorded completion, it need to be right from body-building video
Video frame is extracted, and the attribute that the corresponding two dimensional image of each video frame is uploaded to body-building movement respectively in order is determined
In equipment.Certainly, in the extraction process for carrying out video frame to body-building video, guaranteeing that the attribute for not influencing body-building movement is determining
It, can be in order to reduce the operand that the attribute of body-building movement determines that equipment determines body-building action attributes in the case where accuracy
Video frame is extracted at the extraction interval of setting, for example, being to extract interval to extract body-building video with two frames, to obtain
Corresponding video frame.Certainly, it to interval is extracted without specifically limiting, can be set according to the actual situation.It is to be understood that
Different body-building type of action, between two frame video frame of front and back, the corresponding body-building amplitude in each position of user's body is also different
, in embodiment, it can be set using body-building type of action as foundation to interval is extracted.For example, acting class according to body-building
Type can be divided into deep-knee-bend, sit-ups, push-up, flat support etc..When the corresponding two frame video frame of front and back of body-building type of action it
Between body-building amplitude it is larger when, extract interval occurrence can set it is relatively small;Conversely, when body-building type of action is corresponding
Two frame video frame of front and back between body-building amplitude it is smaller when, the occurrence for extracting interval can set relatively large, can drop
The attribute of low body-building movement determines the operand of equipment.
S120, the three-dimensional prediction model for training video frame input in advance, obtain corresponding practical three-dimensional (3 D) manikin.
Wherein, three-dimensional prediction model is by the standard three-dimensional manikin of respective action in two dimensional image and two dimensional image
Input deep neural network training obtains.In embodiment, Fig. 2 is that a kind of three-dimensional prediction model provided in an embodiment of the present invention is raw
At the flow chart of method, with reference to Fig. 2, the three-dimensional prediction model generating method, it may include step S1201-S1203:
The two dimensional image that the body-building of S1201, acquisition user at same visual angle acts.
Wherein, the two dimensional image of the body-building movement at same visual angle, it can be understood as by the mobile end for being configured with camera
The body-building movement to different user in the same angle is held to carry out two dimensional image obtained from shooting, collecting.In embodiment, it is
The forecasting accuracy for guaranteeing the three-dimensional prediction model that training obtains, can obtain a large amount of two dimensional image, and can obtain difference
The two dimensional image that body-building of the user at same visual angle of figure, different height acts.Wherein, the X-Y scheme of the body-building movement of user
As several visual angles can be divided into, for example, positive angle, depression angle, side view angle, looking up angle etc..In embodiment, in order to
Guarantee acquired two dimensional image, can accurately show corresponding body-building movement, can be adopted by the determination of body-building type of action
Collect the visual angle of two dimensional image.It is non-limiting as example, it is deep-knee-bend for body-building type of action, it can be directly from positive angle to strong
The two dimensional image of body movement is acquired;It is flat support for body-building type of action, it can be directly dynamic to body-building from side view angle
The two dimensional image of work is acquired;For other body-building type of action, can be set as the case may be.
Certainly, in order to guarantee the accuracy for training obtained three-dimensional prediction model according to collected two dimensional image, right
Three-dimensional prediction model is trained, and the body-building movement in the two dimensional image acquired is standard, also, can set two dimensional image
In body-building movement standard degree cannot be below preset standard degree, if be lower than preset standard degree, abandon the two dimensional image.
S1202, corresponding standard three-dimensional manikin is found according to the body-building movement in two dimensional image.
Wherein, standard three-dimensional manikin can be understood as the corresponding three-dimensional (3 D) manikin of standard body-building movement.In reality
It applies in example, after getting the two dimensional image that same body-building of the different user at same visual angle acts, according in two dimensional image
Body-building movement find the standard three-dimensional manikin of corresponding body-building movement.Wherein, standard three-dimensional manikin can according to
The different building shape at family and different heights are set.
S1203, two dimensional image and standard three-dimensional manikin input deep neural network are trained, are obtained corresponding
Three-dimensional prediction model.
Wherein, the working principle of deep neural network is to imitate human brain form of thinking, so that speech recognition speed is more
Fastly, recognition accuracy is also higher.In embodiment, by the two dimensional image acted comprising standard body-building and corresponding standard body-building movement
Standard three-dimensional manikin inputted respectively as training sample in the model of deep neural network, and pass through deep neural network
Model it is trained, corresponding three-dimensional prediction model can be obtained.Wherein, three-dimensional prediction model is got for basis
Any video frame comprising body-building movement in prediction obtain active user and correspond to the practical three-dimensional (3 D) manikin that body-building acts, with
Solve the problems, such as that two dimensional image lacks depth information.
It in embodiment, can be straight by the corresponding two dimensional image of video frame after obtaining the video frame comprising body-building movement
Input three-dimensional prediction model trained in advance is connect, can be obtained by the corresponding two dimensional image of video frame comprising corresponding body-building movement
Practical three-dimensional (3 D) manikin.Fig. 3 is that a kind of video frame provided in an embodiment of the present invention is converted to practical three-dimensional (3 D) manikin
Display schematic diagram.With reference to shown in the left figure of Fig. 3, Fig. 3 left figure is the frame two dimensional image extracted from body-building video, that is, is regarded
Frequency frame can be corresponded to after video frame input three-dimensional prediction model trained in advance according to three-dimensional prediction model prediction
Practical three-dimensional (3 D) manikin, i.e. the practical three-dimensional (3 D) manikin as shown in Fig. 3 right figure.
S130, according to the parameter of key point each in practical three-dimensional (3 D) manikin, determine the actual motion parameter of user.
Wherein, the parameter of each key point can be understood as the three-dimensional coordinate of each artis of user body parts.In reality
It applies in example, it, can be according to the fixation position of artis from each practical 3 D human body mould after obtaining practical three-dimensional (3 D) manikin
The three-dimensional coordinate of each artis is got in type, and the actual motion parameter of user is determined according to the parameter of each key point.Its
In, kinematic parameter includes at least one of following: motion profile, each key point relative position variation and each key point relative angle
Degree variation.In embodiment, actual motion parameter can be understood as actual motion track, the practical relative position of each key point becomes
Change and the practical relative angle of each key point changes.Wherein, actual motion track can be understood as each artis of user
Practical motion track;The practical relative position variation of each key point can be understood as the actual bit between each artis of user
The relative difference set;The practical relative angle variation of each key point can be understood as the actual corners between each artis of user
The relative difference of degree.Certainly, this programme is illustratively with the relative position variation of motion profile, each key point and each key
Kinematic parameter is illustrated for the relative angle variation of point.In the practical operation that the attribute acted to body-building determines,
Other kinematic parameters can be used to determine body-building movement progress attribute, be not limited thereto.Or passing through movement ginseng
It, only need to be dynamic to body-building by at least one of which parameter in kinematic parameter during several pairs of body-building movement progress attributes determine
Make progress attribute to determine.
It should be noted that the actual motion parameter of user corresponding body-building in several frame video frames need to be acted into
Row analysis, can just determine to obtain.It is to be understood that in the actual mechanical process of actual motion parameter for determining user,
It obtains after the parameter of each key point, continuing to obtain next frame video frame in practical three-dimensional (3 D) manikin or extracts the video at interval
Then the parameter of each key point in the corresponding practical three-dimensional (3 D) manikin of frame determines the actual motion track of user, practical phase again
To the data information of change in location and the variation of practical relative angle.
Fig. 4 is a kind of display schematic diagram of key point provided in an embodiment of the present invention.As shown in the left figure in Fig. 4, user
A body movement in the wrong is being done, this is bent into the corresponding two dimensional image of body movement and is input to three-dimensional prediction model, correspondence can be obtained
Practical three-dimensional (3 D) manikin, and the three-dimensional coordinate for extracting each key point, such as Fig. 4 can be identified from practical three-dimensional (3 D) manikin
Each key point is shown shown in right figure.For example, showing 15 key points in figure as shown in right in Figure 4, respectively
For key point 1, key point 2, key point 3 ... key point 15.Wherein, each key point, that is, different artis, for example, key point
1 is head, and key point 2 is neck, and key point 3 is left shoulder, and key point 4 is left elbow, and key point 5 is left wrist, and key point 6 is the right side
Shoulder, key point 7 are right elbow, and key point 8 is right wrist, and key point 9 is abdomen, and key point 10 is left buttocks, and key point 11 is left knee
Lid, key point 12 are left foot point, and key point 13 is right hips, and key point 14 is right knee, and key point 15 is right crus of diaphragm point.Certainly,
Body-building movement attribute determine actual mechanical process in, in order to improve body-building movement attribute recognition accuracy, with
During the body-building of family, user can be shown on the mobile terminal for acquiring image and does the duration acted, for example, user does
Body movement in the wrong, maintains three seconds, then the prompt letter of " you body in the wrong 3 seconds " is shown on the display screen of mobile terminal
Breath.Certainly, preparing, standing etc. in various processes, can show that " you have had been prepared for N on the display screen of mobile terminal
The prompt informations such as second " or " you have stood N seconds ".
S140, according to actual motion parameter and standard movement parameter, determine the attribute that body-building in body-building video acts.
Wherein, standard movement parameter can be understood as the corresponding kinematic parameter of standard body-building movement.In embodiment, exist
After obtaining the actual motion parameter of user, the actual motion parameter of user and the body-building are directly acted into corresponding standard movement
Parameter is compared, according to whether actual motion parameter and standard movement parameter consistent or actual motion parameter and mark
Difference between quasi-moving parameter determines the attribute that body-building acts in body-building video whether within the scope of preset difference value.Wherein, it is good for
The attribute of body movement includes at least one of following: standard degree, the type of body-building movement of body-building movement.In embodiment, if it is real
Difference between border kinematic parameter and standard movement parameter is within the scope of preset difference value, it is determined that body-building acts in body-building video
Standard degree reaches standard degree threshold value, i.e. the body-building movement of the user is standard, also, actual motion parameter and standard movement are joined
Body-building movement corresponding to difference minimum between number is determined as the type of body-building movement;Conversely, if actual motion parameter and
Difference between standard movement parameter is not within the scope of preset difference value, it is determined that the standard degree that body-building acts in body-building video does not reach
Body-building movement to standard degree threshold value, the i.e. user is non-type.
It should be noted that can use in the practical operation that attribute determines with real-time recording acting body-building
Then body-building movement in family acts progress attribute to the body-building in real time and determines.It is to be understood that passing through during user for body-building
Mobile terminal configured with camera records the body-building movement of user, and after recording a frame body-building image, this is good for
Body image is input to three-dimensional prediction model trained in advance, strong by what is recorded again after then recording a frame body-building image again
Body image is input to three-dimensional prediction model trained in advance, and so on, after user completes body-building, user can be directly obtained
Body-building movement whether standard as a result, and body-building movement standard degree occurrence.Certainly, it can also complete to user for body-building
The recording of movement and then uniformly in the body-building video body-building movement carry out attribute determine.
The technical solution of the present embodiment, by obtaining the body-building video of user, extraction obtains the video frame in body-building video;
By video frame input three-dimensional prediction model trained in advance, corresponding practical three-dimensional (3 D) manikin is obtained, three-dimensional prediction model is
The standard three-dimensional manikin input deep neural network training of respective action in two dimensional image and two dimensional image is obtained;Root
Factually in the three-dimensional (3 D) manikin of border each key point parameter, determine the actual motion parameter of user;According to actual motion parameter and
Standard movement parameter determines the attribute that body-building acts in body-building video, realizes while reducing use cost, ensure that strong
The recognition accuracy of the attribute of body movement.
On the basis of the above embodiments, the attribute of body-building movement includes at least one of following: the standard degree of body-building movement,
The type of body-building movement.In order to determine the standard degree of body-building movement, according to actual motion parameter and standard movement parameter, determine strong
The attribute that body-building acts in body video, specifically: according to actual motion parameter and standard movement parameter, determines in body-building video and be good for
The standard degree of body movement.
In embodiment, if the actual motion parameter of user and standard movement parameter show to be good within the scope of preset difference value
The standard degree that body-building acts in body video reaches standard degree threshold value, i.e. the body-building movement of the user is standard;Conversely, if practical
Difference between kinematic parameter and standard movement parameter then shows body-building movement in body-building video not within the scope of preset difference value
Standard degree is not up to standard degree threshold value, i.e. the body-building movement of the user is non-type.
On the basis of the above embodiments, it in order to determine the type of body-building movement, is transported according to actual motion parameter and standard
Dynamic parameter, determines the attribute that body-building acts in body-building video, specifically: according to actual motion parameter and standard movement parameter, really
Determine the type that body-building acts in body-building video.
In embodiment, after obtaining actual motion parameter, actual motion parameter and standard movement parameter are compared
It is body-building view by the smallest corresponding body-building movement of the difference between actual motion parameter and standard movement parameter to analysis
The type for the body-building movement that user is done in frequency.Illustratively, it is assumed that the actual motion parameter of user and the standard movement of deep-knee-bend
Difference between parameter is respectively 12,14,20;Difference between the standard movement parameter of sit-ups is respectively 60,70,
90, and so on, corresponding standard movement parameter is acted with other types of body-building and is compared, and the determining standard with deep-knee-bend is transported
Difference between dynamic parameter is minimum, then can determine that the type that body-building acts in body-building video is deep-knee-bend.Certainly, determining that body-building is dynamic
When the type of work, the type that body-building acts can also be divided into strength type, beautifying, body building type etc., be not limited thereto.?
In the practical operation for determining the type of body-building movement, when the type of identified body-building movement is with training three-dimensional prediction model in advance
The type of pre-set body-building movement is related.It is non-limiting as example, it is assumed that two dimensional image and standard three-dimensional human mould
During type is trained, barbell will be used to complete deep-knee-bend and be set as power-type sports, then according to actual motion parameter and
Standard movement parameter determines that the type of body-building movement is strength type, and not deep-knee-bend and act barbell.
Fig. 5 is the flow chart of the attribute determining method of another body-building movement provided in an embodiment of the present invention.The present embodiment
It is on the basis of the above embodiments, further embody to be made to the attribute determining method of body-building movement.Referring to Fig. 5, this is strong
The attribute determining method of body movement specifically comprises the following steps:
S210, the body-building video for obtaining user, extraction obtain the video frame in body-building video.
S220, the three-dimensional prediction model for training video frame input in advance, obtain corresponding practical three-dimensional (3 D) manikin.
Wherein, three-dimensional prediction model is by the standard three-dimensional manikin of respective action in two dimensional image and two dimensional image
Input deep neural network training obtains.
S230, corresponding practical three-dimensional (3 D) manikin is formed into practical three-dimensional (3 D) manikin according to the extraction sequence of video frame
Sequence.
Wherein, practical three-dimensional (3 D) manikin sequence is by the corresponding each practical three-dimensional (3 D) manikin institute group of each video frame
At sequence.It should be noted that each two dimensional image can be corresponding with a three-dimensional (3 D) manikin.In embodiment,
What each body-building video was made of several video frames, i.e., a body-building video corresponds to multiple practical three-dimensional (3 D) manikins,
It, can after getting body-building video in order to which the attribute acted by practical three-dimensional (3 D) manikin to body-building is accurately determined
Multiple video frames are extracted from body-building video to extract interval, and each video frame is obtained by three-dimensional prediction model prediction and is corresponded to
Practical three-dimensional (3 D) manikin, then, according to extracting the sequence of video frame for all reality three corresponding in the body-building video
Dimension manikin rearranges practical three-dimensional (3 D) manikin sequence.
S240, the reality three for extracting each key point in each practical three-dimensional (3 D) manikin in practical three-dimensional (3 D) manikin sequence
Tie up coordinate.
Wherein, the practical three-dimensional coordinate of each key point refers to practical three of each artis on each position of user's body
Tie up coordinate.In embodiment, by three-dimensional prediction model prediction obtain the corresponding practical three-dimensional (3 D) manikin of each video frame it
It afterwards, can be according to the practical three-dimensional coordinate of each artis of fixation position acquisition of artis, for example, coordinate corresponding to x, y and z
Value.Certainly, the body-building movement in body-building video is real-time change, i.e., practical three-dimensional people corresponding to two adjacent video frames
Body Model may also be different, the reality between each artis extracted in two adjacent practical three-dimensional (3 D) manikins
Border three-dimensional coordinate is also different.
S250, combination is ranked up to the practical three-dimensional coordinate of each key point according to the extraction sequence of video frame, generated real
Border three-dimensional coordinate sequence.
In embodiment, it extracts to obtain the practical three-dimensional coordinate of corresponding each artis from each practical three-dimensional (3 D) manikin
Later, in order to be accurately determined to the attribute of body-building movement, need to successively will first according to the extraction sequence of video frame
The practical three-dimensional coordinate of each artis is ranked up combination, generates practical three-dimensional coordinate sequence.
S260, the actual motion parameter that user is determined according to practical three-dimensional coordinate sequence.
It in embodiment, can be according to two phases in practical three-dimensional coordinate sequence after obtaining practical three-dimensional coordinate sequence
The occurrence of adjacent practical three-dimensional coordinate, determines the actual motion track of user;And according to two in practical three-dimensional coordinate sequence
The difference of a adjacent practical three-dimensional coordinate determines relative position variation and the relative angle of each key point of each key point
Variation.For example, during doing deep-knee-bend, according to the practical three-dimensional coordinate of the practical three-dimensional coordinate of knee and tiptoe, it may be determined that
The actual motion track of knee, the change of the situation of change of relative distance and knee itself relative angle between knee and tiptoe
Change situation.
Certainly, in the actual motion parameter for determining user, only according to the occurrences of two adjacent practical three-dimensional coordinates,
The actual motion parameter of user can not be accurately determined, it can be understood as, it need to be successively to two phases corresponding to the body-building video
Adjacent practical three-dimensional coordinate is compared, to determine the actual motion parameter of user.
S270, according to actual motion parameter and standard movement parameter, determine the attribute that body-building in body-building video acts.
In embodiment, in the variation of practical relative position and each key point for obtaining actual motion track, each key point
After practical relative angle variation, directly each actual motion parameter and corresponding standard movement parameter are compared, if real
Border kinematic parameter and corresponding standard movement parameter within the error range, determine that body-building movement is to comply with standard in body-building video
, i.e., standard degree reaches standard degree threshold value, and can determine the type of body-building movement.
The technical solution of the present embodiment is further advanced by the basis of above scheme and extracts practical 3 D human body mould
In type sequence in each practical three-dimensional (3 D) manikin each key point practical three-dimensional coordinate;According to the extraction sequence of video frame to each
The practical three-dimensional coordinate of key point is ranked up combination, generates practical three-dimensional coordinate sequence;It is true according to practical three-dimensional coordinate sequence
The actual motion parameter for determining user, realize in determining body-building video body-building movement type and body-building movement whether standard
Effect.
Fig. 6 is the flow chart of the attribute determining method of another body-building movement provided in an embodiment of the present invention.The present embodiment
In fact on the basis of the above embodiments, as a preferred embodiment, the attribute determining method of body-building movement is specifically described.Ginseng
Fig. 6 is examined, the attribute determining method of body-building movement specifically comprises the following steps:
S310, the body-building video for obtaining user.
S320, body-building video is carried out to take out frame processing, obtains video frame.
In embodiment, take out frame processing mode can there are many, for example, can by way of extracting frame by frame to body-building regard
Frequency carries out taking out frame processing, can also be carrying out taking out frame processing to body-building video in a manner of extracting interval.Wherein, it extracts frame by frame
Mode can be understood as the mode that a frame is divided between extracting.
S330, the three-dimensional prediction model for training video frame input in advance, obtain corresponding practical three-dimensional (3 D) manikin.
S340, practical three-dimensional (3 D) manikin is formed into practical three-dimensional (3 D) manikin sequence according to the extraction sequence of video frame.
S350, the reality three for extracting each key point in each practical three-dimensional (3 D) manikin in practical three-dimensional (3 D) manikin sequence
Tie up coordinate.
S360, combination is ranked up to the practical three-dimensional coordinate of each key point according to the extraction sequence of video frame, generated real
Border three-dimensional coordinate sequence.
S370, the actual motion parameter that user is determined according to practical three-dimensional coordinate sequence.
Whether S380, actual motion parameter are within the error range of standard movement parameter, if so, thening follow the steps
S390;If it is not, thening follow the steps S3100.
S390, determine that the type of body-building movement and body-building movement are standards in body-building video.
S3100, determine that the type of body-building movement and body-building movement are non-type in body-building video.
The technical solution of the present embodiment, the view extracted from body-building video by three-dimensional prediction model prediction trained in advance
The corresponding practical three-dimensional (3 D) manikin comprising identical body-building movement of frequency frame, and extracted from practical three-dimensional (3 D) manikin and obtain reality
Border three-dimensional coordinate, and determine according to practical three-dimensional coordinate the actual motion parameter of user, then according to actual motion parameter and mark
Quasi-moving parameter determines the standard degree that body-building acts in body-building video, solves missing artis depth information in two dimensional image,
And the problem that depth camera is at high cost and popularity rate is low, it realizes while reducing use cost, ensure that body-building acts
Whether standard and determine body-building movement type recognition accuracy.
Fig. 7 is a kind of structural block diagram of the attribute determining device of body-building movement provided in an embodiment of the present invention.The present embodiment
The attribute determining device of body-building movement be configured in server, with reference to Fig. 7, the attribute determining device packet of body-building movement
It includes: obtaining module 410, the first determining module 420, the second determining module 430 and third determining module 440.
Wherein, module 410 is obtained, for obtaining the body-building video of user, extraction obtains the video frame in body-building video;
First determining module 420 obtains corresponding reality for the three-dimensional prediction model that video frame input is trained in advance
Three-dimensional (3 D) manikin, the three-dimensional prediction model are by the standard three-dimensional human mould of respective action in two dimensional image and two dimensional image
Type input deep neural network training obtains;
Second determining module 430 determines the reality of user for the parameter according to key point each in practical three-dimensional (3 D) manikin
Border kinematic parameter;
Third determining module 440, for determining body-building in body-building video according to actual motion parameter and standard movement parameter
The attribute of movement.
Technical solution provided in this embodiment, by obtaining the body-building video of user, extraction obtains the view in body-building video
Frequency frame;By video frame input three-dimensional prediction model trained in advance, corresponding practical three-dimensional (3 D) manikin, three-dimensional prediction mould are obtained
Type is that the standard three-dimensional manikin of respective action in two dimensional image and two dimensional image is inputted deep neural network is trained
It arrives;According to the parameter of key point each in practical three-dimensional (3 D) manikin, the actual motion parameter of user is determined;Joined according to actual motion
Several and standard movement parameter determines the attribute that body-building acts in body-building video, realizes while reducing use cost, guarantees
The recognition accuracy of the attribute of body-building movement.
On the basis of the above embodiments, the attribute of body-building movement includes at least one of following: the standard degree of body-building movement,
The type of body-building movement.
On the basis of the above embodiments, three-dimensional prediction model is corresponding will to act in two dimensional image and two dimensional image
The input deep neural network training of standard three-dimensional manikin obtains, and is specifically used for:
Acquire the two dimensional image that body-building of the user at same visual angle acts;
Corresponding standard three-dimensional manikin is found according to the body-building movement in two dimensional image;
Two dimensional image and standard three-dimensional manikin input deep neural network are trained, obtained corresponding three-dimensional pre-
Survey model.
On the basis of the above embodiments, described device, further includes:
Comprising modules, for obtaining corresponding practical three-dimensional in the three-dimensional prediction model that video frame input is trained in advance
After manikin, corresponding practical three-dimensional (3 D) manikin is formed into practical three-dimensional (3 D) manikin according to the extraction sequence of video frame
Sequence.
On the basis of the above embodiments, second determining module, comprising:
Extraction unit, for extracting in practical three-dimensional (3 D) manikin sequence each key point in each practical three-dimensional (3 D) manikin
Practical three-dimensional coordinate;
Combination producing unit is ranked up the practical three-dimensional coordinate of each key point for the extraction sequence according to video frame
Combination generates practical three-dimensional coordinate sequence;
Determination unit, for determining the actual motion parameter of user according to practical three-dimensional coordinate sequence.
On the basis of the above embodiments, kinematic parameter includes at least one of following: motion profile, each key point it is opposite
The variation of the relative angle of change in location and each key point.
The category of body-building movement provided by any embodiment of the invention can be performed in the attribute determining device of above-mentioned body-building movement
Property determines method, has the corresponding functional module of execution method and beneficial effect.
Fig. 8 is that a kind of attribute of body-building movement provided in an embodiment of the present invention determines the structural schematic diagram of equipment.With reference to figure
8, the attribute of body-building movement determines that equipment includes: processor 510, memory 520, input unit 530 and output device
540.The attribute of body-building movement determines that the quantity of processor 510 in equipment can be one or more, at one in Fig. 8
For reason device 510.The attribute of body-building movement determines that the quantity of memory 520 in equipment can be one or more, in Fig. 8
By taking a memory 520 as an example.The attribute of body-building movement determines the processor 510, memory 520, input unit 530 of equipment
And output device 540 can be connected by bus or other modes, in Fig. 8 for being connected by bus.In embodiment,
The attribute of body-building movement determines that equipment can be PC machine.
Memory 520 is used as a kind of computer readable storage medium, can be used for storing software program, journey can be performed in computer
The attribute of sequence and module, the body-building movement as described in any embodiment of that present invention determines the corresponding program instruction/module of equipment
(for example, acquisition module 410, the first determining module 420,430 and of the second determining module in the attribute determining device of body-building movement
Third determining module 440).Memory 520 can mainly include storing program area and storage data area, wherein storing program area can
Application program needed for storage program area, at least one function;Storage data area can be stored to be created according to using for equipment
Data etc..In addition, memory 520 may include high-speed random access memory, it can also include nonvolatile memory, example
Such as at least one disk memory, flush memory device or other non-volatile solid state memory parts.In some instances, it stores
Device 520 can further comprise the memory remotely located relative to processor 510, these remote memories can be connected by network
It is connected to equipment.The example of above-mentioned network includes but is not limited to internet, intranet, local area network, mobile radio communication and its group
It closes.
Input unit 530 can be used for receiving the number or character information of input, and generate the user setting with equipment
And the related key signals input of function control, it can also be the camera for obtaining image and obtain picking up for audio data
Sound equipment.Output device 540 may include the audio frequency apparatuses such as loudspeaker.It should be noted that input unit 530 and output device
540 concrete composition may be set according to actual conditions.
Software program, instruction and the module that processor 510 is stored in memory 520 by operation, thereby executing setting
Standby various function application and data processing realizes the attribute determining method of above-mentioned body-building movement.
It is dynamic that the attribute of the body-building movement of above-mentioned offer determines that equipment can be used for executing the body-building that above-mentioned any embodiment provides
The attribute determining method of work has corresponding function and beneficial effect.
The embodiment of the present invention also provides a kind of storage medium comprising computer executable instructions, and the computer is executable
It instructs when being executed by computer processor for executing a kind of attribute determining method of body-building movement, comprising:
The body-building video of user is obtained, extraction obtains the video frame in body-building video;
By video frame input three-dimensional prediction model trained in advance, corresponding practical three-dimensional (3 D) manikin is obtained, described three
Tieing up prediction model is that the standard three-dimensional manikin of respective action in two dimensional image and two dimensional image is inputted depth nerve net
Network training obtains;
According to the parameter of key point each in practical three-dimensional (3 D) manikin, the actual motion parameter of user is determined;
According to actual motion parameter and standard movement parameter, the attribute that body-building acts in body-building video is determined.
Certainly, a kind of storage medium comprising computer executable instructions, computer provided by the embodiment of the present invention
Any implementation of the invention can also be performed in the attribute determining method operation for the body-building movement that executable instruction is not limited to the described above
Relevant operation in the attribute determining method of the movement of body-building provided by example, and have corresponding function and beneficial effect.
By the description above with respect to embodiment, it is apparent to those skilled in the art that, the present invention
It can be realized by software and required common hardware, naturally it is also possible to which by hardware realization, but in many cases, the former is more
Good embodiment.Based on this understanding, technical solution of the present invention substantially in other words contributes to the prior art
Part can be embodied in the form of software products, which can store in computer readable storage medium
In, floppy disk, read-only memory (Read-Only Memory, ROM), random access memory (Random such as computer
Access Memory, RAM), flash memory (FLASH), hard disk or CD etc., including some instructions are with so that a computer is set
Standby (can be robot, personal computer, server or the network equipment etc.) executes and is good for described in any embodiment of that present invention
The attribute determining method of body movement.
It is worth noting that, included each unit and module are in the attribute determining device of above-mentioned body-building movement
It is divided according to the functional logic, but is not limited to the above division, as long as corresponding functions can be realized;Separately
Outside, the specific name of each functional unit is also only for convenience of distinguishing each other, the protection scope being not intended to restrict the invention.
It should be appreciated that each section of the invention can be realized with hardware, software, firmware or their combination.Above-mentioned
In embodiment, software that multiple steps or method can be executed in memory and by suitable instruction execution system with storage
Or firmware is realized.It, and in another embodiment, can be under well known in the art for example, if realized with hardware
Any one of column technology or their combination are realized: having a logic gates for realizing logic function to data-signal
Discrete logic, with suitable combinational logic gate circuit specific integrated circuit, programmable gate array (PGA), scene
Programmable gate array (FPGA) etc..
In the description of this specification, reference term " one embodiment ", " some embodiments ", " example ", " specifically show
The description of example " or " some examples " etc. means specific features, structure, material or spy described in conjunction with this embodiment or example
Point is included at least one embodiment or example of the invention.In the present specification, schematic expression of the above terms are not
Centainly refer to identical embodiment or example.Moreover, particular features, structures, materials, or characteristics described can be any
One or more embodiment or examples in can be combined in any suitable manner.
Note that the above is only a better embodiment of the present invention and the applied technical principle.It will be appreciated by those skilled in the art that
The invention is not limited to the specific embodiments described herein, be able to carry out for a person skilled in the art it is various it is apparent variation,
It readjusts and substitutes without departing from protection scope of the present invention.Therefore, although being carried out by above embodiments to the present invention
It is described in further detail, but the present invention is not limited to the above embodiments only, without departing from the inventive concept, also
It may include more other equivalent embodiments, and the scope of the invention is determined by the scope of the appended claims.
Claims (10)
1. a kind of attribute determining method of body-building movement characterized by comprising
The body-building video of user is obtained, extraction obtains the video frame in the body-building video;
By video frame input three-dimensional prediction model trained in advance, corresponding practical three-dimensional (3 D) manikin is obtained, described three
Tieing up prediction model is that the standard three-dimensional manikin of respective action in two dimensional image and two dimensional image is inputted depth nerve net
Network training obtains;
According to the parameter of each key point in the practical three-dimensional (3 D) manikin, the actual motion parameter of user is determined;
According to the actual motion parameter and standard movement parameter, the attribute that body-building acts in the body-building video is determined.
2. the attribute determining method of body-building movement according to claim 1, which is characterized in that the attribute of the body-building movement
Including at least one of following: standard degree, the type of body-building movement of body-building movement.
3. the attribute determining method of body-building according to claim 1 movement, which is characterized in that the three-dimensional prediction model is
Corresponding standard three-dimensional manikin input deep neural network training will be acted in two dimensional image and two dimensional image to obtain, and is wrapped
It includes:
Acquire the two dimensional image that body-building of the user at same visual angle acts;
Corresponding standard three-dimensional manikin is found according to the body-building movement in the two dimensional image;
The two dimensional image and standard three-dimensional manikin input deep neural network are trained, obtain corresponding three
Tie up prediction model.
4. the attribute determining method of body-building movement according to claim 1, which is characterized in that inputted by the video frame
Three-dimensional prediction model trained in advance, after obtaining corresponding practical three-dimensional (3 D) manikin, further includes:
Corresponding practical three-dimensional (3 D) manikin is formed into practical three-dimensional (3 D) manikin sequence according to the extraction sequence of video frame.
5. the attribute determining method of body-building movement according to claim 3, which is characterized in that described according to the reality three
The parameter for tieing up each key point in manikin, determines the actual motion parameter of user, comprising:
Extract the reality three of each key point in each practical three-dimensional (3 D) manikin in the practical three-dimensional (3 D) manikin sequence
Tie up coordinate;
Combination is ranked up to the practical three-dimensional coordinate of each key point according to the extraction sequence of video frame, generates practical three-dimensional coordinate
Sequence;
The actual motion parameter of user is determined according to the practical three-dimensional coordinate sequence.
6. the attribute determining method of body-building movement according to claim 1, which is characterized in that under the kinematic parameter includes
State at least one: the relative angle of motion profile, the relative position variation of each key point and each key point changes.
7. a kind of attribute determining device of body-building movement characterized by comprising
Module is obtained, for obtaining the body-building video of user, extraction obtains the video frame in the body-building video;
First determining module obtains corresponding reality three for the three-dimensional prediction model that video frame input is trained in advance
Manikin is tieed up, the three-dimensional prediction model is by the standard three-dimensional human mould of respective action in two dimensional image and two dimensional image
Type input deep neural network training obtains;
Second determining module determines the reality of user for the parameter according to each key point in the practical three-dimensional (3 D) manikin
Kinematic parameter;
Third determining module is good for for according to the actual motion parameter and standard movement parameter, determining in the body-building video
The attribute of body movement.
8. the attribute determining device of body-building movement according to claim 7, which is characterized in that further include:
Comprising modules, for obtaining corresponding practical three-dimensional in the three-dimensional prediction model that video frame input is trained in advance
After manikin, corresponding practical three-dimensional (3 D) manikin is formed into practical three-dimensional (3 D) manikin according to the extraction sequence of video frame
Sequence.
9. a kind of attribute of body-building movement determines equipment characterized by comprising memory and one or more processors;
The memory, for storing one or more programs;
When one or more of programs are executed by one or more of processors, so that one or more of processors are real
Now such as the attribute determining method of body-building as claimed in any one of claims 1 to 6 movement.
10. a kind of storage medium comprising computer executable instructions, which is characterized in that the computer executable instructions by
For executing the attribute determining method such as body-building as claimed in any one of claims 1 to 6 movement when computer processor executes.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811638461.6A CN109621332A (en) | 2018-12-29 | 2018-12-29 | A kind of attribute determining method, device, equipment and the storage medium of body-building movement |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811638461.6A CN109621332A (en) | 2018-12-29 | 2018-12-29 | A kind of attribute determining method, device, equipment and the storage medium of body-building movement |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109621332A true CN109621332A (en) | 2019-04-16 |
Family
ID=66054656
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811638461.6A Pending CN109621332A (en) | 2018-12-29 | 2018-12-29 | A kind of attribute determining method, device, equipment and the storage medium of body-building movement |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109621332A (en) |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110170159A (en) * | 2019-06-27 | 2019-08-27 | 郭庆龙 | A kind of human health's action movement monitoring system |
CN110782482A (en) * | 2019-10-21 | 2020-02-11 | 深圳市网心科技有限公司 | Motion evaluation method and device, computer equipment and storage medium |
CN111105347A (en) * | 2019-11-19 | 2020-05-05 | 贝壳技术有限公司 | Method, device and storage medium for generating panoramic image with depth information |
CN111275031A (en) * | 2020-05-07 | 2020-06-12 | 西南交通大学 | Flat plate support detection method, device, equipment and medium based on human body key points |
CN112233770A (en) * | 2020-10-15 | 2021-01-15 | 郑州师范学院 | Intelligent gymnasium management decision-making system based on visual perception |
CN112237730A (en) * | 2019-07-17 | 2021-01-19 | 腾讯科技(深圳)有限公司 | Body-building action correcting method and electronic equipment |
CN112287868A (en) * | 2020-11-10 | 2021-01-29 | 上海依图网络科技有限公司 | Human body action recognition method and device |
US11055835B2 (en) | 2019-11-19 | 2021-07-06 | Ke.com (Beijing) Technology, Co., Ltd. | Method and device for generating virtual reality data |
CN113657266A (en) * | 2021-08-16 | 2021-11-16 | 江苏动泰运动用品有限公司 | Fitness training management method and system based on intelligent bracelet and human body three-dimensional reconstruction |
CN114356100A (en) * | 2022-01-13 | 2022-04-15 | 上海健指树智能系统有限公司 | Body-building action guiding method and device, electronic equipment and storage medium |
CN114827730A (en) * | 2022-04-19 | 2022-07-29 | 咪咕文化科技有限公司 | Video cover selecting method, device, equipment and storage medium |
WO2023040449A1 (en) * | 2021-09-17 | 2023-03-23 | 支付宝(杭州)信息技术有限公司 | Triggering of client operation instruction by using fitness action |
WO2023061269A1 (en) * | 2021-10-15 | 2023-04-20 | 北京京东方技术开发有限公司 | Fitness plan information generation method, apparatus and system |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104952104A (en) * | 2014-03-26 | 2015-09-30 | 联想(北京)有限公司 | Three-dimensional human body gesture estimating method and device thereof |
CN107293162A (en) * | 2017-07-31 | 2017-10-24 | 广东欧珀移动通信有限公司 | Move teaching auxiliary and device, terminal device |
CN107943291A (en) * | 2017-11-23 | 2018-04-20 | 乐蜜有限公司 | Recognition methods, device and the electronic equipment of human action |
CN108921926A (en) * | 2018-07-02 | 2018-11-30 | 广州云从信息科技有限公司 | A kind of end-to-end three-dimensional facial reconstruction method based on single image |
EP3549645A1 (en) * | 2007-11-05 | 2019-10-09 | Brian Francis Mooney | Method and system for analysing a golf swing |
-
2018
- 2018-12-29 CN CN201811638461.6A patent/CN109621332A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP3549645A1 (en) * | 2007-11-05 | 2019-10-09 | Brian Francis Mooney | Method and system for analysing a golf swing |
CN104952104A (en) * | 2014-03-26 | 2015-09-30 | 联想(北京)有限公司 | Three-dimensional human body gesture estimating method and device thereof |
CN107293162A (en) * | 2017-07-31 | 2017-10-24 | 广东欧珀移动通信有限公司 | Move teaching auxiliary and device, terminal device |
CN107943291A (en) * | 2017-11-23 | 2018-04-20 | 乐蜜有限公司 | Recognition methods, device and the electronic equipment of human action |
CN108921926A (en) * | 2018-07-02 | 2018-11-30 | 广州云从信息科技有限公司 | A kind of end-to-end three-dimensional facial reconstruction method based on single image |
Cited By (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110170159A (en) * | 2019-06-27 | 2019-08-27 | 郭庆龙 | A kind of human health's action movement monitoring system |
CN112237730A (en) * | 2019-07-17 | 2021-01-19 | 腾讯科技(深圳)有限公司 | Body-building action correcting method and electronic equipment |
CN110782482A (en) * | 2019-10-21 | 2020-02-11 | 深圳市网心科技有限公司 | Motion evaluation method and device, computer equipment and storage medium |
US11055835B2 (en) | 2019-11-19 | 2021-07-06 | Ke.com (Beijing) Technology, Co., Ltd. | Method and device for generating virtual reality data |
CN111105347B (en) * | 2019-11-19 | 2020-11-13 | 贝壳找房(北京)科技有限公司 | Method, device and storage medium for generating panoramic image with depth information |
CN111105347A (en) * | 2019-11-19 | 2020-05-05 | 贝壳技术有限公司 | Method, device and storage medium for generating panoramic image with depth information |
US11721006B2 (en) | 2019-11-19 | 2023-08-08 | Realsee (Beijing) Technology Co., Ltd. | Method and device for generating virtual reality data |
CN111275031A (en) * | 2020-05-07 | 2020-06-12 | 西南交通大学 | Flat plate support detection method, device, equipment and medium based on human body key points |
CN112233770A (en) * | 2020-10-15 | 2021-01-15 | 郑州师范学院 | Intelligent gymnasium management decision-making system based on visual perception |
CN112233770B (en) * | 2020-10-15 | 2023-07-07 | 郑州师范学院 | Gymnasium intelligent management decision-making system based on visual perception |
CN112287868A (en) * | 2020-11-10 | 2021-01-29 | 上海依图网络科技有限公司 | Human body action recognition method and device |
CN113657266A (en) * | 2021-08-16 | 2021-11-16 | 江苏动泰运动用品有限公司 | Fitness training management method and system based on intelligent bracelet and human body three-dimensional reconstruction |
WO2023040449A1 (en) * | 2021-09-17 | 2023-03-23 | 支付宝(杭州)信息技术有限公司 | Triggering of client operation instruction by using fitness action |
WO2023061269A1 (en) * | 2021-10-15 | 2023-04-20 | 北京京东方技术开发有限公司 | Fitness plan information generation method, apparatus and system |
CN114356100A (en) * | 2022-01-13 | 2022-04-15 | 上海健指树智能系统有限公司 | Body-building action guiding method and device, electronic equipment and storage medium |
CN114827730A (en) * | 2022-04-19 | 2022-07-29 | 咪咕文化科技有限公司 | Video cover selecting method, device, equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109621332A (en) | A kind of attribute determining method, device, equipment and the storage medium of body-building movement | |
CN109409348A (en) | A kind of determination method, apparatus, equipment and the storage medium of user's sign | |
CN108256433B (en) | Motion attitude assessment method and system | |
CN109934111B (en) | Fitness posture estimation method and system based on key points | |
CN108764120B (en) | Human body standard action evaluation method | |
CN112237730B (en) | Body-building action correcting method and electronic equipment | |
CN104217454B (en) | A kind of human face animation generation method of video drive | |
CN107485844A (en) | A kind of limb rehabilitation training method, system and embedded device | |
CN110532874B (en) | Object attribute recognition model generation method, storage medium and electronic device | |
CN110222611A (en) | Human skeleton Activity recognition method, system, device based on figure convolutional network | |
US20210110146A1 (en) | Action recognition method and apparatus and electronic equipment | |
CN107930048B (en) | Space somatosensory recognition motion analysis system and motion analysis method | |
CN111437583A (en) | Badminton basic action auxiliary training system based on Kinect | |
CN109934847A (en) | The method and apparatus of weak texture three-dimension object Attitude estimation | |
CN105005995B (en) | A kind of method for calculating three-dimensional point cloud model bone | |
CN109978975A (en) | A kind of moving method and device, computer equipment of movement | |
CN105243375B (en) | A kind of motion characteristic extracting method and device | |
CN109472764A (en) | Method, apparatus, equipment and the medium of image synthesis and the training of image synthetic model | |
CN110298279A (en) | A kind of limb rehabilitation training householder method and system, medium, equipment | |
CN105446362B (en) | Posture detection based on computer science auxiliary adjusts devices and methods therefor | |
CN108211310A (en) | The methods of exhibiting and device of movement effects | |
US20200254299A1 (en) | System and non-intrusive method for exercise posture detection | |
CN110070052A (en) | A kind of robot control method based on mankind's demonstration video, device and equipment | |
CN110909621A (en) | Body-building guidance system based on vision | |
CN109407826A (en) | Ball game analogy method, device, storage medium and electronic equipment |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190416 |
|
RJ01 | Rejection of invention patent application after publication |