CN114287917A - Method and system for constructing motion recognition model of lower limbs of human body - Google Patents
Method and system for constructing motion recognition model of lower limbs of human body Download PDFInfo
- Publication number
- CN114287917A CN114287917A CN202111439552.9A CN202111439552A CN114287917A CN 114287917 A CN114287917 A CN 114287917A CN 202111439552 A CN202111439552 A CN 202111439552A CN 114287917 A CN114287917 A CN 114287917A
- Authority
- CN
- China
- Prior art keywords
- data
- marking
- motion
- data segment
- state
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 230000033001 locomotion Effects 0.000 title claims abstract description 145
- 210000003141 lower extremity Anatomy 0.000 title claims abstract description 64
- 238000000034 method Methods 0.000 title claims abstract description 39
- 230000009471 action Effects 0.000 claims abstract description 66
- 230000015654 memory Effects 0.000 claims description 34
- 238000003062 neural network model Methods 0.000 claims description 31
- 238000012360 testing method Methods 0.000 claims description 30
- 238000002372 labelling Methods 0.000 claims description 29
- 238000012549 training Methods 0.000 claims description 27
- 238000004590 computer program Methods 0.000 claims description 12
- 230000001133 acceleration Effects 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000010586 diagram Methods 0.000 description 10
- 244000309466 calf Species 0.000 description 6
- 210000003127 knee Anatomy 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 210000002414 leg Anatomy 0.000 description 5
- 210000000689 upper leg Anatomy 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 125000004122 cyclic group Chemical group 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000009191 jumping Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000007787 long-term memory Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
Images
Landscapes
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The embodiment of the invention provides a method and a system for constructing a motion recognition model of a lower limb of a human body. The method comprises the following steps: the method comprises the steps of determining action types of lower limbs of a human body, obtaining motion data of multiple parts of the lower limbs of the human body, obtaining pressure data of soles of feet of the human body, carrying out state marking on the motion data according to the pressure data to obtain state marking data, repeating the steps for multiple times to obtain state marking data corresponding to multiple groups of action types, and constructing an action recognition model of the lower limbs of the human body by adopting the state marking data corresponding to the multiple groups of action types.
Description
Technical Field
The invention relates to the technical field of motion recognition, in particular to a method and a system for constructing a motion recognition model of a lower limb of a human body.
Background
At present, a method for constructing a motion recognition model of a human body generally includes the steps of firstly placing an inertial sensor at each part of the human body, then capturing motion capture data of each joint point of the human body in a certain motion space, and training a neural network model by using the motion capture data, so as to realize construction of the motion recognition model. However, lower limb motions such as walking, running, leg lifting, kicking and the like of a human body have high similarity, and accurate lower limb motion recognition cannot be realized by adopting the motion recognition model constructed by the prior technical scheme.
Disclosure of Invention
In view of the above problems, embodiments of the present invention are provided to provide a method for constructing a motion recognition model of a lower limb of a human body and a corresponding system for constructing a motion recognition model of a lower limb of a human body, which overcome or at least partially solve the above problems.
In order to solve the above problems, an embodiment of the present invention discloses a method for constructing a motion recognition model of a lower limb of a human body, the method including:
determining the action type of the lower limbs of the human body;
acquiring motion data of a plurality of parts of the lower limbs of the human body;
acquiring pressure data of soles of the feet;
according to the pressure data, carrying out state marking on the motion data to obtain state marking data;
repeating the steps for multiple times to obtain state marking data corresponding to multiple groups of action types;
and constructing a motion recognition model of the lower limbs of the human body by adopting the state marking data corresponding to the plurality of groups of motion types.
Optionally, the motion data carries a first timestamp, the pressure data includes a second timestamp, a left foot pressure value and a right foot pressure value, the motion data is subjected to state labeling according to the pressure data, and the step of obtaining state labeling data includes:
time synchronizing the motion data and the pressure data according to the first time stamp and the second time stamp;
when the pressure value of the left foot is continuously zero, marking a corresponding data segment in the motion data as a left foot non-contact state data segment;
when the pressure value of the right foot is continuously zero, marking a corresponding data segment in the motion data as a data segment of a right foot non-contact state;
when the pressure value of the left foot is gradually increased from zero, marking the corresponding data segment in the motion data as a data segment of the initial contact state of the left foot;
when the pressure value of the right foot is gradually increased from zero, marking the corresponding data segment in the motion data as a data segment of the contact starting state of the right foot;
when the left foot pressure value is continuously not zero, marking a corresponding data segment in the motion data as a left foot continuous contact state data segment;
when the pressure value of the right foot is continuously not zero, marking a corresponding data segment in the motion data as a data segment of a continuous contact state of the right foot;
when the pressure value of the left foot is gradually reduced to zero, marking a corresponding data segment in the motion data as a left foot contact-free state data segment;
and when the pressure value of the right foot is gradually reduced to zero, marking the corresponding data segment in the motion data as a data segment of the contact-free state of the right foot.
Optionally, the step of constructing the motion recognition model of the lower limb of the human body by using the state labeling data corresponding to the plurality of groups of motion types includes:
dividing the state marking data corresponding to the plurality of groups of action types into a training sample set and a testing sample set;
constructing an initial long-short term memory neural network model;
training the initial long-short term memory neural network model by adopting the training sample set;
calculating the test error of the initial long-short term memory neural network model by adopting the test sample set;
and when the test error is lower than a preset threshold value, determining the initial long and short memory neural network model as a motion recognition model of the lower limbs of the human body.
Optionally, the motion data comprises acceleration data and angular velocity data.
The embodiment of the invention also discloses a system for constructing the motion recognition model of the lower limbs of the human body, which comprises the following steps:
the action type determining module is used for determining the action type of the lower limbs of the human body;
the motion data acquisition module is used for acquiring motion data of a plurality of parts of the lower limbs of the human body;
the pressure data acquisition module is used for acquiring pressure data of soles of feet;
the state labeling module is used for performing state labeling on the motion data according to the pressure data to obtain state labeling data;
the state marking data acquisition module is used for repeating the steps for multiple times to obtain state marking data corresponding to multiple groups of action types;
and the action recognition model building module is used for building an action recognition model of the lower limbs of the human body by adopting the state marking data corresponding to the plurality of groups of action types.
Optionally, the motion data carries a first timestamp, the pressure data includes a second timestamp, a left foot pressure value and a right foot pressure value, and the state labeling module includes:
the time synchronization submodule is used for performing time synchronization on the motion data and the pressure data according to the first time stamp and the second time stamp;
the left foot non-contact state marking submodule is used for marking a corresponding data segment in the motion data as a left foot non-contact state data segment when the left foot pressure value is continuously zero;
the right foot non-contact state marking sub-module is used for marking the corresponding data segment in the motion data as a right foot non-contact state data segment when the pressure value of the right foot is continuously zero;
the left foot initial contact state marking submodule is used for marking a corresponding data segment in the motion data as a left foot initial contact state data segment when the left foot pressure value is gradually increased from zero;
the right foot initial contact state marking submodule is used for marking a corresponding data segment in the motion data as a right foot initial contact state data segment when the pressure value of the right foot is gradually increased from zero;
the left foot continuous contact state marking submodule is used for marking a corresponding data segment in the motion data as a left foot continuous contact state data segment when the left foot pressure value is continuously not zero;
the right foot continuous contact state marking submodule is used for marking a corresponding data segment in the motion data as a right foot continuous contact state data segment when the right foot pressure value is continuously not zero;
the left foot contact-breaking state marking submodule is used for marking a corresponding data segment in the motion data as a left foot contact-breaking state data segment when the left foot pressure value is gradually reduced to zero;
and the right foot contact-free state marking submodule is used for marking the corresponding data segment in the motion data as a right foot contact-free state data segment when the pressure value of the right foot is gradually reduced to zero.
Optionally, the motion recognition model building module comprises:
the sample set classification submodule is used for classifying the state marking data corresponding to the plurality of groups of action types into a training sample set and a testing sample set;
the initial long-short term memory neural network model building submodule is used for building an initial long-short term memory neural network model;
the model training submodule is used for training the initial long-short term memory neural network model by adopting the training sample set;
the test error calculation submodule is used for calculating the test error of the initial long-short term memory neural network model by adopting the test sample set;
and the action recognition model determining submodule is used for determining the initial long and short memory neural network model as an action recognition model of the lower limbs of the human body when the test error is lower than a preset threshold value.
Optionally, the motion data comprises acceleration data and angular velocity data.
The embodiment of the invention has the following advantages: the method comprises the steps of determining action types of lower limbs of a human body, obtaining motion data of multiple parts of the lower limbs of the human body, obtaining pressure data of soles of feet of the human body, carrying out state marking on the motion data according to the pressure data to obtain state marking data, repeating the steps for multiple times to obtain state marking data corresponding to multiple groups of action types, and constructing an action recognition model of the lower limbs of the human body by adopting the state marking data corresponding to the multiple groups of action types.
Drawings
Fig. 1 is a flowchart illustrating steps of a first embodiment of a method for recognizing a motion of a lower limb of a human body according to the present invention.
Fig. 2 is a block diagram of a first embodiment of a human lower limb motion recognition model system according to the present invention.
Detailed Description
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
Referring to fig. 1, a flowchart illustrating steps of a first embodiment of a method for recognizing a motion of a lower limb of a human body according to the present invention is shown, which may specifically include the following steps:
the action types of the lower limbs of the human body can comprise walking, running, squatting, leg lifting, jumping and other actions. When a human subject makes a certain action, the action type thereof is determined. For example, when the subject person makes an action of raising the leg, the type of the action of the lower limb of the human body is determined as a leg raising action.
102, acquiring motion data of a plurality of parts of the lower limbs of a human body;
in the embodiment of the invention, the motion data of a plurality of parts of the lower limb of the human body is firstly acquired. The plurality of portions of the lower limb may include a left thigh portion, a right thigh portion, a left knee portion, a right knee portion, a left calf portion, a right calf portion. Therefore, the motion data includes motion data of a right thigh region, motion data of a left knee region, motion data of a right knee region, motion data of a left calf region, and motion data of a right calf region.
The motion data includes acceleration data and angular velocity data. Therefore, the acceleration sensor and the angular velocity sensor can be respectively placed at the left thigh part, the right thigh part, the left knee part, the right knee part, the left calf part and the right calf part of the human body to realize the motion data of a plurality of parts of the lower limb of the human body. It should be noted that the motion data carries a first time stamp.
103, acquiring pressure data of soles of the feet;
specifically, a pressure sensor is respectively worn on the sole part of the left foot and the sole part of the right foot of the feet, so that the pressure data of the soles of the feet can be acquired, and the pressure data comprises left foot pressure data and right foot sole pressure data. It should be noted that the pressure data carries a second time stamp.
104, performing state labeling on the motion data according to the pressure data to obtain state labeling data;
the motion data carries a first timestamp, the pressure data comprises a second timestamp, a left foot pressure value and a right foot pressure value, state labeling is carried out on the motion data according to the pressure data, and the step of obtaining state labeling data comprises the following steps:
time synchronizing the motion data and the pressure data according to the first time stamp and the second time stamp;
when the pressure value of the left foot is continuously zero, marking a corresponding data segment in the motion data as a left foot non-contact state data segment;
when the pressure value of the right foot is continuously zero, marking a corresponding data segment in the motion data as a data segment of a right foot non-contact state;
when the pressure value of the left foot is gradually increased from zero, marking the corresponding data segment in the motion data as a data segment of the initial contact state of the left foot;
when the pressure value of the right foot is gradually increased from zero, marking the corresponding data segment in the motion data as a data segment of the contact starting state of the right foot;
when the left foot pressure value is continuously not zero, marking a corresponding data segment in the motion data as a left foot continuous contact state data segment;
when the pressure value of the right foot is continuously not zero, marking a corresponding data segment in the motion data as a data segment of a continuous contact state of the right foot;
when the pressure value of the left foot is gradually reduced to zero, marking a corresponding data segment in the motion data as a left foot contact-free state data segment;
for example, from 0 second to 3 seconds, the pressure value of the sole of the left foot is continuously zero, and then the data segment of 0 second to 3 seconds in the motion data is labeled as the left foot non-contact state data segment.
And gradually increasing the pressure value of the sole of the right foot from zero from 3 seconds to 4 seconds, and marking the data segment of 3 seconds to 4 seconds in the motion data as the data segment of the contact starting state of the right foot.
When the pressure value of the sole of the left foot is continuously not zero from 5 seconds to 6 seconds, the data segment of 5 seconds to 6 seconds in the motion data is marked as the data segment of the continuous contact state of the left foot.
From 6 seconds to 7 seconds, the pressure value of the sole of the left foot gradually decreases to zero, and then the data segment of 6 seconds to 7 seconds in the exercise data is marked as the left foot out-of-contact state data segment.
105, repeating the steps to obtain state marking data corresponding to a plurality of groups of action types;
in the embodiment of the invention, after the steps are repeated for multiple times, the state marking data corresponding to multiple groups of action types can be obtained. For example, the status label data corresponding to the squat action type, the status label data corresponding to the leg raising action type, the status label data corresponding to the jumping action type, and so on.
And 106, constructing a motion recognition model of the lower limbs of the human body by adopting the plurality of groups of state labeling data.
In the embodiment of the present invention, the step of constructing the motion recognition model of the lower limb of the human body by using the state labeling data corresponding to the plurality of groups of motion types includes:
dividing the state marking data corresponding to the plurality of groups of action types into a training sample set and a testing sample set;
specifically, the plurality of sets of state labeling data may be randomly divided into a training sample set and a testing sample set.
Constructing an initial long-short term memory neural network model;
the long-term and short-term memory network is a time-cycle neural network and is specially designed for solving the long-term dependence problem of the common cyclic neural network, and all the cyclic neural networks have a chain form of repeated neural network modules.
Training the initial long-short term memory neural network model by adopting the training sample set;
after randomly dividing a plurality of groups of state labeling data into a training sample set and a testing sample set, training the initial long-short term memory neural network model by adopting the training sample set.
Calculating the test error of the initial long-short term memory neural network model by adopting the test sample set;
and when the test error is lower than a preset threshold value, determining the initial long and short memory neural network model as a motion recognition model of the lower limbs of the human body.
And when the test error is lower than a preset threshold value, stopping training the initial long-short term memory neural network model, and determining the initial long-short term memory neural network model at the moment as a motion recognition model of the lower limbs of the human body.
In the embodiment of the invention, the action type of the lower limb of the human body is determined, the motion data of a plurality of parts of the lower limb of the human body is obtained, the pressure data of soles of both feet is obtained, the motion data is subjected to state marking according to the pressure data to obtain state marking data, the steps are repeated for a plurality of times to obtain state marking data corresponding to a plurality of groups of action types, the state marking data corresponding to the plurality of groups of action types are adopted to construct the action recognition model of the lower limb of the human body, and the action recognition model constructed by adopting the method can more accurately recognize the action of the lower limb with high similarity.
It should be noted that, for simplicity of description, the method embodiments are described as a series of acts or combination of acts, but those skilled in the art will recognize that the present invention is not limited by the illustrated order of acts, as some steps may occur in other orders or concurrently in accordance with the embodiments of the present invention. Further, those skilled in the art will appreciate that the embodiments described in the specification are presently preferred and that no particular act is required to implement the invention.
Referring to fig. 2, a block diagram of a first embodiment of a system for constructing a motion recognition model of a lower limb of a human body according to the present invention is shown, and may specifically include the following modules:
an action type determining module 201, configured to determine an action type of a lower limb of a human body;
the motion data acquisition module 202 is used for acquiring motion data of a plurality of parts of the lower limbs of the human body;
the pressure data acquisition module 203 is used for acquiring pressure data of soles of feet;
the state labeling module 204 is configured to perform state labeling on the motion data according to the pressure data to obtain state labeling data;
a status labeling data obtaining module 205, configured to repeat the above steps for multiple times to obtain status labeling data corresponding to multiple groups of action types;
and the action recognition model building module 206 is configured to build an action recognition model of the lower limb of the human body by using the state labeling data corresponding to the plurality of groups of action types.
In an embodiment of the present invention, the motion data carries a first timestamp, the pressure data includes a second timestamp, a left foot pressure value, and a right foot pressure value, and the state labeling module includes:
the time synchronization submodule is used for performing time synchronization on the motion data and the pressure data according to the first time stamp and the second time stamp;
the left foot non-contact state marking submodule is used for marking a corresponding data segment in the motion data as a left foot non-contact state data segment when the left foot pressure value is continuously zero;
the right foot non-contact state marking sub-module is used for marking the corresponding data segment in the motion data as a right foot non-contact state data segment when the pressure value of the right foot is continuously zero;
the left foot initial contact state marking submodule is used for marking a corresponding data segment in the motion data as a left foot initial contact state data segment when the left foot pressure value is gradually increased from zero;
the right foot initial contact state marking submodule is used for marking a corresponding data segment in the motion data as a right foot initial contact state data segment when the pressure value of the right foot is gradually increased from zero;
the left foot continuous contact state marking submodule is used for marking a corresponding data segment in the motion data as a left foot continuous contact state data segment when the left foot pressure value is continuously not zero;
the right foot continuous contact state marking submodule is used for marking a corresponding data segment in the motion data as a right foot continuous contact state data segment when the right foot pressure value is continuously not zero;
the left foot contact-breaking state marking submodule is used for marking a corresponding data segment in the motion data as a left foot contact-breaking state data segment when the left foot pressure value is gradually reduced to zero;
and the right foot contact-free state marking submodule is used for marking the corresponding data segment in the motion data as a right foot contact-free state data segment when the pressure value of the right foot is gradually reduced to zero. In the embodiment of the present invention, the action recognition model building module includes:
the sample set classification submodule is used for classifying the state marking data corresponding to the plurality of groups of action types into a training sample set and a testing sample set;
the initial long-short term memory neural network model building submodule is used for building an initial long-short term memory neural network model;
the model training submodule is used for training the initial long-short term memory neural network model by adopting the training sample set;
the test error calculation submodule is used for calculating the test error of the initial long-short term memory neural network model by adopting the test sample set;
and the action recognition model determining submodule is used for determining the initial long and short memory neural network model as an action recognition model of the lower limbs of the human body when the test error is lower than a preset threshold value.
In an embodiment of the invention, the motion data comprises acceleration data and angular velocity data.
For the device embodiment, since it is basically similar to the method embodiment, the description is simple, and for the relevant points, refer to the partial description of the method embodiment.
An embodiment of the present invention further provides an apparatus, including:
the method comprises a processor, a memory and a computer program which is stored on the memory and can run on the processor, wherein when the computer program is executed by the processor, each process of the embodiment of the method for constructing the motion recognition model of the lower limbs of the human body is realized, the same technical effect can be achieved, and the method is not repeated herein for avoiding repetition.
The embodiment of the invention also provides a computer-readable storage medium, wherein a computer program is stored on the computer-readable storage medium, and when being executed by a processor, the computer program realizes each process of the above-mentioned method for constructing and controlling the motion recognition model of the lower limbs of the human body, and can achieve the same technical effect, and is not repeated here in order to avoid repetition.
The embodiments in the present specification are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, apparatus, or computer program product. Accordingly, embodiments of the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, embodiments of the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
Embodiments of the present invention are described with reference to flowchart illustrations and/or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing terminal to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing terminal to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing terminal to cause a series of operational steps to be performed on the computer or other programmable terminal to produce a computer implemented process such that the instructions which execute on the computer or other programmable terminal provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While preferred embodiments of the present invention have been described, additional variations and modifications of these embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all such alterations and modifications as fall within the scope of the embodiments of the invention.
Finally, it should also be 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 terminal 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 terminal. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or terminal that comprises the element.
The method for constructing the motion recognition model of the lower limbs of the human body and the system for constructing the motion recognition model of the lower limbs of the human body are described in detail, specific examples are applied to explain the principle and the implementation mode of the invention, and the description of the examples is only used for helping to understand the method and the core idea of the invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present specification should not be construed as a limitation to the present invention.
Claims (9)
1. A method for constructing a motion recognition model of a lower limb of a human body is characterized by comprising the following steps:
determining the action type of the lower limbs of the human body;
acquiring motion data of a plurality of parts of the lower limbs of the human body;
acquiring pressure data of soles of the feet;
according to the pressure data, carrying out state marking on the motion data to obtain state marking data;
repeating the steps for multiple times to obtain state marking data corresponding to multiple groups of action types;
and constructing a motion recognition model of the lower limbs of the human body by adopting the state marking data corresponding to the plurality of groups of motion types.
2. The method according to claim 1, wherein the motion data carries a first timestamp, the pressure data comprises a second timestamp, a left foot pressure value and a right foot pressure value, and the step of performing state labeling on the motion data according to the pressure data to obtain state labeling data comprises:
time synchronizing the motion data and the pressure data according to the first time stamp and the second time stamp;
when the pressure value of the left foot is continuously zero, marking a corresponding data segment in the motion data as a left foot non-contact state data segment;
when the pressure value of the right foot is continuously zero, marking a corresponding data segment in the motion data as a data segment of a right foot non-contact state;
when the pressure value of the left foot is gradually increased from zero, marking the corresponding data segment in the motion data as a data segment of the initial contact state of the left foot;
when the pressure value of the right foot is gradually increased from zero, marking the corresponding data segment in the motion data as a data segment of the contact starting state of the right foot;
when the left foot pressure value is continuously not zero, marking a corresponding data segment in the motion data as a left foot continuous contact state data segment;
when the pressure value of the right foot is continuously not zero, marking a corresponding data segment in the motion data as a data segment of a continuous contact state of the right foot;
when the pressure value of the left foot is gradually reduced to zero, marking a corresponding data segment in the motion data as a left foot contact-free state data segment;
and when the pressure value of the right foot is gradually reduced to zero, marking the corresponding data segment in the motion data as a data segment of the contact-free state of the right foot.
3. The method according to claim 1, wherein the step of constructing the motion recognition model of the lower limbs of the human body by using the state labeling data corresponding to the plurality of groups of motion types comprises:
dividing the state marking data corresponding to the plurality of groups of action types into a training sample set and a testing sample set;
constructing an initial long-short term memory neural network model;
training the initial long-short term memory neural network model by adopting the training sample set;
calculating the test error of the initial long-short term memory neural network model by adopting the test sample set;
and when the test error is lower than a preset threshold value, determining the initial long and short memory neural network model as a motion recognition model of the lower limbs of the human body.
4. The method of claim 1, wherein the motion data comprises acceleration data and angular velocity data.
5. A system for constructing a motion recognition model of a lower limb of a human body, the system comprising:
the action type determining module is used for determining the action type of the lower limbs of the human body;
the motion data acquisition module is used for acquiring motion data of a plurality of parts of the lower limbs of the human body;
the pressure data acquisition module is used for acquiring pressure data of soles of feet;
the state labeling module is used for performing state labeling on the motion data according to the pressure data to obtain state labeling data;
the state marking data acquisition module is used for repeating the steps for multiple times to obtain state marking data corresponding to multiple groups of action types;
and the action recognition model building module is used for building an action recognition model of the lower limbs of the human body by adopting the state marking data corresponding to the plurality of groups of action types.
6. The system of claim 5, wherein the motion data carries a first timestamp, wherein the pressure data comprises a second timestamp, a left foot pressure value, and a right foot pressure value, and wherein the status labeling module comprises:
the time synchronization submodule is used for performing time synchronization on the motion data and the pressure data according to the first time stamp and the second time stamp;
the left foot non-contact state marking submodule is used for marking a corresponding data segment in the motion data as a left foot non-contact state data segment when the left foot pressure value is continuously zero;
the right foot non-contact state marking sub-module is used for marking the corresponding data segment in the motion data as a right foot non-contact state data segment when the pressure value of the right foot is continuously zero;
the left foot initial contact state marking submodule is used for marking a corresponding data segment in the motion data as a left foot initial contact state data segment when the left foot pressure value is gradually increased from zero;
the right foot initial contact state marking submodule is used for marking a corresponding data segment in the motion data as a right foot initial contact state data segment when the pressure value of the right foot is gradually increased from zero;
the left foot continuous contact state marking submodule is used for marking a corresponding data segment in the motion data as a left foot continuous contact state data segment when the left foot pressure value is continuously not zero;
the right foot continuous contact state marking submodule is used for marking a corresponding data segment in the motion data as a right foot continuous contact state data segment when the right foot pressure value is continuously not zero;
the left foot contact-breaking state marking submodule is used for marking a corresponding data segment in the motion data as a left foot contact-breaking state data segment when the left foot pressure value is gradually reduced to zero;
and the right foot contact-free state marking submodule is used for marking the corresponding data segment in the motion data as a right foot contact-free state data segment when the pressure value of the right foot is gradually reduced to zero.
7. The system of claim 5, wherein the action recognition model building module comprises:
the sample set classification submodule is used for classifying the state marking data corresponding to the plurality of groups of action types into a training sample set and a testing sample set;
the initial long-short term memory neural network model building submodule is used for building an initial long-short term memory neural network model;
the model training submodule is used for training the initial long-short term memory neural network model by adopting the training sample set;
the test error calculation submodule is used for calculating the test error of the initial long-short term memory neural network model by adopting the test sample set;
and the action recognition model determining submodule is used for determining the initial long and short memory neural network model as an action recognition model of the lower limbs of the human body when the test error is lower than a preset threshold value.
8. The system of claim 5, wherein the motion data comprises acceleration data and angular velocity data.
9. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method for constructing a model for identifying the movement of a lower limb of a human body according to any one of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111439552.9A CN114287917B (en) | 2021-11-30 | 2021-11-30 | Method and system for constructing action recognition model of human lower limb |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111439552.9A CN114287917B (en) | 2021-11-30 | 2021-11-30 | Method and system for constructing action recognition model of human lower limb |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114287917A true CN114287917A (en) | 2022-04-08 |
| CN114287917B CN114287917B (en) | 2023-12-01 |
Family
ID=80965967
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111439552.9A Active CN114287917B (en) | 2021-11-30 | 2021-11-30 | Method and system for constructing action recognition model of human lower limb |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114287917B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104799862A (en) * | 2015-04-02 | 2015-07-29 | 鲁东大学 | Human body unbalance warning method and system |
| KR20170043308A (en) * | 2015-10-13 | 2017-04-21 | 충북대학교 산학협력단 | Method for identificating Person on the basis gait data |
| CN107092861A (en) * | 2017-03-15 | 2017-08-25 | 华南理工大学 | Lower limb movement recognition methods based on pressure and acceleration transducer |
| CN107368796A (en) * | 2017-07-06 | 2017-11-21 | 中国科学院合肥物质科学研究院 | A kind of gait obtains and interactive device and its exchange method |
| US20200000373A1 (en) * | 2014-04-22 | 2020-01-02 | The Trustees Of Columbia University In The City Of New York | Gait Analysis Devices, Methods, and Systems |
| CN112754468A (en) * | 2021-01-07 | 2021-05-07 | 华南理工大学 | Human body lower limb movement detection and identification method based on multi-source signals |
-
2021
- 2021-11-30 CN CN202111439552.9A patent/CN114287917B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20200000373A1 (en) * | 2014-04-22 | 2020-01-02 | The Trustees Of Columbia University In The City Of New York | Gait Analysis Devices, Methods, and Systems |
| CN104799862A (en) * | 2015-04-02 | 2015-07-29 | 鲁东大学 | Human body unbalance warning method and system |
| KR20170043308A (en) * | 2015-10-13 | 2017-04-21 | 충북대학교 산학협력단 | Method for identificating Person on the basis gait data |
| CN107092861A (en) * | 2017-03-15 | 2017-08-25 | 华南理工大学 | Lower limb movement recognition methods based on pressure and acceleration transducer |
| CN107368796A (en) * | 2017-07-06 | 2017-11-21 | 中国科学院合肥物质科学研究院 | A kind of gait obtains and interactive device and its exchange method |
| CN112754468A (en) * | 2021-01-07 | 2021-05-07 | 华南理工大学 | Human body lower limb movement detection and identification method based on multi-source signals |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114287917B (en) | 2023-12-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Abreu et al. | Learning to run faster in a humanoid robot soccer environment through reinforcement learning | |
| Dikovski et al. | Evaluation of different feature sets for gait recognition using skeletal data from Kinect | |
| CN111563487A (en) | Dance scoring method based on gesture recognition model and related equipment | |
| US9183431B2 (en) | Apparatus and method for providing activity recognition based application service | |
| CN113392742A (en) | Abnormal action determination method and device, electronic equipment and storage medium | |
| CN112686075A (en) | Behavior detection method, device and system | |
| CN112998700B (en) | Apparatus, system and method for assisting assessment of a motor function of an object | |
| CN110782482A (en) | Motion evaluation method and device, computer equipment and storage medium | |
| CN111274932B (en) | State identification method and device based on human gait in video and storage medium | |
| CN113392743B (en) | Abnormal action detection method, abnormal action detection device, electronic equipment and computer storage medium | |
| Ashapkina et al. | Metric for exercise recognition for telemedicine systems | |
| CN114287917B (en) | Method and system for constructing action recognition model of human lower limb | |
| Worsey et al. | Automatic classification of running surfaces using an ankle-worn inertial sensor | |
| Wu et al. | Human gait-labeling uncertainty and a hybrid model for gait segmentation | |
| Pardoel et al. | Grouping successive freezing of gait episodes has neutral to detrimental effect on freeze detection and prediction in Parkinson’s disease | |
| Adel et al. | Gait-based Person Identification using Multiple Inertial Sensors. | |
| CN108903947A (en) | Gait analysis method, gait analysis device and readable storage medium storing program for executing | |
| Josiński et al. | Assessment of local dynamic stability in gait based on univariate and multivariate time series | |
| Kailas | Basic human motion tracking using a pair of gyro+ accelerometer MEMS devices | |
| CN113392744A (en) | Dance motion aesthetic feeling confirmation method and device, electronic equipment and storage medium | |
| CN115105057A (en) | Gait phase dividing and identifying method and system for transverse walking | |
| Gutiérrez et al. | Human stability assessment and fall detection based on dynamic descriptors | |
| US20230190138A1 (en) | Information processing apparatus, information processing method, program, information processing systems, and generation method | |
| Palma et al. | Hmm based evaluation of physical therapy movements using kinect tracking | |
| Yilmaz et al. | Gait motion simulator for kinematic tests of above knee prostheses |
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 | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20230927 Address after: Room 116-2, 1st Floor, No. 198 Xinghua Road, Xiangzhou District, Zhuhai City, Guangdong Province, 519099 Applicant after: Fantasy (Zhuhai) Technology Co.,Ltd. Applicant after: GUANGZHOU HUANTEK Co.,Ltd. Applicant after: DONGGUAN YILIAN INTERATION INFORMATION TECHNOLOGY Co.,Ltd. Address before: 510635 self made room 01-011, third floor, No. 721, Tianhe North Road, Tianhe District, Guangzhou City, Guangdong Province (office only) Applicant before: GUANGZHOU HUANTEK Co.,Ltd. Applicant before: DONGGUAN YILIAN INTERATION INFORMATION TECHNOLOGY Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |