CN114764819A - Human body posture estimation method and device based on filtering algorithm - Google Patents

Human body posture estimation method and device based on filtering algorithm Download PDF

Info

Publication number
CN114764819A
CN114764819A CN202210058731.6A CN202210058731A CN114764819A CN 114764819 A CN114764819 A CN 114764819A CN 202210058731 A CN202210058731 A CN 202210058731A CN 114764819 A CN114764819 A CN 114764819A
Authority
CN
China
Prior art keywords
human body
matrix
key points
posture estimation
filtering
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
Application number
CN202210058731.6A
Other languages
Chinese (zh)
Inventor
马捷径
李海洋
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Deck Intelligent Technology Co ltd
Original Assignee
Beijing Deck Intelligent Technology Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Beijing Deck Intelligent Technology Co ltd filed Critical Beijing Deck Intelligent Technology Co ltd
Priority to CN202210058731.6A priority Critical patent/CN114764819A/en
Publication of CN114764819A publication Critical patent/CN114764819A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/20—Analysis of motion
    • G06T7/246—Analysis of motion using feature-based methods, e.g. the tracking of corners or segments
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/20—Image enhancement or restoration by the use of local operators
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20024—Filtering details
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30196—Human being; Person

Abstract

The embodiment of the invention discloses a human body posture estimation method and a human body posture estimation device based on a filtering algorithm, wherein the method comprises the following steps: extracting a plurality of human body key points in a target video frame based on a pre-stored human body posture estimation model; setting the width of a filtering window, and acquiring a plurality of data points in the filtering window; fitting the plurality of human body key points and the plurality of data points based on a pre-stored filter algorithm model to obtain a combined matrix aiming at any human body key point; and obtaining the attitude positions corresponding to the plurality of human body key points according to the joint matrix. According to the method, on the basis of the existing human body posture estimation, a joint matrix is obtained by utilizing the optimization of a filtering algorithm, the corresponding posture position is obtained through the joint matrix, the action relation among the associated key points is fully considered, the output posture position is more coherent and accurate, and the technical problems of poor stability and low accuracy of the human body posture estimation in the prior art are solved.

Description

Human body posture estimation method and device based on filtering algorithm
Technical Field
The invention relates to the technical field of human body pose detection methods, in particular to a human body pose estimation method and device based on a filtering algorithm.
Background
This section provides background information related to the present disclosure only and is not necessarily prior art.
Human body posture estimation is an important branch of computer vision, can detect human body joint point position information from a given image or a section of video, and is widely applied to the fields of pedestrian posture capture, motion analysis, pedestrian re-identification and the like.
At present, the estimation of human body posture mainly utilizes deep learning and is roughly divided into two types: top-down methods and bottom-up methods. Detecting a human body from top to bottom, and then estimating the posture of a single person; and the human body joint points are detected from bottom to top and then connected into a human body skeleton according to the detected joint points.
However, when the existing human body posture estimation method is used for motion analysis, the stability and accuracy of human body posture estimation are low, so that the subsequent motion recognition accuracy cannot be improved.
Disclosure of Invention
Therefore, the embodiment of the invention provides a human body posture estimation method and device based on a filtering algorithm, so as to at least partially solve the technical problems of poor stability and low accuracy of human body posture estimation in the prior art.
In order to achieve the above object, the embodiments of the present invention provide the following technical solutions:
A human body posture estimation method based on a filtering algorithm comprises the following steps:
extracting a plurality of human body key points in a target video frame based on a pre-stored human body posture estimation model;
setting the width of a filtering window, and acquiring a plurality of data points in the filtering window;
fitting the plurality of human body key points and the plurality of data points based on a pre-stored filter algorithm model to obtain a combined matrix aiming at any human body key point;
and obtaining the attitude positions corresponding to the plurality of human body key points according to the joint matrix.
Further, the setting of the width of the filtering window and the obtaining of the plurality of data points in the filtering window specifically include:
the filter window width w is set using the following equation:
w＝2m+1
wherein m is a constant and m is a positive integer;
the multiple data points are each x ═ (-m, -m +1, …, m-1, m).
Further, when the window width of the filter is set, the value range of the constant m is 2 to 4.
Further, fitting the plurality of human keypoints and the plurality of data points using the following formula:
y＝ao+a1x+a2x2+…+ak-1xk-1
wherein y represents the observed data points to be fitted;
ao，a1，a2、…，ak-1representing polynomial coefficients;
x represents a measurement point;
k denotes the polynomial order.
Further, the joint matrix for any of the human body key points is:
Pi＝M·Ai+Ei
wherein the content of the first and second substances,
Figure BDA0003474062810000031
representing a pose matrix formed by continuous 2m +1 frames of the ith key point;
Figure BDA0003474062810000032
representing a polynomial matrix;
Figure BDA0003474062810000033
a coefficient matrix representing the ith keypoint;
Figure BDA0003474062810000034
the fitting error matrix of the ith keypoint is represented.
Further, obtaining the corresponding points of the plurality of human body key points according to the joint matrixAttitude position P ofi' is:
Pi′＝M·Ai＝M(MTM)-1MTPi；
wherein, PiRepresenting a pose matrix formed by 2m +1 continuous frames of the ith key point;
m represents a polynomial matrix;
Aia coefficient matrix representing the ith keypoint;
t denotes a matrix transposition.
The invention also provides a human body posture estimation device, which comprises:
the key point extraction unit is used for extracting a plurality of human body key points in the target video frame based on a pre-stored human body posture estimation model;
the data point extraction unit is used for setting the width of a filtering window and acquiring a plurality of data points in the filtering window;
the filtering optimization unit is used for fitting the plurality of human key points and the plurality of data points based on a pre-stored filtering algorithm model to obtain a combined matrix aiming at any human key point;
And the attitude position output unit is used for obtaining the attitude positions corresponding to the plurality of human key points according to the joint matrix.
The invention also provides an electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps of the method as described above when executing the program.
The invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the method as described above.
The invention also provides a computer program product comprising a computer program which, when executed by a processor, carries out the steps of the method as described above.
The human body posture estimation method based on the filtering algorithm extracts a plurality of human body key points in a target video frame through a pre-stored human body posture estimation model; then setting the width of a filtering window, acquiring a plurality of data points in the filtering window, and fitting the plurality of human body key points and the plurality of data points based on a pre-stored filtering algorithm model to obtain a combined matrix for any human body key point; and then obtaining the corresponding posture positions of the plurality of human body key points according to the joint matrix. Therefore, the method integrates a filtering algorithm on the basis of the existing human body posture estimation, obtains a joint matrix by utilizing the optimization of the filtering algorithm, obtains the corresponding posture position through the joint matrix, fully considers the action relation among the associated key points, enables the output posture position to be more coherent and accurate, and solves the technical problems of poor stability and low accuracy of the human body posture estimation in the prior art.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below. It should be apparent that the drawings in the following description are merely exemplary, and that other embodiments can be derived from the drawings provided by those of ordinary skill in the art without inventive effort.
The structures, ratios, sizes, and the like shown in the present specification are only used for matching with the contents disclosed in the specification, so as to be understood and read by those skilled in the art, and are not used to limit the conditions that the present invention can be implemented, so that the present invention has no technical significance, and any structural modifications, changes in the ratio relationship, or adjustments of the sizes, without affecting the effects and the achievable by the present invention, should still fall within the range that the technical contents disclosed in the present invention can cover.
FIG. 1 is a flowchart of an embodiment of a method for estimating a human body posture according to the present invention;
FIGS. 2 and 3 are diagrams illustrating the effect of a human pose estimation method according to an embodiment;
FIG. 4 is a block diagram of a human body posture estimation method according to an embodiment of the present invention;
Fig. 5 is a block diagram of a computer device according to the present invention.
Detailed Description
The present invention is described in terms of specific embodiments, and other advantages and benefits of the present invention will become apparent to those skilled in the art from the following disclosure. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In the scene of motion analysis, the detected human body posture needs to be continuously analyzed, and the stability and accuracy of human body posture estimation directly influence the result of motion analysis. Therefore, the method combines a combined filtering algorithm on the basis of the human body posture estimation model, takes the human body posture estimation sequence as input, realizes the real-time stable output of the human body posture key points by expanding the S-G filtering algorithm, effectively ensures the stability of the real-time video output human body key point positions, and improves the subsequent action recognition precision.
Referring to fig. 1, fig. 1 is a flowchart of a human body posture estimation method according to an embodiment of the present invention.
In a specific embodiment, as shown in fig. 1, the method for estimating a human body posture based on a filtering algorithm provided by the present invention includes the following steps:
s1: and extracting a plurality of human key points in the target video frame based on a pre-stored human posture estimation model. That is, human key points in video frames are extracted using a human pose estimation model
Figure BDA0003474062810000061
Wherein the content of the first and second substances,
Figure BDA0003474062810000062
represents the t-th frameKeypoint locations of s keypoints. The human pose estimation model may be embodied as OpenPose, AlphaPose or BlazePose.
S2: setting the width of a filtering window, and acquiring a plurality of data points in the filtering window.
Specifically, the filter window width w can be set using the following equation:
w＝2m+1
where m is a constant and m is a positive integer, the plurality of data points are each x ═ (-m, -m +1, …, m-1, m).
The window width of the filter can be set according to the real-time requirement of the algorithm and the fluctuation degree of data, and the filter is not suitable to be too long or too short, the result is too smooth due to too long, and the filtering effect is not obvious due to too short. In this embodiment, m takes on a value of [2,4] for processing real-time interaction scenarios.
S3: fitting the plurality of human body key points and the plurality of data points based on a pre-stored filter algorithm model to obtain a joint matrix for any human body key point.
In this embodiment, the above filtering algorithm model may be selected from the Savitzky-Golay filter. In principle, the Savitzky-Golay filter is a low-pass filter which is used for smoothing and denoising data streams, is a method for performing best fitting by a least square method through a moving window based on a polynomial in a time domain, and can better keep distribution characteristics such as relative maximum, minimum and width compared with other similar averaging methods. The human body posture estimation sequence is used as input, and the real-time stable output of the human body posture key points is realized by expanding an S-G filtering algorithm.
Specifically, when fitting the plurality of human body key points and the plurality of data points, according to the S-G algorithm, assuming that the filter window width w is 2m +1(m ∈ N), and each measurement point is x (-m, -m +1, …, m-1, m), fitting the data points in the window by using a k-1 degree polynomial:
y＝ao+a1x+a2x2+…+ak-1xk-1 (1)
wherein y represents the observed data points to be fitted;
ao，a1，a2、…，ak-1representing polynomial coefficients;
x represents a measurement point;
k denotes the polynomial order.
According to equation (1), the coordinate fit of the ith keypoint at the (-m, -m +1, …, m-1, m) frame can be combined as follows:
Figure BDA0003474062810000071
formula (2) is abbreviated as a joint matrix for any of the human body key points:
Pi＝M·Ai+Ei
Wherein, the first and the second end of the pipe are connected with each other,
Figure BDA0003474062810000081
representing a pose matrix formed by 2m +1 continuous frames of the ith key point;
Figure BDA0003474062810000082
representing a polynomial matrix;
Figure BDA0003474062810000083
a coefficient matrix representing the ith keypoint;
Figure BDA0003474062810000084
the fitting error matrix of the ith keypoint is represented.
S4: and obtaining the attitude positions corresponding to the plurality of human body key points according to the joint matrix.
Obtaining the attitude positions P corresponding to the plurality of human body key points according to the combined matrix after least square filteringi' is:
Pi′＝M·Ai＝M(MTM)-1MTPi (3)
wherein, PiRepresenting a pose matrix formed by 2m +1 continuous frames of the ith key point;
m represents a polynomial matrix;
Aia coefficient matrix representing the ith keypoint;
t denotes a matrix transposition.
The following is a brief description of the implementation process of the human body posture estimation method provided by the present invention, taking a specific use scenario as an example.
Taking openpos as a human body posture estimation model as an example, as shown in fig. 2, the model can output 25 key points. The key points are not isolated, and a connecting line between the key points can be simply understood as a connecting relation between two key points, but other connections can be constructed. For example, the position 0 is connected to the points {1,15,16}, so that these points need to be taken into account to construct a fitting equation, such as equation (4), when constructing S-G filtering.
Figure BDA0003474062810000091
The meaning of the neutron matrix in the formula (4) can be referred to the formula (3), P0、P1、P15、P16Respectively representing pose matrixes formed by continuous 2m +1 frames of key points 0, 1, 15 and 16; a. the0、A1、A15、A16Polynomial matrixes respectively representing key points of numbers 0, 1, 15 and 16; e0、E1、E15、E16Error matrixes respectively representing key points of 0, 1, 15 and 16;
Figure BDA0003474062810000092
a union matrix is constructed.
When all the key points of OpenPose are filtered, the matrix M is combined to clearly see whether the key points are connected or notposeCan be represented as a figure3, in fig. 3, each dark block represents an M matrix and the white blocks represent a zero matrix. Assuming the position P of all key pointspose＝(P0,P1,…,P24)TFiltered keypoint locations, also according to least squares
Figure BDA0003474062810000101
Figure BDA0003474062810000102
In this way, in the above specific embodiment, the human body posture estimation method based on the filtering algorithm provided by the present invention extracts a plurality of human body key points in the target video frame through the pre-stored human body posture estimation model; then setting the width of a filtering window, acquiring a plurality of data points in the filtering window, and fitting the plurality of human body key points and the plurality of data points based on a pre-stored filtering algorithm model to obtain a combined matrix for any human body key point; and then obtaining the corresponding posture positions of the plurality of human body key points according to the joint matrix. Therefore, the method integrates a filtering algorithm on the basis of the existing human body posture estimation, obtains a joint matrix by utilizing the optimization of the filtering algorithm, obtains the corresponding posture position through the joint matrix, fully considers the action relation among the associated key points, enables the output posture position to be more coherent and accurate, and solves the technical problems of poor stability and low accuracy of the human body posture estimation in the prior art.
In addition to the above method, the present invention also provides a human body posture estimating apparatus, as shown in fig. 4, the apparatus comprising:
a key point extracting unit 100, configured to extract a plurality of human key points in a target video frame based on a pre-stored human pose estimation model;
a data point extracting unit 200, configured to set a width of a filtering window, and obtain a plurality of data points in the filtering window;
a filtering optimization unit 300, configured to fit the plurality of human body key points and the plurality of data points based on a pre-stored filtering algorithm model to obtain a joint matrix for any human body key point;
and the posture position output unit 400 is configured to obtain posture positions corresponding to the plurality of human body key points according to the joint matrix.
In the above specific embodiment, the human body posture estimation device based on the filtering algorithm provided by the invention extracts a plurality of human body key points in a target video frame through a pre-stored human body posture estimation model; then setting the width of a filtering window, acquiring a plurality of data points in the filtering window, and fitting the plurality of human body key points and the plurality of data points based on a pre-stored filtering algorithm model to obtain a combined matrix for any human body key point; and then obtaining the corresponding posture positions of the plurality of human body key points according to the joint matrix. Therefore, the device integrates a filtering algorithm on the basis of the existing human body posture estimation, obtains a joint matrix by utilizing the optimization of the filtering algorithm, obtains a corresponding posture position through the joint matrix, fully considers the action relation among related key points, enables the output posture position to be more coherent and accurate, and solves the technical problems of poor stability and low accuracy of the human body posture estimation in the prior art.
In one embodiment, a computer device is provided, which may be a server, the internal structure of which may be as shown in fig. 5. The computer device includes a processor, a memory, and a network interface connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a model prediction. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The model prediction of the computer device is used to store static information and dynamic information data. The network interface of the computer device is used for communicating with an external terminal through a network connection. Which computer program is executed by a processor to carry out the steps in the above-described method embodiments.
Those skilled in the art will appreciate that the configuration shown in fig. 5 is a block diagram of only a portion of the configuration associated with aspects of the present invention and is not intended to limit the computing devices to which aspects of the present invention may be applied, and that a particular computing device may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
In correspondence with the above embodiments, embodiments of the present invention also provide a computer storage medium containing one or more program instructions. Wherein the one or more program instructions are for performing the method described above by a weight verification system.
The invention also provides a computer program product comprising a computer program, storable on a non-transitory computer readable storage medium, which, when executed by a processor, is capable of executing the above method by a computer.
In an embodiment of the invention, the processor may be an integrated circuit chip having signal processing capability. The Processor may be a general purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other programmable logic device, discrete Gate or transistor logic device, discrete hardware component.
The various methods, steps and logic blocks disclosed in the embodiments of the present invention may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of the method disclosed in connection with the embodiments of the present invention may be directly implemented by a hardware decoding processor, or implemented by a combination of hardware and software modules in the decoding processor. The software module may be located in ram, flash memory, rom, prom, or eprom, registers, etc. storage media as is well known in the art. The processor reads the information in the storage medium and completes the steps of the method in combination with the hardware.
The storage medium may be a memory, for example, which may be volatile memory or nonvolatile memory, or which may include both volatile and nonvolatile memory.
The nonvolatile Memory may be a Read-Only Memory (ROM), a Programmable ROM (PROM), an Erasable PROM (EPROM), an Electrically Erasable PROM (EEPROM), or a flash Memory.
The volatile Memory may be a Random Access Memory (RAM) which serves as an external cache. By way of example and not limitation, many forms of RAM are available, such as Static Random Access Memory (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced SDRAM (ESDRAM), SLDRAM (SLDRAM), and Direct Rambus RAM (DRRAM).
The storage media described in connection with the embodiments of the invention are intended to comprise, without being limited to, these and any other suitable types of memory.
Those skilled in the art will appreciate that the functionality described in the present invention may be implemented in a combination of hardware and software in one or more of the examples described above. When software is applied, the corresponding functionality may be stored on or transmitted over as one or more instructions or code on a computer-readable medium. Computer-readable media includes both computer storage media and communication media including any medium that facilitates transfer of a computer program from one place to another. A storage media may be any available media that can be accessed by a general purpose or special purpose computer.
The above embodiments are only for illustrating the embodiments of the present invention and are not to be construed as limiting the scope of the present invention, and any modifications, equivalent substitutions, improvements and the like made on the basis of the embodiments of the present invention shall be included in the scope of the present invention.

Claims (10)

1. A human body posture estimation method based on a filtering algorithm is characterized by comprising the following steps:
extracting a plurality of human body key points in a target video frame based on a pre-stored human body posture estimation model;
setting the width of a filtering window, and acquiring a plurality of data points in the filtering window;
fitting the plurality of human body key points and the plurality of data points based on a pre-stored filter algorithm model to obtain a combined matrix aiming at any human body key point;
and obtaining the corresponding posture positions of the plurality of human body key points according to the joint matrix.
2. The method according to claim 1, wherein the setting of the width of the filtering window and the obtaining of the plurality of data points within the filtering window specifically comprises:
the filter window width w is set using the following equation:
w＝2m+1
wherein m is a constant and m is a positive integer;
the multiple data points are each x ═ (-m, -m +1, …, m-1, m).
3. The human body posture estimation method according to claim 2, characterized in that the value range of the constant m is 2-4 when setting the width of the filter window.
4. The human pose estimation method of claim 3, wherein the plurality of human key points and the plurality of data points are fitted using the following formula:
y＝a0+a1x+a2x2+…+ak-1xk-1
Wherein y represents the observed data points to be fitted;
a0，a1，a2、…，ak-1representing polynomial coefficients;
x represents a measurement point;
k denotes the polynomial order.
5. The human body pose estimation method according to claim 4, wherein the joint matrix for any of the human body key points is:
Pi＝M·Ai+Ei
wherein the content of the first and second substances,
Figure FDA0003474062800000021
representing a pose matrix formed by 2m +1 continuous frames of the ith key point;
Figure FDA0003474062800000022
representing a polynomial matrix;
Figure FDA0003474062800000023
a coefficient matrix representing the ith keypoint;
Figure FDA0003474062800000024
the fitting error matrix of the ith keypoint is represented.
6. The method according to claim 5, wherein the plurality of human poses are obtained from the joint matrixPosture position P 'corresponding to human body key point'iComprises the following steps:
P′i＝M·Ai＝M(MTM)-1MTPi；
wherein, PiRepresenting a pose matrix formed by continuous 2m +1 frames of the ith key point;
m represents a polynomial matrix;
Aia coefficient matrix representing the ith keypoint;
t denotes a matrix transpose.
7. An apparatus for estimating a posture of a human body, the apparatus comprising:
the key point extraction unit is used for extracting a plurality of human body key points in the target video frame based on a pre-stored human body posture estimation model;
the data point extraction unit is used for setting the width of a filtering window and acquiring a plurality of data points in the filtering window;
The filtering optimization unit is used for fitting the plurality of human body key points and the plurality of data points based on a pre-stored filtering algorithm model to obtain a joint matrix aiming at any human body key point;
and the attitude position output unit is used for obtaining the attitude positions corresponding to the plurality of human key points according to the joint matrix.
8. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the method according to any of claims 1 to 6 when executing the program.
9. A non-transitory 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 according to any one of claims 1 to 6.
10. A computer program product comprising a computer program, characterized in that the computer program realizes the steps of the method according to any one of claims 1 to 6 when executed by a processor.
CN202210058731.6A 2022-01-17 2022-01-17 Human body posture estimation method and device based on filtering algorithm Pending CN114764819A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210058731.6A CN114764819A (en) 2022-01-17 2022-01-17 Human body posture estimation method and device based on filtering algorithm

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210058731.6A CN114764819A (en) 2022-01-17 2022-01-17 Human body posture estimation method and device based on filtering algorithm

Publications (1)

Publication Number Publication Date
CN114764819A true CN114764819A (en) 2022-07-19

Family

ID=82364753

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210058731.6A Pending CN114764819A (en) 2022-01-17 2022-01-17 Human body posture estimation method and device based on filtering algorithm

Country Status (1)

Country Link
CN (1) CN114764819A (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110666791A (en) * 2019-08-29 2020-01-10 江苏大学 RGBD robot nursing system and method based on deep learning
CN111079695A (en) * 2019-12-30 2020-04-28 北京华宇信息技术有限公司 Human body key point detection and self-learning method and device
WO2020135529A1 (en) * 2018-12-25 2020-07-02 浙江商汤科技开发有限公司 Pose estimation method and apparatus, and electronic device and storage medium
CN112949569A (en) * 2021-03-25 2021-06-11 南京邮电大学 Effective extraction method of human body posture points for tumble analysis
CN113158981A (en) * 2021-05-17 2021-07-23 广东中卡云计算有限公司 Riding posture analysis method based on cascade convolution neural network
CN113255522A (en) * 2021-05-26 2021-08-13 山东大学 Personalized motion attitude estimation and analysis method and system based on time consistency

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2020135529A1 (en) * 2018-12-25 2020-07-02 浙江商汤科技开发有限公司 Pose estimation method and apparatus, and electronic device and storage medium
CN110666791A (en) * 2019-08-29 2020-01-10 江苏大学 RGBD robot nursing system and method based on deep learning
CN111079695A (en) * 2019-12-30 2020-04-28 北京华宇信息技术有限公司 Human body key point detection and self-learning method and device
CN112949569A (en) * 2021-03-25 2021-06-11 南京邮电大学 Effective extraction method of human body posture points for tumble analysis
CN113158981A (en) * 2021-05-17 2021-07-23 广东中卡云计算有限公司 Riding posture analysis method based on cascade convolution neural network
CN113255522A (en) * 2021-05-26 2021-08-13 山东大学 Personalized motion attitude estimation and analysis method and system based on time consistency

Similar Documents

Publication Publication Date Title
WO2020098250A1 (en) Character recognition method, server, and computer readable storage medium
TWI485632B (en) Image alignment method and image alignment system
US8280173B2 (en) Feature point location determination method and apparatus
CN111160232B (en) Front face reconstruction method, device and system
EP2622576A1 (en) Method and apparatus for solving position and orientation from correlated point features in images
CN112733767B (en) Human body key point detection method and device, storage medium and terminal equipment
CN110941989A (en) Image verification method, image verification device, video verification method, video verification device, equipment and storage medium
CN113988112A (en) Method, device and equipment for detecting lane line and storage medium
CN111898571A (en) Action recognition system and method
CN114119777B (en) Stereo matching method and system based on deep learning
CN113592706B (en) Method and device for adjusting homography matrix parameters
CN110956131B (en) Single-target tracking method, device and system
CN114764819A (en) Human body posture estimation method and device based on filtering algorithm
CN109063601B (en) Lip print detection method and device, computer equipment and storage medium
CN112364738A (en) Human body posture estimation method, device, system and medium based on deep learning
KR101515847B1 (en) Device and method for correcting motion data collected by nui device
CN110660091A (en) Image registration processing method and device and photographing correction operation system
CN112989872A (en) Target detection method and related device
US20230069608A1 (en) Object Tracking Apparatus and Method
CN113255700B (en) Image feature map processing method and device, storage medium and terminal
CN115435790A (en) Method and system for fusing visual positioning and visual odometer pose
US20210224947A1 (en) Computer Vision Systems and Methods for Diverse Image-to-Image Translation Via Disentangled Representations
CN109087351B (en) Method and device for carrying out closed-loop detection on scene picture based on depth information
CN115205094A (en) Neural network training method, image detection method and equipment thereof
WO2021087812A1 (en) Method for determining depth value of image, image processor and module

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