CN112184898A - Digital human body modeling method based on motion recognition - Google Patents
Digital human body modeling method based on motion recognition Download PDFInfo
- Publication number
- CN112184898A CN112184898A CN202011131281.6A CN202011131281A CN112184898A CN 112184898 A CN112184898 A CN 112184898A CN 202011131281 A CN202011131281 A CN 202011131281A CN 112184898 A CN112184898 A CN 112184898A
- Authority
- CN
- China
- Prior art keywords
- human body
- joint
- joint points
- modeling method
- points
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T13/00—Animation
- G06T13/20—3D [Three Dimensional] animation
- G06T13/40—3D [Three Dimensional] animation of characters, e.g. humans, animals or virtual beings
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Graphics (AREA)
- Geometry (AREA)
- Software Systems (AREA)
- Length Measuring Devices By Optical Means (AREA)
Abstract
The invention discloses a digital human body modeling method based on action recognition, which belongs to the technical field of digital modeling and comprises the following steps: acquiring three-dimensional depth images of characters and scenes through a camera; separating the human body from the background by performing image processing on the depth image, and identifying the human body part; obtaining coordinates of human joint points according to each human body part, and integrating all the joint points to form a human skeleton model; and selecting specific joint points in the human skeleton model as target characteristics, and calculating the human body action through the target characteristics. The human body image is collected, the human body image part is separated from the background, the skeleton model is built according to the human body part, and the motion state of a person is calculated according to the state of the joint points in the skeleton model, so that the human body can be accurately simulated by the digital modeling method, a medical student can explore the secret of life conveniently, and the medical problem is solved.
Description
Technical Field
The invention relates to the technical field of digital modeling, in particular to a digital human body modeling method based on motion recognition.
Background
The digital human body is a leading-edge subject of world scientific research of combined research of life science and information science, and is an important field of interdisciplinary research. The digital human body mainly digitizes the human body structure through modern high and new technology, explores the secret of life through the digital human body modeling, and solves the medical problem. The digital human body mathematical model refers to mathematical expressions for describing the interaction among human body elements, subsystems and layers and the interaction between a human body system and the environment. The construction of the digital human body mathematical model is a tool for quantitatively and microscopically qualitatively analyzing a human body system, and indicates a direction for the development of medicine. With the acquisition of CT, MRT and milled specimen images, the application of remote sensing technology, a supercomputer and a high-performance workstation, a foundation is laid for constructing a digital human mathematical model. The complex mathematical equations, such as higher-order algebraic equations and nonlinear differential equations, are solved by a computer, and the operation procedure is greatly simplified. The close combination of computational mathematics and high-performance computers is a remarkable characteristic of a digital human body mathematical model. On the basis of an advanced calculation method, scientific graphical operation and a high-speed computer, scientific visualization becomes a big bright point of a digital human body mathematical model.
The conventional digital human body mathematics model building general steps are complex, a complete human body model can be built only by providing a large amount of basic data, the motion posture of the human body is convenient to capture, the current motion of the human body is difficult to accurately show through a digital model, and inconvenience is brought to relevant treatment performed by using the human body model.
Disclosure of Invention
The invention aims to provide a digital human body modeling method based on motion recognition, aiming at solving the problems that the digital human body model is complex in building steps and the built model cannot accurately show corresponding motions of a human body, and the method has the advantages of simple building of a digital human body skeleton model, low use cost and more accurate capturing of the motions of the human body.
The invention realizes the aim through the following technical scheme, and a digital human body modeling method based on motion recognition comprises the following steps:
acquiring three-dimensional depth images of characters and scenes;
identifying a human body part by performing image processing on the depth image;
obtaining coordinates of human joint points according to each human body part, and integrating all the joint points to form a human skeleton model;
and selecting specific joint points in the human skeleton model as target characteristics, and calculating the human body action through the target characteristics.
Preferably, the three-dimensional depth image is acquired through an infrared camera with an infrared emitter, and the infrared emitter is used for projecting a near infrared spectrum, so that speckles are formed and read by the infrared camera after infrared light irradiates a rough object or penetrates frosted glass.
Preferably, the method for three-dimensional depth images of the person and the scene comprises:
A. taking a reference surface at regular intervals and recording speckle patterns of the reference surface;
B. combining the reference surfaces to form a speckle pattern in the whole space, thereby completing light source calibration;
C. and performing cross-correlation operation on the speckle pattern of each object and the speckle pattern of each reference surface respectively, and determining the spatial position and shape of the object according to the operation result.
Preferably, the human body part identification separates the human body from the background through an edge detection algorithm, a noise threshold processing algorithm and an algorithm for extracting target feature points, and identifies the part of the human body part to which the pixel belongs by scanning the pixel of the human body part.
Preferably, the human body joint points are skeleton joint points at the joint of two human body parts, the joint points are divided into three types including trunk joint points, height joint points and limb joint points, and the approximate shape of the skeleton of the human body is determined through the trunk joint points; the height joint point judges the height of the human body through the vertical distance from the head joint to the left foot joint and the right foot joint, and the standing state or sitting state of the human body can be judged through the height joint point; the four-limb joint points are used for determining the motion posture characteristics of the human body.
Preferably, the target feature includes a joint point angle, a joint point vector direction angle, and a joint point position difference, the degree of bending of the joint is calculated by the joint point angle, the position of the joint is calculated by the joint point vector direction angle, and the joint point position difference calculates coordinates of other joint points by selecting one joint point coordinate as a reference coordinate point.
Compared with the prior art, the invention has the beneficial effects that:
1. the human body image is collected, the human body image part is separated from the background, the skeleton model is built according to the human body part, and the motion state of a person is calculated according to the state of the joint points in the skeleton model, so that the human body can be accurately simulated by the digital modeling method, a medical student can explore the secret of life conveniently, and the medical problem is solved.
2. The skeleton joints are classified, and the human body action can be accurately calculated according to the joint point angle and the joint point vector direction angle, so that the human body skeleton model can visually display the human body action, and the interaction requirement of human body rehabilitation training is met.
Drawings
FIG. 1 is a flow chart of a digital human body model construction method of the present invention.
FIG. 2 is a schematic diagram of a human skeleton model formed by the joint points of the present invention.
Fig. 3 is a schematic representation of the joint point of the present invention in a coordinate system.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. 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.
Referring to fig. 1, a digital human body modeling method based on motion recognition includes the following steps:
step S101, acquiring three-dimensional depth images of characters and scenes through a camera, wherein the camera comprises an infrared camera with an infrared emitter, the infrared emitter is used for projecting a near infrared spectrum, so that speckles are formed after infrared light irradiates a rough object or penetrates through frosted glass and are read by the infrared camera, the infrared camera transmits read speckle signals to an image processing chip, and the method for processing the images by the image processing chip to obtain the three-dimensional depth images of the characters and the scenes comprises the following steps:
A. a reference surface is taken every 5mm distance, speckle patterns of the reference surface are recorded, and as the narrower the spacing is, the more dense the speckle patterns are, the more fine the formed three-dimensional image is, but the more difficult the data processing is, the best the reference surface is taken at the spacing of 3mm-8mm according to the image processing capacity;
B. combining the reference surfaces to form a speckle pattern in the whole space, thereby completing light source calibration;
C. because of the randomness of laser speckles, the speckle pattern of each object in the space is different, the speckle pattern of each object is respectively subjected to cross-correlation operation with the speckle pattern of each reference surface, and the space position and the shape of the object can be determined according to the operation result.
Step S102, separating the human body from the background by processing the depth image, and identifying the human body part, wherein the human body and the background are separated from each other by an edge detection algorithm, a noise threshold processing algorithm and an algorithm for extracting target characteristic points, the human body part is identified by scanning pixels of the human body part to judge the part of the human body part to which the pixels belong, the pixels of each human body part are different, standard human body pixels are stored in a database, when the pixel image of each part of the human body in the depth image is scanned, the standard pixels are compared with the standard pixels in the database, the probability of the corresponding human body part is judged, if a certain pixel has a probability of 80% being a foot, a probability of 50% being a head and a probability of 20% being a hand, the human body part to which the pixel belongs is not directly judged by the maximum probability, but all the possible human body parts are input into the next processing step, and judging until all the body parts are completely identified, selecting the probability with highest probability and no repetition to determine the body parts, and judging the body parts according to the body part pixels of the adjacent parts of the pixels if the probability with repetition exists.
Step S103, acquiring coordinates of human body joint points according to each human body part, and integrating all the joint points to form a human body skeleton model;
and step S104, selecting specific joint points in the human skeleton model as target characteristics, and calculating human body actions according to the target characteristics.
The human body joint points are skeleton joint points at the joint of two human body parts, the joint points are divided into three types including trunk joint points, height joint points and four limb joint points, and the approximate shape of the human skeleton is determined through the trunk joint points; the height joint point judges the height of the human body through the vertical distance from the head joint to the left foot joint and the right foot joint, and the standing state or sitting state of the human body can be judged through the height joint point; the four-limb joint points are used for determining the motion posture characteristics of a human body, as shown in fig. 2, the trunk joint points mainly comprise eight joint points such as a head, a left shoulder, a right shoulder, a spine, a left hip and a right hip, the eight joint points respectively comprise joint points 1.2.3.4.11.12.13.14, the height joint points can calculate the height of the human body through the distance from the joint point 1 to the joint point 19 and the joint point 20, the four-limb joint points comprise joint points 5.6.7.8.9.10.15.16.17.18.19.20, the target characteristics comprise joint point angles, joint point vector direction angles and joint point position differences, the bending degree of the joint is calculated through the joint point angles, the positions of the joints are calculated through the joint point vector direction angles, and the joint point position differences calculate the coordinates of other joint points by selecting one joint point coordinate as a reference coordinate point. As shown in FIG. 3, the motion of raising the left arm selects the elbow joint p6And the wrist joint p8Vector p of doing relationship nodes8p6The vector makes included angles alpha, beta and gamma with the X axis, the Y axis and the Z axis, and the space position of the joint can be determined by the included angles.
The joint point vector can be expressed as
p8p6＝(x6-x8)i+(y6-y8)j+(z6-z8)k
The vector direction angle can be expressed as
Angle to the X-axis:
angle to Y axis:
included angle with Z axis:
in this way, the spatial position of each segment of the joint can be uniquely determined.
The position difference of the joint point is normalized by selecting a certain joint as a normalized joint reference, and if a joint point p is selected11As reference joint point, joint point p2To p11The length of the joint is used as a normalized reference, and the rest joint points can be expressed as
x′i＝xi-x11 1≤i≤20
y′j＝yi-y11 1≤j≤20
z′k＝zk-z11 1≤k≤20
Thus, new position coordinates (x ', y', z ') of 20 joint points can be obtained, and the spatial position difference of the joint points is D ═ x'1，y′1，z′1，...，x′20，y′20，z′20) Articulation point p2To p11Can be expressed as
The unified skeletal joint point can be represented as
Thus, the standardization of the joint point coordinates is realized.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned.
Furthermore, it should be understood that although the present description refers to embodiments, not every embodiment may contain only a single embodiment, and such description is for clarity only, and those skilled in the art should integrate the description, and the embodiments may be combined as appropriate to form other embodiments understood by those skilled in the art.
Claims (6)
1. A digital human body modeling method based on motion recognition is characterized by comprising the following steps:
acquiring three-dimensional depth images of characters and scenes;
identifying a human body part by performing image processing on the depth image;
obtaining coordinates of human joint points according to each human body part, and integrating all the joint points to form a human skeleton model;
and selecting specific joint points in the human skeleton model as target characteristics, and calculating the human body action through the target characteristics.
2. The digital human body modeling method based on motion recognition according to claim 1, wherein the three-dimensional depth image is acquired by an infrared camera with an infrared emitter, and the infrared emitter is used for projecting a near infrared spectrum, so that infrared light irradiates a rough object or ground glass, forms speckles and is read by the infrared camera.
3. The digital human body modeling method based on motion recognition as claimed in claim 2, wherein the method of three-dimensional depth image of the person and the scene is:
A. taking a reference surface at regular intervals and recording speckle patterns of the reference surface;
B. combining the reference surfaces to form a speckle pattern in the whole space, thereby completing light source calibration;
C. and performing cross-correlation operation on the speckle pattern of each object and the speckle pattern of each reference surface respectively, and determining the spatial position and shape of the object according to the operation result.
4. The digital human body modeling method based on motion recognition according to claim 1, wherein the human body part recognition separates the human body from the background through an edge detection algorithm, a noise threshold processing algorithm and an algorithm for extracting target feature points, and the recognition of the human body part judges the part of the human body part to which the pixel belongs by scanning the pixel of the human body part.
5. The digital human body modeling method based on motion recognition as claimed in claim 1, wherein the human body joint points are skeleton joint points at the connection of two human body parts, the joint points are divided into three categories including a trunk joint point, a height joint point and a limb joint point, and the trunk joint point is used for determining the approximate shape of the human skeleton; the height joint point judges the height of the human body through the vertical distance from the head joint to the left foot joint and the right foot joint, and the standing state or sitting state of the human body can be judged through the height joint point; the four-limb joint points are used for determining the motion posture characteristics of the human body.
6. The digital human body modeling method based on motion recognition as claimed in claim 1, wherein the target feature comprises a joint point angle by which a degree of bending of the joint is calculated, a joint point vector direction angle by which a position of the joint is calculated, and a joint point position difference by which coordinates of other joint points are calculated by selecting one joint point coordinate as a reference coordinate point.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011131281.6A CN112184898A (en) | 2020-10-21 | 2020-10-21 | Digital human body modeling method based on motion recognition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011131281.6A CN112184898A (en) | 2020-10-21 | 2020-10-21 | Digital human body modeling method based on motion recognition |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112184898A true CN112184898A (en) | 2021-01-05 |
Family
ID=73923162
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011131281.6A Pending CN112184898A (en) | 2020-10-21 | 2020-10-21 | Digital human body modeling method based on motion recognition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112184898A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112990089A (en) * | 2021-04-08 | 2021-06-18 | 重庆大学 | Method for judging human motion posture |
| CN115797560A (en) * | 2022-11-28 | 2023-03-14 | 广州市碳码科技有限责任公司 | Head model construction method and system based on near infrared spectrum imaging |
| CN116563939A (en) * | 2023-03-20 | 2023-08-08 | 南通锡鼎智能科技有限公司 | Experimenter nonstandard behavior detection method and device based on depth information |
-
2020
- 2020-10-21 CN CN202011131281.6A patent/CN112184898A/en active Pending
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112990089A (en) * | 2021-04-08 | 2021-06-18 | 重庆大学 | Method for judging human motion posture |
| CN112990089B (en) * | 2021-04-08 | 2023-09-26 | 重庆大学 | Method for judging human motion gesture |
| CN115797560A (en) * | 2022-11-28 | 2023-03-14 | 广州市碳码科技有限责任公司 | Head model construction method and system based on near infrared spectrum imaging |
| CN115797560B (en) * | 2022-11-28 | 2023-07-25 | 广州市碳码科技有限责任公司 | Near infrared spectrum imaging-based head model construction method and system |
| CN116563939A (en) * | 2023-03-20 | 2023-08-08 | 南通锡鼎智能科技有限公司 | Experimenter nonstandard behavior detection method and device based on depth information |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111126272A (en) | Posture acquisition method, and training method and device of key point coordinate positioning model | |
| CN112184898A (en) | Digital human body modeling method based on motion recognition | |
| JP6560480B2 (en) | Image processing system, image processing method, and program | |
| CN108629946B (en) | Human body falling detection method based on RGBD sensor | |
| CN108256504A (en) | A kind of Three-Dimensional Dynamic gesture identification method based on deep learning | |
| JP7015152B2 (en) | Processing equipment, methods and programs related to key point data | |
| CN107357427A (en) | A kind of gesture identification control method for virtual reality device | |
| CN109758756B (en) | Gymnastics video analysis method and system based on 3D camera | |
| CN113239797B (en) | Human body action recognition method, device and system | |
| JP7164045B2 (en) | Skeleton Recognition Method, Skeleton Recognition Program and Skeleton Recognition System | |
| CN110060284A (en) | A kind of binocular vision environmental detecting system and method based on tactilely-perceptible | |
| WO2023078169A1 (en) | Medical image mark point recognition method and apparatus, electronic device, and storage medium | |
| CN113658211A (en) | User posture evaluation method and device and processing equipment | |
| CN111652047A (en) | Human body gesture recognition method based on color image and depth image and storage medium | |
| CN113229807A (en) | Human body rehabilitation evaluation device, method, electronic device and storage medium | |
| Angladon et al. | The toulouse vanishing points dataset | |
| CN104732586A (en) | Fast reconstruction method for three-dimensional human body dynamic form and fast construction method for three-dimensional movement light stream | |
| CN114273826A (en) | Automatic identification method for welding position of large-sized workpiece to be welded | |
| Cai et al. | Tracking multiple surgical instruments in a near-infrared optical system | |
| WO2023159517A1 (en) | System and method of capturing three-dimensional human motion capture with lidar | |
| Jain et al. | Human computer interaction–Hand gesture recognition | |
| Munn et al. | FixTag: An algorithm for identifying and tagging fixations to simplify the analysis of data collected by portable eye trackers | |
| Westfeld et al. | Automatic techniques for 3D reconstruction of critical workplace body postures from range imaging data | |
| Chen et al. | An integrated sensor network method for safety management of construction workers | |
| Wu et al. | Explore on Doctor's Head Orientation Tracking for Patient's Body Surface Projection Under Complex Illumination Conditions |
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 |