A digital evaluation method and system for individual soldier's bare-handed formation movements

Abstract

The present invention discloses a digital evaluation method and system for individual soldier's bare-handed queue action, which respectively determines the morphological variables of the tested person according to the collected three-dimensional point cloud data in posture A and the set standard queue action posture. β and posture variables i Based on the morphological and posture variables, a three-dimensional parametric mesh model of the human body of the person being tested is reconstructed, and the position information of each end joint point and skeleton joint point of the person being tested in each posture is calculated to obtain a simplified skeleton representation; based on the simplified skeleton representation, the standard queue action quantization rules are compared to obtain a digital evaluation result. The present invention obtains three-dimensional point cloud data of the human body in various postures through a non-contact method, establishes a three-dimensional digital model of an individual soldier, and calculates a simplified skeleton representation to make the positioning of the joints and skeleton more accurate; by comparing with the standard queue action quantization rules, the standardization degree of the individual soldier's queue action is intelligently evaluated.

Description

Technical Field

[0001] The present invention relates to the field of computer graphics, and in particular to a digital evaluation method and system for individual soldier's bare-handed formation movements. Background Art

[0002] The standardization of individual soldier formation movements plays a crucial role in overall formation effectiveness. Currently, formation training in my country relies primarily on individuals and simple auxiliary tools. Different individuals may have different understandings of standard movements. Furthermore, as training progresses, fatigue among trainers increases, leading to variations in training standards. Auxiliary tools are primarily categorized as contact and quasi-contact. Contact tools include wooden supports for participants to maintain an upright upper body and arm swings to achieve a uniform fist angle. These tools can correct specific movements of parade participants, but long-term use can cause discomfort. Quasi-contact tools include straight lines for measuring kick height and standardizers for feet at ease. These tools establish critical thresholds for movement standards and do not involve direct contact with individuals, but they can also cause psychological discomfort to trainers, leading to distortions in movement. More importantly, none of these auxiliary devices provide a quantitative assessment of individual soldier formation movements, relying instead on subjective judgment by trainers or participants. Summary of the Invention

[0003] The present invention aims to solve the technical problem that the current use of auxiliary equipment for individual soldier formation movement training does not provide quantitative evaluation of the individual soldier's formation movement, and still requires trainers or tested personnel to make subjective judgments. A digital evaluation method and system for individual soldier's bare-hand formation movement is provided.

[0004] In order to achieve the above technical objectives, the present invention adopts the following technical solutions.

[0005] In a first aspect, the present invention provides a digital evaluation method for individual soldier's bare-handed formation movement, comprising: collecting three-dimensional point cloud data of the measured person in posture A and in a set standard formation movement posture;

[0006] According to the three-dimensional point cloud data of the measured person in posture A and the set standard queue action posture, the morphological variables β and posture variables θ are determined respectively; according to the morphological variables β and posture variables θ, the three-dimensional human body parameter grid model of the measured person is reconstructed, and the three-dimensional human body parameter grid model expresses the human body morphological variables β, posture variables θ, constant Φ to the three-dimensional human body space R 3 The mapping relationship;

[0007] Based on the three-dimensional human body parametric mesh model, the position information of the end joints and skeleton joints of the subject in each posture is calculated to obtain a simplified skeleton representation;

[0008] The digital evaluation results are obtained by comparing the simplified skeleton representation with the standard queue action quantification rules.

[0009] Furthermore, based on the three-dimensional human body parametric mesh model, the position information of each end joint point and the skeleton joint point of the measured person in each posture is calculated, including:

[0010] Based on the morphological variables β of the person being measured and the corresponding three-dimensional human body parameterized mesh model in posture A, the positions of the skeleton joints in posture A are determined using a linear regression model;

[0011] Determine the position of the end joint of the person being measured in posture A;

[0012] According to the positions of the skeleton joints and terminal joints of the person being tested in posture A, combined with the three-dimensional parametric mesh model of the human body in the set standard queue action posture, the rigid transformation corresponding to each rigid bone is calculated, and the position information of the terminal joints and skeleton joints in each posture is obtained.

[0013] Furthermore, the linear regression model is expressed as:

[0014]

[0015] where a β is a constant, β i Represents the i-th dimension element of the morphological variable vector, a i is the constant parameter corresponding to the i-th dimension element of the morphological variable vector;

[0016] The training method of the linear regression model includes:

[0017] For a certain person’s skeleton joint point j, assuming that the position of the skeleton joint point j in posture A is p j , the bones associated with skeleton joint point j are b1 and b2;

[0018] The position p of the skeleton joint point j in posture A is obtained by solving the following equation j :

[0019] Where K is the number of postures in the training set, is the three-dimensional rotation matrix of bone b1 in posture k, is the three-dimensional translation vector of bone b1 in posture k, is the three-dimensional rotation matrix of bone b2 in posture k, is the three-dimensional translation vector of bone b2 in posture k; sym(j) represents the symmetrical joint of skeleton joint point j, p sym(j)represents the position of the symmetrical joint of the skeleton joint point j in posture A; sym(b1) and sym(b2) represent the symmetrical bones of bone b1 and bone b2 respectively, and M is the symmetry transformation matrix, which is expressed as:

[0020]

[0021] is the 3D rotation matrix of the bone symmetrical to bone b1 in posture k, is the 3D rotation matrix of the bone symmetrical to bone b2 in posture k, is the three-dimensional translation vector of the bone symmetrical to bone b1 in posture k, is the three-dimensional translation vector of the bone symmetrical to bone b2 in posture k; α is a setting parameter.

[0022] By solving the above equations, we can obtain the morphological variables β and the position p of the skeleton joint point j under posture A. j , thereby constructing a linear regression model training set;

[0023] The linear regression model is trained using the training set to determine the constant parameter corresponding to the i-th dimension element of the morphological variable β (vector), and the position p of the skeleton joint point j in the A posture is obtained. j and the linear relationship between the morphological variables β.

[0024] Furthermore, the simplified skeleton model of the three-dimensional human body is represented by S={J,B}, J={j1,...,j |J|} represents the set of joint points, B={b1,...,b |B|} represents a set of bones connected by adjacent joints.

[0025] Furthermore, determining the position of the end joint point of the measured person in posture A includes: determining the center point of all adjacent points within a set distance from the end joint point as the position of the end joint point.

[0026] Furthermore, the position of the skeleton joint point j after rigid transformation is expressed as:

[0027]

[0028] where p j ′ is the position of the skeleton joint point j after rigid transformation, is the three-dimensional rotation matrix of bone b1 at the posture variable θ, is the three-dimensional translation vector of bone b1 at the posture variable θ, is the three-dimensional rotation matrix of bone b2 at the posture variable θ, is the three-dimensional translation vector of bone b2 at the posture variable θ.

[0029] Furthermore, the standard queue movement postures are set to include standing at attention posture, left hand in front marching posture, right hand in front marching posture, left hand in front marching posture, right hand in front marching posture and saluting posture.

[0030] Furthermore, the end joints include the top of the head, toes and fingertips.

[0031] In a second aspect, the present invention provides a digital evaluation system for individual soldier's bare-handed formation movements, comprising an image acquisition module, a three-dimensional human body modeling module, a simplified skeleton representation module, and a digital evaluation module;

[0032] The image acquisition module is used to collect three-dimensional point cloud data of the measured person in posture A and the set standard queue action posture;

[0033] The three-dimensional human body modeling module is used to determine the morphological variables β and posture variables θ of the measured person according to the three-dimensional point cloud data in posture A and the set standard queue action posture; reconstruct the three-dimensional human body parameter grid model of the measured person according to the morphological variables β and posture variables θ, and the three-dimensional human body parameter grid model expresses the human body morphological variables β, posture variables θ, constant Φ to the three-dimensional human body space The mapping relationship;

[0034] The simplified skeleton representation module is used to calculate the position information of the end joints and skeleton joints of the measured person in various postures based on the three-dimensional human body parameterized mesh model to obtain the simplified skeleton representation;

[0035] The digital evaluation module is used to obtain digital evaluation results by comparing the standard queue action quantification rules based on the simplified skeleton representation.

[0036] In a third aspect, the present invention further provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the method provided in any possible implementation manner of the first aspect.

[0037] The beneficial technical effects achieved by the present invention are as follows: the present invention obtains three-dimensional point cloud data of the human body in various postures through a non-contact method, establishes a three-dimensional digital model of a single soldier, and calculates a simplified skeleton representation to make the positioning of joints and skeletons more accurate; by comparing the quantitative rules of standard queue movements, the standardization degree of individual soldier queue movements is intelligently evaluated. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 A flowchart of a digital evaluation method provided for a specific embodiment;

[0039] Figure 2Schematic diagram of a three-dimensional human body parametric mesh model and a simplified skeleton representation in a specific embodiment;

[0040] Figure 3 A schematic diagram of display results of a digital evaluation module in a digital evaluation system provided in a specific embodiment;

[0041] Figure 4 To obtain the standing posture evaluation result of the person to be evaluated in a specific embodiment;

[0042] Figure 5 To obtain the goose-step posture evaluation result of the person to be evaluated in the specific embodiment;

[0043] Figure 6 To obtain the assessment results of the marching posture of the person to be assessed in a specific embodiment;

[0044] Figure 7 To obtain the evaluation result of the saluting posture of the person to be evaluated in a specific embodiment. DETAILED DESCRIPTION

[0045] The 3D human body can be represented mathematically. Assume that a single 3D human body is discretized and represented as a triangle mesh, including vertices V = {v1,...,v |V|},E={e1,...,e |E|} and face F={f1,...,f |F| A 3D human body parametric mesh model is constructed using a triangular mesh. The 3D human body parametric mesh model is a clean body model of the human body. Based on the original point cloud data collected from multiple perspectives, the specific embodiment comprehensively utilizes prior knowledge such as height and weight, skin detection, and the location of the clean body model inside clothing to reconstruct the 3D human body parametric mesh model. It expresses the human body related parameters into three-dimensional human body space Mapping Among them, the morphological variable β is the morphological space parameter, the posture variable θ is the posture space parameter, and the constant Φ is the parameter pre-learned through data training. 3D human body parametric mesh model The three-dimensional human body is described by low-dimensional features. Here, the low-dimensional features are morphological variables β (height, weight, measurements, etc., a total of 42 dimensions) and posture variables θ (a set of Rodriguez rotation vectors of rigid parts such as the trunk, upper arms, and lower arms, a total of 19 vectors). Φ is the linear coefficient matrix mapped from the low-dimensional space to the three-dimensional human body, R 3 Space refers to the vertex position V of the 3D human body model. 3D human body parametric mesh model The vertices on the image are clustered according to the bones closest to them.

[0046] The mesh model generated by the 3D human body parametric mesh model is an effective method for digital human body representation and can be formally expressed as:

[0047] The Φ parameter is pre-trained. The morphological deformation D(β) sub-mathematical model captures the diversity of 3D human morphology. The posture deformation P(θ) and Q(θ) sub-mathematical models capture the diversity of 3D human posture.

[0048] It should be noted that the specific method for constructing the three-dimensional human body parametric mesh model has been disclosed in the previous patent application of the inventor of the present application and is prior art, so it will not be repeated here.

[0049] The technical solution of the present invention will be further described below in conjunction with the accompanying drawings, and the following embodiments are only used to illustrate the present invention more clearly.

[0050] Example 1: Figure 1 As shown, a digital evaluation method for individual soldier's bare-handed formation action includes: collecting three-dimensional point cloud data of the tested person in posture A and in a set standard formation action posture;

[0051] According to the three-dimensional point cloud data of the measured person in posture A and the set standard queue action posture, the morphological variables β and posture variables θ are determined respectively; according to the morphological variables β and posture variables θ, the three-dimensional human body parametric mesh model of the measured person is reconstructed, and the three-dimensional human body parametric mesh model expresses the human body morphological variables β, posture variables θ, constant Φ to the three-dimensional human body space The mapping relationship;

[0052] Based on the three-dimensional human body parametric mesh model, the position information of the end joints and skeleton joints of the measured person in each posture is calculated to obtain a simplified skeleton representation;

[0053] The digital evaluation results are obtained by comparing the simplified skeleton representation with the standard queue action quantification rules.

[0054] In this embodiment, a multi-view structured light scanner is used to collect point cloud data of the measured person.

[0055] The morphological variable β of the person being tested is calculated by combining prior knowledge such as height and weight, skin detection, and the location of the 3D human body parametric model inside clothing. This specific calculation method has been disclosed in a previous invention patent application filed by the inventor of this application and is prior art. In short, it involves solving a parametric 3D human body mesh model that meets the constraints of height and weight, skin detection, and location inside clothing, and will not be further described here.

[0056] In this embodiment, the standard queue action postures are set to include standing at attention posture, left hand in front marching posture, right hand in front marching posture, left hand in front marching posture, right hand in front marching posture and saluting posture.

[0057] Under the condition that the morphological variable β is determined, the posture variable θ under each posture is determined according to the 3D point cloud data of the human body in the other 6 postures, and the 3D human body parametric mesh model is reconstructed. The human body model is a joint model, which can be divided into rigid parts such as the torso, upper arms, and forearms. Changes in human posture can be considered to be caused by the combined action of these rigid parts. The rigid transformation of each rigid part can be expressed in the form of the Rodriguez vector of the rotation matrix. θ represents the set of Rodriguez vectors of all rigid parts.

[0058] Calculating the position information of the end joints and skeleton joints of the subject in each posture to obtain a simplified skeleton representation, including: determining the positions of the skeleton joints in posture A using a linear regression model based on the morphological variables β of the subject and the corresponding three-dimensional human body parameterized mesh model in posture A;

[0059] Determine the position of the end joint of the person being measured in posture A;

[0060] According to the positions of the skeleton joints and terminal joints of the person being tested in posture A, combined with the three-dimensional parametric mesh model of the human body in the set standard queue action posture, the rigid transformation corresponding to each rigid bone is calculated to obtain the position information of each terminal joint and skeleton joint of the person being tested in each posture.

[0061] In this invention, joints are divided into two categories: one is the end node (i.e., the end joint point), denoted as J1, such as the top of the head, toes, and fingertips. For this type of joint, since the ends are all rigid transformations, their positions can be set as the center points of adjacent vertices on the three-dimensional model; the other is the intermediate node (i.e., the skeleton joint point), denoted as J2, which is shared by more than two bones (Bone), such as the elbow joint and knee joint. For this type of joint, associated bones can be defined for it.

[0062] 3D human body parametric mesh model The corresponding simplified skeleton is represented as S = {J, B}, J = {j1,...,j |J|} represents the set of joint points, B={b1,...,b |B|} represents a set of bones connected by adjacent joints, such as Figure 2 As shown, Figure 2 The posture shown is posture A, which is a normalized posture. The different human body shapes in other postures are mainly caused by different posture variables θ.

[0063] Different people have different morphologies and different skeleton positions. In order to characterize the relationship between the skeleton in pose A and the morphological variables β, it is necessary to establish a training dataset to train a linear regression model. The training process includes:

[0064] For an individual with morphological variables β, the 3D human body parameterized mesh model of the A pose θ A The posture variable corresponding to posture A is represented by the aforementioned 3D human body parameterized mesh model to generate 72 3D models in different postures.

[0065] For the J2 type node, since it is located in the connection area of ​​multiple bones, it produces non-rigid deformation when the posture deforms. Therefore, setting its position similar to the J1 type node will produce a large error. Therefore, the following calculation method is used. Taking the joint j in J2 as an example, assuming that its position in posture A is p j , the associated bones are b1, b2, for any posture k, b1 and b2 correspond to the rigid transformation and It can be based on and The divided vertices are calculated directly, because b1 and b2 share joint j, so the positions of b1 and b2 after rigid transformation of j are the same, that is, There are currently 72 different postures. Considering the symmetry of the left and right joints, the position p of joint j in posture A can be obtained by solving the following equation: j :

[0066]

[0067] Among them, sym(j) represents the symmetric joint of joint j, sym(b1) and sym(b2) represent the symmetric bones of b1 and b2, and M is the symmetric transformation matrix.

[0068] α is an artificially set parameter used to constrain the symmetrical joint point positions to meet certain conditions and can be set to 100.

[0069] By solving the above equations, we can obtain the morphological variables β of different soldiers and the position p of the skeleton joint point j in posture A. j , thereby constructing the model training set.

[0070] Set p j The mathematical model between and β is a simple linear regression model, that is where a β Is a constant, then the model constant parameter a can be calculated through the training set i For other J2 type nodes, a mathematical model that changes with the morphological space parameter β can also be trained similarly.

[0071] In this embodiment, according to the current soldier's morphological variable β and the corresponding A posture three-dimensional model The linear regression model obtained by the above training is used to calculate the positions of all the nodes of the J2 class under the A posture; the positions of the joint points of the J1 class are set to The upper distance from the end joint point sets the center point of all adjacent points within the distance as its position.

[0072] Calculate the simplified skeleton representation of a soldier in different postures, including: the model of the current soldier (personnel to be tested) under the posture variable θ and the morphological variable β And the corresponding A posture 3D model Calculate the rigid transformation corresponding to each rigid bone. For the J2 joint point j, its position p in the A posture j It has been calculated. Let the rigid transformation corresponding to the associated bones b1 and b2 be and Then the position at the attitude variable θ is equal to the average of the transformed positions of b1 and b2: The position of the J1 joint point can be transformed in posture A by using the rigid transformation of the bone where it is located.

[0073] Standard queue action quantification rules include:

[0074] (1) Stand at attention

[0075] The action standards set out in the regulations are as follows: Figure 4 As shown: the heels are close together and aligned, and the toes are separated by about 60 degrees; the legs are straight; the lower abdomen is slightly retracted, and the chest is naturally raised; the upper body is upright and slightly tilted forward; the shoulders are level and slightly stretched back; the arms are hanging down and stretched naturally, the fingers are together and naturally slightly bent, the tip of the thumb is placed on the second section of the index finger, and the middle finger is placed on the trouser seam; the head is upright, the neck is straight, the mouth is closed, the lower jaw is slightly retracted, and the eyes are looking straight forward.

[0076] Toe separation: the angle between the left foot (toe to ankle) and the right foot (toe to ankle). Toe separation value = 1 - |angle - 60| / 60;

[0077] Anterior-posterior tilt: the angle between the upper body (the line segment between the top of the head and the hip nodes) and the coronal plane; Anterior-posterior tilt value = 1-angle / 60 (if the angle is > 60, then 60 degrees is used);

[0078] Left-right sway: The angle between the upper body (the line segment between the top of the head and the hip nodes) and the sagittal plane; left-right sway value = 1-angle / 60 (if the angle is > 60, use 60 degrees);

[0079] Shoulder height difference: the height difference between the left and right shoulder points; if the height difference is less than or equal to 2 cm, the shoulder height difference is 1, otherwise it is 0;

[0080] The quantitative value of standing at attention = (the distance between the two toes + the front and back tilt value + the left and right sway value + the difference in shoulder height) / 4.

[0081] The definitions of marching (left hand in front), marching (right hand in front), marching in step (left hand in front), marching in step (right hand in front), and saluting are similar and will not be repeated here.

[0082] In a specific embodiment, a quantitative value is calculated based on the results obtained by comparing the standard queue action quantification rules, and its comprehensive evaluation score is given, which is displayed in the form of a three-dimensional model and a two-dimensional report, and quantitative training improvement suggestions are put forward.

[0083] In this embodiment, the evaluation results of the personnel to be tested are as follows: Figures 4 to 7 As shown, Figure 4 Corresponding standing posture assessment results.

[0084] Key points for standing at attention: keep your heels close together and aligned, and your toes pointed outward about 60 degrees; keep your legs straight; slightly retract your abdomen and naturally lift your chest; keep your upper body upright and slightly lean forward; keep your shoulders level and slightly stretched back; let your arms hang naturally straight, keep your fingers together and slightly bent, with the tip of your thumb touching the second joint of your index finger and your middle finger touching the crotch of your trousers; keep your head upright, your neck straight, your mouth closed, your jaw slightly retracted, and your eyes looking straight ahead.

[0085] The results of this assessment are: the angle between the two feet is 25.6 degrees; the body is tilted 1.8 degrees forward and backward, and 2.5 degrees left and right; the height difference between the two shoulders is 0.5 cm when standing at attention. The assessment result of standing at attention is: 83.90 points.

[0086] Figure 5 The corresponding goose step posture evaluation results include the left-hand-in-front goose step posture evaluation results and the right-hand-in-front goose step posture evaluation results.

[0087] Essentials of marching posture: Kick the front foot forward about 75 cm (keep the legs straight, toes pressed down, soles of the feet parallel to the ground, about 30 cm from the ground); keep the upper body upright and slightly lean forward: when swinging the arm forward, bend the elbow, make the forearm slightly horizontal, palms facing inward and slightly downward, and the lower edge of the wrist should be about 15 cm higher than the lowest button of the spring and autumn uniform, and about 10 cm away from the body: when swinging the arm backward, the front of the wrist should be about 30 cm away from the trouser seam.

[0088] The results of this assessment are as follows: the right foot is 9.2 cm from the ground; the angle between the right foot and the horizontal plane is 14.1 degrees; the angle between the left forearm and the horizontal plane is 3.0 degrees; the angle between the left upper arm and the left forearm is 87.3 degrees; the body is tilted 4.1 degrees forward and backward, and 0.4 degrees left and right; the right hand is 25.3 cm away from the trouser line; the left marching (i.e., the left hand is in front of the marching posture) assessment result is 70.03 points.

[0089] The results of this assessment are: the left foot is 6.5 cm from the ground; the angle between the left foot and the horizontal plane is 14.4 degrees; the angle between the right forearm and the horizontal plane is 1.1 degrees; the angle between the right upper arm and the right forearm is 84.2 degrees; the body is tilted 14.4 degrees forward and backward, and 6.4 degrees left and right; the left hand is 22.5 cm from the trouser line, and the right marching posture (that is, the right hand is in front of the marching posture) assessment result is: 64.60 points.

[0090] Figure 6 The corresponding marching posture evaluation results include the left-hand-in-front marching posture evaluation results and the right-hand-in-front marching posture evaluation results.

[0091] Key points of marching posture: When swinging your arms forward, bend your elbows, naturally close your forearms inward, turn your palms inward and slightly downward, align the base of your thumbs with the button line, and be about 5 cm higher than the lowest button of your spring and autumn uniform, and about 30 cm away from your body; when swinging your arms backward, straighten your arms naturally, and keep the front of your wrists about 30 cm away from the trouser seam.

[0092] The results of this assessment are: the left wrist is 0.6 cm away from the midline; the left wrist is 0.7 cm away from the body; the right wrist is 0.9 cm away from the body; the body is tilted 0.9 degrees forward and backward, and 1.0 degrees left and right. The assessment result of left march (i.e., the left hand is in front of the marching posture) is 81.86 points.

[0093] The results of this assessment are: the right wrist is 0.6 cm away from the midline; the right wrist is 0.9 cm away from the body; the left wrist is 0.7 cm away from the body; the body is tilted 0.9 degrees forward and backward, and 1.0 degrees left and right. The assessment result of right march (i.e., marching with the right hand in front) is 87.43 points.

[0094] Figure 7 Corresponding salute posture assessment results.

[0095] Key points of saluting posture: Keep your upper body upright, raise your right hand quickly, put your five fingers together and stretch them naturally, with your middle finger slightly touching the right corner of the hat brim about 2 cm in front, palm facing down, slightly outward (about 20 degrees), the wrist must not be bent, and the right upper arm should be slightly horizontal, in line with the shoulders.

[0096] The test results are: the angle between the two feet is 25.6 degrees; the body is tilted 1.8 degrees forward and backward, and 2.5 degrees left and right; the height difference between the two shoulders is 0.5 cm. The salute posture test result is 83.90 points.

[0097] The present invention uses non-contact structured light to scan the human body from multiple perspectives, establishes a three-dimensional digital model of an individual soldier, calculates a simplified skeleton representation, defines a quantitative evaluation method for queue movements, and intelligently evaluates the degree of standardization of individual soldier queue movements.

[0098] Corresponding to the digital evaluation method for individual soldier's bare-handed formation movement provided in the above embodiment, embodiment 2 provides a digital evaluation system for individual soldier's bare-handed formation movement, comprising: an image acquisition module, a 3D human body modeling module, a simplified skeleton representation module, and a digital evaluation module;

[0099] The image acquisition module is used to collect three-dimensional point cloud data of the measured person in posture A and the set standard queue action posture;

[0100] The three-dimensional human body modeling module is used to determine the morphological variables β and posture variables θ of the measured person according to the three-dimensional point cloud data in posture A and the set standard queue action posture; reconstruct the three-dimensional human body parameter grid model of the measured person according to the morphological variables β and posture variables θ, and the three-dimensional human body parameter grid model expresses the human body morphological variables β, posture variables θ, constant Φ to the three-dimensional human body space The mapping relationship;

[0101] The simplified skeleton representation module is used to calculate the position information of each end joint point and skeleton joint point of the measured person in each posture based on the three-dimensional human body parameterized mesh model to obtain a simplified skeleton representation of the measured person;

[0102] The digital assessment module is used to compare the simplified skeleton representation of the person being measured with the standard queue action quantification rules to obtain the digital assessment results.

[0103] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0104] This embodiment further provides a computer-readable storage medium, which stores a computer program, and is characterized in that when the computer program is executed by a processor, the steps of the method described in the above embodiment are implemented.

[0105] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0106] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0107] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0108] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0109] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the purpose of the present invention and the claims, which are all protected by the present invention.

Claims

1. A digital evaluation method for individual soldier's bare-handed formation action, characterized by: include: Collect 3D point cloud data of the person being measured in posture A and in the set standard queue action posture; Determine the morphological variable β and posture variable θ of the person being measured in posture A and in the set standard queue action posture according to the three-dimensional point cloud data; The three-dimensional human body parameter grid model of the measured person is reconstructed according to the morphological variables β and the posture variables θ. The three-dimensional human body parameter grid model describes the human body morphological variables β, the posture variables β, the constant To the three-dimensional human body space The mapping relationship; Based on the three-dimensional human body parametric mesh model, the position information of each end joint and skeleton joint of the measured person in each posture is calculated to obtain a simplified skeleton representation; According to the simplified skeleton representation, the standard queue action quantification rules are compared to obtain the digital evaluation results; Based on the 3D human body parametric mesh model, the position information of each end joint and skeleton joint of the subject in each posture is calculated to obtain its simplified skeleton representation, including: Based on the morphological variables β of the person being measured and the corresponding three-dimensional human body parameterized mesh model in posture A, the positions of the skeleton joints in posture A are determined using a linear regression model; Determine the position of the end joint of the person being measured in posture A; Based on the positions of the skeleton joints and terminal joints of the person being tested in posture A, combined with a three-dimensional parametric mesh model of the human body in a set standard queue action posture, the rigid transformation corresponding to each rigid bone is calculated to obtain the position information of the terminal joints and skeleton joints in each posture; The linear regression model is expressed as: ； in is a constant, β i Represents the i-th dimension element of the morphological variable vector, a i is the constant parameter corresponding to the i-th dimension element of the morphological variable vector; The training method of the linear regression model includes: For a certain person’s skeleton joint point j, assuming that the position of the skeleton joint point j in posture A is p j , the bones associated with skeleton joint point j are b1 and b2; The position p of the skeleton joint point j in posture A is obtained by solving the following equation j : ； Where K is the number of postures in the training set, is the three-dimensional rotation matrix of bone b1 in posture k, is the three-dimensional translation vector of bone b1 in posture k, is the three-dimensional rotation matrix of bone b2 in posture k, is the three-dimensional translation vector of bone b2 in posture k; sym(j) represents the symmetrical joint of skeleton joint point j, represents the position of the symmetrical joint of the skeleton joint point j in posture A; sym(b1) and sym(b2) represent the symmetrical bones of bone b1 and bone b2 respectively, and M is the symmetry transformation matrix, which is expressed as: ； is the 3D rotation matrix of the bone symmetrical to bone b1 in posture k, The 3D rotation matrix of the bone symmetrical to bone b2 in pose k, is the three-dimensional translation vector of the bone symmetrical to bone b1 in posture k, is the three-dimensional translation vector of the bone symmetrical to bone b2 in posture k; It is to set parameters; By solving the above equations, we can obtain the morphological variables β and the position p of the skeleton joint point j under posture A. j , thereby constructing a linear regression model training set; The linear regression model is trained using the training set to determine the constant parameter corresponding to the i-th dimension element of the morphological variable vector, and the position p of the skeleton joint point j in the A posture is obtained. j and the linear relationship between the morphological variables β.

2. A digital evaluation method for individual soldier's bare-handed formation action according to claim 1, characterized in that: The simplified skeleton of a 3D human body is represented as , represents a set of joint points, Represents a set of bones connected by adjacent joints.

3. The digital evaluation method for individual soldier's bare-handed formation action according to claim 1 is characterized in that: Determine the position of the end joint of the person being measured in posture A, including: The center point of all adjacent points within a set distance from the end joint point is determined as the end joint point position.

4. The digital evaluation method for individual soldier's bare-handed formation action according to claim 1 is characterized in that: The position of the skeleton joint point j after rigid transformation is expressed as: ； in is the position of the skeleton joint point j after rigid transformation, For bone b1 in the pose variable The three-dimensional rotation matrix when For bone b1 in the pose variable The three-dimensional translation vector at time , For bone b2 in the pose variable The three-dimensional rotation matrix when For bone b2 in the pose variable The three-dimensional translation vector at time .

5. The digital evaluation method for individual soldier's bare-handed formation action according to claim 1 is characterized in that: The standard queue movement postures include standing at attention, marching with left hand in front, marching with right hand in front, marching in unison with left hand in front, marching in unison with right hand in front and saluting.

6. The digital evaluation method for individual soldier's bare-handed formation movement according to claim 1 is characterized in that: The end joints include the top of the head, toes and fingertips.

7. A digital assessment system for individual soldier's bare-handed formation movements, characterized by: A digital assessment method for performing the individual soldier's barehanded formation movement according to any one of claims 1 to 6, comprising: Image acquisition module, 3D human body modeling module, simplified skeleton representation module and digital evaluation module; The image acquisition module is used to collect three-dimensional point cloud data of the measured person in posture A and the set standard queue action posture; The three-dimensional human body modeling module is used to determine the morphological variables β and posture variables θ of the measured person according to the three-dimensional point cloud data in the A posture and the set standard queue action posture; Reconstruct the three-dimensional human body parameter grid model of the person being measured, wherein the three-dimensional human body parameter grid model expresses the human body morphological variables β and posture variables ,constant To the three-dimensional human body space The mapping relationship; The simplified skeleton representation module is used to calculate the position information of the end joints and skeleton joints of the measured person in various postures based on the three-dimensional human body parameterized mesh model to obtain a simplified skeleton representation; The digital evaluation module is used to obtain digital evaluation results by comparing the standard queue action quantification rules based on the simplified skeleton representation.

8. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images