Human bone landmark detection method, evaluation system, equipment and storage medium

Abstract

The present invention discloses a human body bone landmark detection method, evaluation system, equipment and storage medium, including: identifying the bone landmarks of a first user, obtaining a set of static bone landmarks of the user, determining a set of posture and body evaluation indicators, and obtaining a first evaluation picture and video information; determining a predetermined convolution kernel according to the posture and body evaluation indicator set, performing traversal convolution calculation on the first evaluation picture and video information according to the predetermined convolution kernel to obtain a result, obtaining first static posture feature information according to the calculation result, generating a posture skeleton simulation diagram for dynamic demonstration; performing multi-dimensional analysis on the first static posture feature information and the posture skeleton simulation diagram, generating a first posture accuracy coefficient as a result of the analysis, and generating a first reminder instruction when the accuracy coefficient is lower than the preset accuracy coefficient, reminding the first user to correct the posture, so that the posture analysis result is more accurate and efficient, and the posture correction is more timely and reasonable.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence computer image recognition calculation measurement technology, specifically to an intelligent human bone landmark detection method and an intelligent whole-body posture assessment system, electronic equipment, and a computer-readable storage medium. Background Art

[0002] Posture is the spatial position of the human body to maintain and ensure its functional state, and the external manifestation of maintaining appropriate relationships between body segments and between the body and the environment.

[0003] Posture is the positioning of the body's bones, joints, and soft tissues. Good posture is a state of balance between muscles and bones, and plays a vital role in maintaining physical health, preventing joint injuries, and static rehabilitation. Therefore, correct posture assessment is necessary, which is of great significance for monitoring the user's health assessment, correction, and rehabilitation.

[0004] Theoretically, correct posture means maintaining a scientific, balanced alignment of all body parts, in the correct order. The normal human spine has four curvatures, known as physiological curvatures: a slightly forward cervical curve, a slightly backward thoracic curve, a more pronounced forward lumbar curve, and a more pronounced backward sacral curve. These physiological curves not only mitigate shock and protect internal organs, but also help maintain the body's center of gravity.

[0005] From the side: The ears, shoulders, and hips are aligned in a straight line perpendicular to the ground. From the back: The head, back, and hips are aligned, shoulders retracted and naturally lowered, the chin slightly retracted, and the spine maintains its natural curvature. In this position, the spine is minimally stressed and in its healthiest state.

[0006] However, the existing technology has technical problems such as inaccurate posture assessment, slow assessment speed, and inconvenience in large-scale screening and assessment, which in turn affects the user's correction and rehabilitation effects. Therefore, the present invention provides an intelligent human bony landmark detection method and an intelligent whole-body posture assessment system to solve the problems raised in the above background technology. Summary of the Invention

[0007] In order to achieve the above object, the present invention adopts the following technical solutions:

[0008] An intelligent human body bony landmark detection method, the method comprising the following steps:

[0009] The intelligent image acquisition device identifies the bony landmarks of a first user to obtain a set of static bony landmarks of the user, and dynamically marks the set of static bony landmarks of the user; determines a set of posture and body shape assessment indicators based on the set of static bony landmarks of the user; obtains first assessment images and video information through the intelligent image acquisition device, wherein the first assessment images and video information include static images and video information of the first user; determines a predetermined convolution kernel based on the set of posture and body shape assessment indicators, and performs a traversal convolution calculation on each frame of image information of the first assessment images and video information according to the predetermined convolution kernel to obtain a first convolution calculation result; obtains first static posture feature information based on the first convolution calculation result, and generates a posture skeletal simulation diagram based on the first static posture feature information for dynamic demonstration; inputs the first static posture feature information and the posture skeletal simulation diagram into a first static posture analysis model for multi-dimensional analysis to obtain a first posture analysis result; generates a first posture accuracy coefficient based on the first posture analysis result; and generates a first reminder instruction if the first posture accuracy coefficient is lower than a preset accuracy coefficient, wherein the first reminder instruction is used to remind the first user to correct the posture.

[0010] Furthermore, the detection method also includes the following steps: based on the first posture analysis result, obtaining the non-standard evaluation indicators in the posture and body evaluation indicator set; performing correlation analysis based on the non-standard evaluation indicators and the posture and body evaluation indicator set to obtain an indicator correlation set; screening the indicators in the indicator correlation set that exceed a predetermined correlation threshold to obtain a correlation correction indicator; supplementing the first posture analysis result according to the correlation correction indicator to obtain a second posture analysis result.

[0011] Furthermore, the first static posture feature information and the posture skeletal simulation diagram are input into the first static posture analysis model for multi-dimensional analysis to obtain a first posture analysis result, including: obtaining an initial hidden layer value of the recurrent neural network, and obtaining a first input weight matrix based on the initial hidden layer value; using the historical static posture feature information and the historical posture skeletal simulation diagram as input layer information, and training the recurrent neural network according to the input layer information and the first input weight matrix; using the input layer information and the initial hidden layer value as the next hidden layer value, and using the historical posture analysis results as identification information for iterative training in sequence to construct the first static posture analysis model.

[0012] Furthermore, if the first posture accuracy coefficient is lower than a preset accuracy coefficient, a first reminder instruction is generated, and the first reminder instruction is used to remind the first user to correct the posture, including: obtaining a first posture correction coefficient based on the difference between the first posture accuracy coefficient and the preset accuracy coefficient; constructing a standard posture database, inputting the first posture correction coefficient and the standard posture database into a posture correction analysis model to obtain a first posture correction plan; and correcting the posture of the first user according to the first posture correction plan.

[0013] Furthermore, the first posture analysis result is supplemented according to the correlation correction index to obtain a second posture analysis result, including: obtaining static physical sign information of the first user through the intelligent image acquisition device; obtaining a first human body depth image and gradient data based on the static physical sign information; performing statistics on the first human body depth image and gradient data, and obtaining a first posture evaluation index based on the statistical results; and correcting the second posture analysis result according to the first posture evaluation index.

[0014] Furthermore, the construction of a standard posture database, inputting the first posture correction coefficient and the standard posture database into a posture correction analysis model, and obtaining a first posture correction scheme, includes: obtaining a standard posture image data set based on the standard posture database; determining a standard posture image change coefficient based on a static application scenario; performing data amplification on the standard posture image data set based on an image processing algorithm, the data amplification performing a change output according to the standard posture image change coefficient to obtain an amplified standard posture image data set; and performing data expansion on the standard posture database based on the amplified standard posture image data set.

[0015] An intelligent whole-body posture assessment system for implementing an intelligent human bone landmark detection method, the assessment system comprising: a first obtaining unit, the first obtaining unit being used to identify the bone landmarks of a first user through an intelligent image acquisition device, obtain a set of static bone landmarks of the user, and dynamically mark the set of static bone landmarks of the user; a first determining unit, the first determining unit being used to determine a set of posture and body evaluation indicators based on the set of static bone landmarks of the user; a second obtaining unit, the second obtaining unit being used to obtain first evaluation pictures and video information through an image acquisition device, the first evaluation pictures and video information including static pictures and video information of the first user; a third obtaining unit, the third obtaining unit being used to determine a predetermined convolution kernel based on the set of posture and body evaluation indicators, and to dynamically mark the set of static bone landmarks of the user; The first processing unit is used to obtain first static posture feature information according to the first convolution calculation result, and generate a posture skeleton simulation diagram for dynamic demonstration according to the first static posture feature information; the fourth obtaining unit is used to input the first static posture feature information and the posture skeleton simulation diagram into a first static posture analysis model for multi-dimensional analysis to obtain a first posture analysis result; the first generating unit is used to generate a first posture accuracy coefficient based on the first posture analysis result; the second processing unit is used to generate a first reminder instruction if the first posture accuracy coefficient is lower than a preset accuracy coefficient, and the first reminder instruction is used to remind the first user to correct the posture.

[0016] Furthermore, the intelligent whole-body posture assessment system also includes: a fifth acquisition unit, which is used to obtain the non-standard assessment indicators in the posture and body assessment indicator set based on the first posture analysis result; a sixth acquisition unit, which is used to perform correlation analysis based on the non-standard assessment indicators and the posture and body assessment indicator set to obtain an indicator correlation set; a seventh acquisition unit, which is used to screen the indicators in the indicator correlation set that exceed a predetermined correlation threshold to obtain a correlation correction indicator; and an eighth acquisition unit, which is used to supplement the first posture analysis result according to the correlation correction indicator to obtain a second posture analysis result.a ninth obtaining unit, the ninth obtaining unit being used to obtain the static vital sign information of the first user through the intelligent image collector device; a tenth obtaining unit, the tenth obtaining unit being used to obtain a first human body depth image and gradient data according to the static vital sign information; an eleventh obtaining unit, the eleventh obtaining unit being used to perform statistics on the values of the first human body depth image and gradient data, and obtain a first posture evaluation index according to the statistical result; a first correction unit, the first correction unit being used to correct the second posture analysis result according to the first posture evaluation index; a twelfth obtaining unit, the twelfth obtaining unit being used to correct the second posture analysis result according to the first posture accuracy coefficient and the preset accuracy coefficient a first posture correction coefficient is obtained by calculating the difference between the two numbers; a third processing unit is used to construct a standard posture database, input the first posture correction coefficient and the standard posture database into a posture correction analysis model, and obtain a first posture correction scheme; a fourth processing unit is used to correct the posture of the first user according to the first posture correction scheme; a thirteenth obtaining unit is used to obtain a standard posture image data set according to the standard posture database; a second determining unit is used to determine the standard posture image change coefficient according to a static application scenario; a fourteenth obtaining unit is used to obtain a standard posture image data set based on image processing. A processing algorithm performs data amplification on the standard posture image data set, and the data amplification is output according to the change coefficient of the standard posture image to obtain an amplified standard posture image data set; a first expansion unit, the first expansion unit is used to perform data expansion on the standard posture database according to the amplified standard posture image data set; a fifteenth obtaining unit, the fifteenth obtaining unit is used to obtain the basic physiological information of the first user; a first selection unit, the first selection unit is used to select a posture evaluation mode according to the personal training goal of the first user; a first calling unit, the first calling unit is used to select a posture evaluation mode from a static posture analysis mode based on the basic physiological information and the posture evaluation mode. The first static posture analysis model is called in the model library; a sixteenth obtaining unit, the sixteenth obtaining unit is used to obtain the initial hidden layer value of the recurrent neural network, and obtain the first input weight matrix based on the initial hidden layer value; a first correction unit, the first correction unit is used to use the historical static posture feature information and the historical posture skeleton simulation diagram as input layer information, and train the recurrent neural network according to the input layer information and the first input weight matrix; a first construction unit, the first construction unit is used to use the input layer information and the initial hidden layer value as the next hidden layer value, and use the historical posture analysis result as identification information for iterative training in sequence to construct the first static posture analysis model.

[0017] An electronic device comprises: an image collector, a memory, and a processor. The image collector collects human body images to obtain human body bony landmarks. The memory stores a computer program. When the processor executes the computer program, the steps of the above-mentioned detection method are implemented.

[0018] A computer-readable storage medium stores a computer program, which implements the steps of the above detection method when executed by a processor.

[0019] Compared with the prior art, the present invention has the following beneficial effects:

[0020] The invention adopts a technical solution that uses an intelligent image acquisition device to identify the bony landmarks of a first user and dynamically mark the set of identified static bony landmarks of the user, collects static images and video information of the first user through the intelligent image acquisition device, determines a predetermined convolution kernel based on a set of posture and body evaluation indicators, performs traversal convolution calculations on the first evaluation images and video information according to the predetermined convolution kernel, and determines first static posture feature information based on the calculated first convolution calculation results, thereby generating a posture skeletal simulation diagram for dynamic demonstration. The first static posture feature information and the posture skeletal simulation diagram are then input into a first static posture analysis model for multi-dimensional analysis to obtain a first posture analysis result. Based on the first posture analysis result, a first posture accuracy coefficient is generated. If the first posture accuracy coefficient is lower than the preset accuracy coefficient, the first user is reminded to correct the posture. Thus, the technical effect of real-time monitoring of the user's static posture and multi-dimensional analysis of the monitoring results is achieved, thereby making the posture analysis results more accurate and efficient, achieving more timely and reasonable posture correction, and ensuring the user's correction and rehabilitation effects. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 This is a flowchart of an intelligent human bony landmark detection method and posture assessment system for this application;

[0022] Figure 2 This is a flow chart of supplementing the first posture analysis result in an intelligent human bony landmark detection method and posture assessment system of the present application;

[0023] Figure 3 This is a schematic diagram of the process of correcting the second posture analysis result in an intelligent human bone landmark detection method and posture assessment system of the present application;

[0024] Figure 4 This is a schematic diagram of a process for correcting the posture of a first user in an intelligent human bony landmark detection method and posture assessment system according to the present application;

[0025] Figure 5 A schematic diagram of the structure of an intelligent whole-body posture assessment system for this application;

[0026] Figure 6 This is a structural diagram of an electronic device in this application.

[0027] Among them: bus 1110 , processor 1120 , transceiver 1130 , bus interface 1140 , memory 1150 , operating system 1151 , application 1152 , user interface 1160 , image collector 1170 . DETAILED DESCRIPTION

[0028] The technical solutions in the embodiments of the present invention will be described clearly and completely below with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments.

[0029] In the description of this application, those skilled in the art should know that this application can be implemented as a method, an apparatus, an electronic device, and a computer-readable storage medium. Therefore, this application can be specifically implemented in the following forms: complete hardware, complete software (including firmware, resident software, microcode, etc.), or a combination of hardware and software. In addition, in some embodiments, this application can also be implemented in the form of a computer program product in one or more computer-readable storage media, wherein the computer-readable storage medium contains computer program code.

[0030] The computer-readable storage medium may be any combination of one or more computer-readable storage media. Computer-readable storage media include: electrical, magnetic, optical, electromagnetic, infrared or semiconductor systems, devices or components, or any combination thereof. More specific examples of computer-readable storage media include: portable computer disks, hard disks, random access memories, read-only memories, erasable programmable read-only memories, flash memory, optical fibers, optical disc read-only memories, optical storage devices, magnetic storage devices, or any combination thereof. In this application, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device, or component.

[0031] The acquisition, storage, use, and processing of data in this application's technical solution comply with relevant national laws.

[0032] This application describes the provided methods, devices, and electronic devices through flowcharts and / or block diagrams.

[0033] It should be understood that each block in the flowchart and / or block diagram, as well as combinations of blocks in the flowchart and / or block diagram, can be implemented by computer-readable program instructions. These computer-readable program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, thereby producing a machine. These computer-readable program instructions are executed by the computer or other programmable data processing device to produce a device that implements the functions / operations specified in the blocks in the flowchart and / or block diagram.

[0034] These computer-readable program instructions may also be stored in a computer-readable storage medium that enables a computer or other programmable data processing device to operate in a specific manner. In this way, the instructions stored in the computer-readable storage medium produce an instruction device product that implements the functions / operations specified in the blocks in the flowchart and / or block diagram.

[0035] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device, so that a series of operational steps are performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby enabling the instructions executed on the computer or other programmable data processing apparatus to provide a process that implements the functions / operations specified by the blocks in the flowchart and / or block diagram.

[0036] The present application is described below in conjunction with the accompanying drawings.

[0037] Example 1

[0038] like Figure 1 As shown, the present application provides a software method and system for intelligently identifying bony landmarks of the human body. The method is applied to a posture assessment system. The system includes an intelligent image acquisition device, and the system is communicatively connected to an intelligent image acquisition device. The method includes:

[0039] Step S100: identifying the bony landmarks of the first user by the intelligent image collector device, obtaining a set of static bony landmarks of the user, and dynamically marking the set of static bony landmarks of the user;

[0040] Specifically, posture is the positioning of the body's bones, joints, and soft tissues. Good posture is a state of muscle and bone balance that plays a vital role in maintaining good health, preventing joint injuries, and performing static rehabilitation. Therefore, a correct posture assessment is necessary. Posture assessment is the starting point for coaches or therapists to initiate static correction strategies. It can also monitor the patient's progress and is of great significance for monitoring the user's health assessment, correction, and rehabilitation. The intelligent image acquisition device is a general term for devices that use image acquisition technology to intelligently design and develop image acquisition devices for daily use, such as watches, bracelets, glasses, and clothing.

[0041] The intelligent image acquisition device identifies the bony landmarks of a first user, who is a patient requiring posture assessment, and obtains a set of identified static bony landmarks of the user. The set is the bony landmarks assessed during a static posture test, including the acromion, anterior superior iliac spine, patellar center, tibial tuberosity, ankle center, posterior surface of the right temporomandibular joint, shoulder center, iliac crest tuberosity, greater trochanter of the femur, knee center, anterior lateral malleolus, inferior angle of scapula, horizontal angle, medial margin of scapula T5-T8, posterior superior iliac spine, C7-L5 (upper cervical and lumbar spinous processes), and joints such as the shoulder, elbow, hip, knee, and ankle. The set of static bony landmarks is dynamically marked, providing a basis for subsequently comprehensively understanding the real-time posture changes of the assessment tester.

[0042] Step S200: determining a posture and body shape assessment index set based on the user's static bony landmark set;

[0043] 1. Specifically, based on the user's static bony landmark set, a posture and body shape assessment index set is determined. The posture and body shape assessment index set is a discriminant index for accurately assessing the user's posture and body shape, including head and eye tilt, level, and head position; shoulder level; trunk tilt, horizontal position and anteroposterior tilt of the pelvis, horizontal position of both knee joints, hyperextension and hyperflexion of the knee joints, ankle level; scoliosis, Q angle, and foot inversion and varus, etc.

[0044] Step S300: obtaining first evaluation pictures and video information through the intelligent image acquisition device, wherein the first evaluation pictures and video information include static pictures and video information of the first user;

[0045] Step S400: determining a predetermined convolution kernel according to the posture and body evaluation index set, and performing a traversal convolution calculation on each frame of the first evaluation picture and video information according to the predetermined convolution kernel to obtain a first convolution calculation result;

[0046] Specifically, the intelligent image acquisition device monitors and captures static images and video information of the first user. The intelligent image acquisition device can preferably be an infrared camera, providing the monitoring image and video data foundation for the user's static posture assessment. A predetermined convolution kernel is determined based on the posture and body posture assessment indicator set, and convolution kernel feature acquisition is performed according to the posture and body posture assessment indicator set. The predetermined convolution kernel is the body posture feature to be tested and assessed. A convolution kernel is a weighted average of pixels in a small area of a given input image during image processing, where the weights are defined by a function called a convolution kernel.

[0047] The convolution kernel focuses on local features, i.e., predefined standard features, and collects and evaluates the matching degree of the features based on the numerical value of the convolution kernel at the local feature location. A convolution calculation is performed on each frame of the first evaluation image and video information using the predefined convolution kernel to obtain a first convolution calculation result. The first convolution calculation result indicates the characteristics of posture changes during the user test evaluation process.

[0048] Step S500: obtaining first static posture feature information according to the first convolution calculation result, and generating a posture skeleton simulation diagram for dynamic demonstration according to the first static posture feature information;

[0049] Specifically, based on the first convolution calculation result, first static posture feature information of the first user is determined. This first static posture feature information represents the user's posture variation characteristics during a static test assessment, including static amplitude, angle, and other variation information. Based on this first static posture feature information, a posture skeletal simulation diagram is generated for dynamic demonstration. This animated simulation diagram can intuitively and accurately display the static state of each joint of the user.

[0050] Step S600: inputting the first static posture feature information and the posture skeleton simulation diagram into a first static posture analysis model for multi-dimensional analysis to obtain a first posture analysis result;

[0051] Furthermore, step S600 of the present application also includes:

[0052] Step S610: obtaining an initial hidden layer value of the recurrent neural network, and obtaining a first input weight matrix based on the initial hidden layer value;

[0053] Step S620: using the historical static posture feature information and the historical posture skeleton simulation diagram as input layer information, and training the recurrent neural network according to the input layer information and the first input weight matrix;

[0054] Step S630: Using the input layer information and the initial hidden layer value as the next hidden layer value, and using the historical posture analysis results as identification information, iteratively train in sequence to construct the first static posture analysis model.

[0055] Specifically, in order to evaluate whether the user's posture is standard, the first static posture feature information and the posture skeleton simulation diagram are input into the first static posture analysis model for multi-dimensional analysis, and the first static posture analysis model is a recurrent neural network. A recurrent neural network is a type of recurrent neural network that takes sequence data as input, recursively in the direction of sequence evolution, and all nodes (recurrent units) are connected in a chain, including an input layer, a hidden layer, and an output layer. In the process of processing input information in the processing layer of the recurrent neural network, in addition to processing according to the current input information, it also saves the output information of the previous time series, and processes the output information as the input information of the current time series, thereby obtaining the output. As the time series progresses, the processing layer is continuously updated. By using neurons with self-feedback, the recurrent neural network makes the output of the network not only related to the current input, but also related to the output of the previous moment, so when processing time series data of any length, it has short-term memory capabilities.

[0056] The initial hidden layer value can be obtained in a customized manner, and the first input weight matrix is obtained based on the initial hidden layer value. During the processing, the current input information and the output information of the previous time series are predicted according to a certain weight ratio, which is the above-mentioned weight matrix. In addition, during the update of the processing layer, the weight values in the weight matrix are stable. In order to ensure the accuracy of the model evaluation, multiple sets of historical static posture feature information and historical posture skeletal simulation images are used as input layer information. The recurrent neural network is trained based on the input layer information and the first input weight matrix. The input layer of each time and the hidden layer of the previous time are used as the hidden layer of each time. The hidden layer of each time here is the hidden layer value of the next time, and the corresponding posture analysis result is used as the identification information, i.e., the output result.

[0057] Through iterative training, when the output of the recurrent neural network reaches a certain accuracy or converges, supervised training is completed and the first static posture analysis model is constructed. Based on the first static posture analysis model, the model's output, namely the first posture analysis result, is obtained. The first posture analysis result includes information on the accuracy of the first user's posture assessment results, the standard qualification of each posture and body posture assessment indicator, and detailed analysis results. By using the recurrent neural network model to analyze user posture, the posture analysis results are more accurate and reasonable, thereby improving the analysis efficiency and accuracy of the posture assessment results.

[0058] Step S700: generating a first posture accuracy coefficient based on the first posture analysis result;

[0059] Step S800: If the first posture accuracy coefficient is lower than a preset accuracy coefficient, a first reminder instruction is generated, where the first reminder instruction is used to remind the first user to correct the posture.

[0060] Specifically, based on the first posture analysis result, a first posture accuracy coefficient is generated. The first posture accuracy coefficient represents the standard degree of the user's static posture. A larger posture accuracy coefficient indicates a more standard static posture of the user. If the first posture accuracy coefficient is lower than a preset accuracy coefficient, it indicates that the user's static posture is not standard enough and that there is a health problem that requires correction. The first user is reminded to correct their posture according to the first reminder instruction, thereby achieving more timely and reasonable posture correction, thereby ensuring the user's correction and rehabilitation effect.

[0061] like Figure 4 As shown, further, step S800 of this application also includes:

[0062] Step S810: obtaining a first posture correction coefficient according to the difference between the first posture accuracy coefficient and the preset accuracy coefficient;

[0063] Step S820: constructing a standard posture database, inputting the first posture correction coefficient and the standard posture database into a posture correction analysis model, and obtaining a first posture correction solution;

[0064] Step S830: Correcting the posture of the first user according to the first posture correction solution.

[0065] Specifically, the first posture correction coefficient is the degree to which the user's static posture needs to be corrected, and is the difference between the first posture accuracy coefficient and the preset accuracy coefficient. The larger the posture correction coefficient, the less standard the user's static posture is, and the greater the degree of correction required. A standard posture database is constructed. The standard posture database is a database of standard static posture ranges for users with various body shape characteristics. The first posture correction coefficient and the standard posture database are input into a posture correction analysis model. The posture correction analysis model is a neural network model used to analyze and formulate relevant correction plans based on the user's static posture. The training output of the model is obtained, namely the first posture correction plan. The first posture correction plan is a personalized correction and rehabilitation plan for the user, including correction frequency, correction strength, correction method, etc. The posture of the first user is corrected according to the first posture correction plan, and the correction result is more accurate and reasonable, thereby ensuring the user's correction and rehabilitation effects.

[0066] like Figure 2As shown, further, the application steps also include:

[0067] Step S910: obtaining, based on the first posture analysis result, non-compliant evaluation indicators in the posture and body evaluation indicator set;

[0068] Step S920: performing a correlation analysis based on the non-standard evaluation indicator and the posture and body evaluation indicator set to obtain an indicator correlation degree set;

[0069] Step S930: screening the indicators in the indicator correlation set that exceed a predetermined correlation threshold to obtain correlation correction indicators;

[0070] Step S940: Supplement the first posture analysis result according to the correlation correction index to obtain a second posture analysis result.

[0071] Specifically, based on the first posture analysis result, a substandard evaluation indicator in the posture and body posture evaluation indicator set is obtained, such as excessive deviation of the center of gravity during walking, lateral bending of the body, or a small static knee amplitude. A correlation analysis is performed on the substandard evaluation indicator and other indicators in the posture and body posture evaluation indicator set. That is, a corresponding correlation indicator analysis is performed on the substandard indicator to obtain an indicator correlation degree set. The indicator correlation degree set is the correlation degree between the substandard indicator and other indicators in the posture and body posture evaluation indicator set.

[0072] Indicators in the indicator correlation set that exceed a predetermined correlation threshold are screened, where the predetermined correlation threshold is a preset indicator correlation range, to obtain a correlation correction indicator that has a strong correlation with the non-compliant indicator. The first posture analysis result is supplemented based on the correlation correction indicator to obtain a second posture analysis result after the supplementation. For example, if the hip joint static amplitude indicator and the knee joint static amplitude indicator have a strong correlation, the knee joint static amplitude indicator is combined with the hip joint static amplitude indicator when analyzing the reasons for the non-compliant hip joint static amplitude indicator. This combined analysis makes the posture analysis result more comprehensive and accurate, thereby improving the user's posture correction and health rehabilitation effects.

[0073] like Figure 3 As shown, further, step S940 of this application also includes:

[0074] Step S941: Obtaining static vital sign information of the first user through the intelligent image collector device;

[0075] Step S942: acquiring a first human body depth image and gradient data according to the static vital sign information;

[0076] Step S943: performing statistics on the values of the first depth image and the gradient data, and obtaining a first posture evaluation index according to the statistical results;

[0077] Step S944: Correct the second posture analysis result according to the first posture evaluation index.

[0078] Specifically, the intelligent image acquisition device obtains static vital sign information of the first user, including the coronal, sagittal, and horizontal planes during the user's static test. Based on the static vital sign information, a human body depth image and gradient data are obtained, and the data are characterized by the static vital sign information. Statistics are collected on the first human body depth image and gradient data, which indicate changes in the evaluator's posture. The larger the number and value of the gradients, the greater the postural imbalance of the evaluator. Based on the statistical results, a first posture assessment imbalance index is obtained, which is used to indicate the posture imbalance result of the posture evaluator.

[0079] The depth image and gradient data index of the user being assessed for posture can affect changes in physical signs. For example, if the user's depth image and gradient data are low, or if they are too tense or tense during the test, this can lead to deviations in the posture assessment results. The second posture analysis results are corrected based on the first posture assessment imbalance index. By analyzing the static posture of the user using the depth image and gradient data, the posture analysis results are more comprehensive and accurate, thereby improving the user's posture correction and health recovery outcomes.

[0080] Furthermore, after constructing the standard posture database, step S820 of the present application further includes:

[0081] Step S821: obtaining a standard posture image dataset according to the standard posture database;

[0082] Step S822: determining a standard posture image variation coefficient according to a static application scenario;

[0083] Step S823: performing data augmentation on the standard posture image dataset based on an image processing algorithm, wherein the data augmentation is changed and output according to the standard posture image change coefficient to obtain an augmented standard posture image dataset;

[0084] Step S824: performing data expansion on the standard posture database according to the expanded standard posture image dataset.

[0085] Specifically, the standard posture database is a database of standard static posture ranges for users of various body types, including data and images. Based on this standard posture database, a standard posture image dataset is obtained. This standard posture image dataset is a collection of images of users in standard postures. The static application scenarios are static scenarios for test users, including walking, jogging, and postoperative rehabilitation. The standard posture image dataset requires conversion of static angles, static amplitudes, and static speeds for different static scenarios.

[0086] Image conversion is performed according to the standard posture image variation coefficients, such as scaling, amplitude conversion, and static angle conversion. The standard posture image dataset is then augmented using an image processing algorithm. The data augmentation outputs the changes according to the standard posture image variation coefficients to obtain an augmented standard posture image dataset. Based on the augmented standard posture image dataset, the standard posture database is augmented. By augmenting the standard posture database, image augmentation for different static application scenarios is improved, thereby enhancing the comprehensiveness of the standard posture database and the accuracy of posture analysis results.

[0087] Furthermore, the application steps also include:

[0088] Step S1010: Obtaining basic physiological information of the first user;

[0089] Step S1020: selecting a posture assessment mode according to the first user's personal training goal;

[0090] Step S1030: Based on the basic physiological information and the posture evaluation mode, call the first static posture analysis model from a static posture analysis model library.

[0091] Specifically, due to different physiological characteristics of users, their static posture assessments also vary accordingly. The first user's basic physiological information includes the user's gender, age, height, weight, and pathological condition. The first user's personal training goal is the user's static correction goal, such as posture correction, postoperative rehabilitation, postpartum rehabilitation, elderly rehabilitation, and adolescent scoliosis treatment. Different training goals require different posture testing standards. Based on the first user's personal training goal, a posture assessment mode is selected, such as the front, side, back, pelvis, scoliosis, standing position, and sitting position assessment modes.

[0092] The static posture analysis model library is a database of static posture analysis models trained and classified according to historical user static posture analysis results, assessment patterns, and personal physiological information. The posture analysis models in the static posture analysis model library are recurrent neural network models. Based on the basic physiological information and the posture assessment pattern, the first static posture analysis model is retrieved from the static posture analysis model library. The first static posture analysis model is a posture assessment model that matches the first user, achieving personalized matching of user posture assessment models, making the static posture analysis results more accurate, thereby improving the correction and rehabilitation effects on the user's posture.

[0093] In summary, the intelligent human bony landmark detection method and posture assessment system provided by this application have the following technical effects:

[0094] The invention adopts a technical solution that uses an intelligent image acquisition device to identify the bony landmarks of a first user and dynamically mark the set of identified static bony landmarks of the user, collects static images and video information of the first user through the intelligent image acquisition device, determines a predetermined convolution kernel based on a set of posture and body evaluation indicators, performs traversal convolution calculations on the first evaluation images and video information according to the predetermined convolution kernel, and determines first static posture feature information based on the calculated first convolution calculation results, thereby generating a posture skeletal simulation diagram for dynamic demonstration. The first static posture feature information and the posture skeletal simulation diagram are then input into a first static posture analysis model for multi-dimensional analysis to obtain a first posture analysis result. Based on the first posture analysis result, a first posture accuracy coefficient is generated. If the first posture accuracy coefficient is lower than the preset accuracy coefficient, the first user is reminded to correct the posture. Thus, the technical effect of real-time monitoring of the user's static posture and multi-dimensional analysis of the monitoring results is achieved, thereby making the posture analysis results more accurate and efficient, achieving more timely and reasonable posture correction, and ensuring the user's correction and rehabilitation effects.

[0095] Example 2

[0096] Based on the above embodiments, the present invention also provides an intelligent whole body posture assessment system, such as Figure 5 As shown, the system includes:

[0097] A first obtaining unit 11 is configured to identify the bony landmarks of the first user through an intelligent image acquisition device, obtain a set of static bony landmarks of the user, and dynamically mark the set of static bony landmarks of the user;

[0098] A first determining unit 12, configured to determine a posture and body shape assessment index set based on the user's static bony landmark set;

[0099] A second obtaining unit 13, the second obtaining unit 13 is configured to obtain first evaluation pictures and video information through an intelligent image acquisition device, wherein the first evaluation pictures and video information include static pictures and video information of the first user;

[0100] a third obtaining unit 14, configured to determine a predetermined convolution kernel according to the posture and body evaluation indicator set, and perform a traversal convolution calculation on each frame of the first evaluation image and video information according to the predetermined convolution kernel to obtain a first convolution calculation result;

[0101] a first processing unit 15 configured to obtain first static posture feature information based on the first convolution calculation result, and generate a posture skeleton simulation diagram for dynamic demonstration based on the first static posture feature information;

[0102] a fourth obtaining unit 16, configured to input the first static posture feature information and the posture skeleton simulation diagram into a first static posture analysis model for multi-dimensional analysis to obtain a first posture analysis result;

[0103] a first generating unit 17, configured to generate a first posture accuracy coefficient based on the first posture analysis result;

[0104] The second processing unit 18 is configured to generate a first reminder instruction if the first posture accuracy coefficient is lower than a preset accuracy coefficient, wherein the first reminder instruction is configured to remind the first user to correct the posture.

[0105] Furthermore, the system further comprises:

[0106] a fifth obtaining unit, configured to obtain, based on the first posture analysis result, an unqualified evaluation indicator in the posture and body evaluation indicator set;

[0107] a sixth obtaining unit, configured to perform a correlation analysis based on the non-standard evaluation indicator and the posture and body shape evaluation indicator set to obtain an indicator correlation degree set;

[0108] a seventh obtaining unit, configured to screen indicators exceeding a predetermined correlation threshold in the indicator correlation set to obtain correlation correction indicators;

[0109] An eighth obtaining unit is configured to supplement the first posture analysis result according to the correlation correction index to obtain a second posture analysis result.

[0110] Furthermore, the system further comprises:

[0111] a ninth obtaining unit, configured to obtain static vital sign information of the first user through the intelligent image collector device;

[0112] a tenth obtaining unit, configured to obtain a first human body depth image and gradient data according to the static vital sign information;

[0113] an eleventh obtaining unit, configured to perform statistics on the values of the first human body depth image and the gradient data, and obtain a first posture evaluation index according to the statistical results;

[0114] A first correction unit corrects the second posture analysis result according to the first posture evaluation index.

[0115] Furthermore, the system further comprises:

[0116] a twelfth obtaining unit, configured to obtain a first posture correction coefficient according to a difference between the first posture accuracy coefficient and the preset accuracy coefficient;

[0117] a third processing unit, configured to construct a standard posture database, input the first posture correction coefficient and the standard posture database into a posture correction analysis model, and obtain a first posture correction solution;

[0118] A fourth processing unit is configured to perform posture correction on the first user according to the first posture correction solution.

[0119] Furthermore, the system further comprises:

[0120] a thirteenth obtaining unit, configured to obtain a standard posture image dataset based on the standard posture database;

[0121] a second determining unit, configured to determine a standard posture image variation coefficient according to a static application scenario;

[0122] a fourteenth obtaining unit, configured to perform data augmentation on the standard posture image dataset based on an image processing algorithm, wherein the data augmentation is performed by changing and outputting the data according to a change coefficient of the standard posture image to obtain an augmented standard posture image dataset;

[0123] A first expansion unit is configured to expand the standard posture database according to the expanded standard posture image data set.

[0124] Furthermore, the system further comprises:

[0125] a fifteenth obtaining unit, configured to obtain basic physiological information of the first user;

[0126] a first selection unit, configured to select a posture assessment mode according to a personal training goal of the first user;

[0127] A first calling unit is configured to call the first static posture analysis model from a static posture analysis model library based on the basic physiological information and the posture evaluation mode.

[0128] Furthermore, the system further comprises:

[0129] a sixteenth obtaining unit, configured to obtain an initial hidden layer value of the recurrent neural network, and obtain a first input weight matrix based on the initial hidden layer value;

[0130] a first correction unit, configured to use the historical static posture feature information and the historical posture skeletal simulation graph as input layer information, and train the recurrent neural network according to the input layer information and the first input weight matrix;

[0131] The first construction unit is used to use the input layer information and the initial hidden layer value as the next hidden layer value, and use the historical posture analysis results as identification information for iterative training in sequence to construct the first static posture analysis model.

[0132] The foregoing Figure 1 The various variations and specific examples of the intelligent method for identifying human bony landmarks in Example 1 are also applicable to an intelligent whole-body posture assessment system in this embodiment. Through the above detailed description of an intelligent method for identifying human bony landmarks, those skilled in the art can clearly understand the implementation method of an intelligent whole-body posture assessment system in this embodiment, so for the sake of brevity of the specification, it will not be described in detail here.

[0133] In addition, the present application also provides an electronic device, including an image acquisition device, a bus, a transceiver, a memory, a processor, and a computer program stored in the memory and runnable on the processor. The transceiver, the memory, and the processor are respectively connected via a bus. When the computer program is executed by the processor, each process of the above-mentioned method embodiment for controlling output data is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0134] like Figure 6As shown, the present application also provides an electronic device, which includes a bus 1110 , a processor 1120 , a transceiver 1130 , a bus interface 1140 , a memory 1150 , a user interface 1160 and an image collector 1170 .

[0135] In the present application, the electronic device also includes: after the image collector 1170 collects images, videos, bony landmarks and other data, the computer program is stored on the memory 1150 and can be run on the processor 1120. When the computer program is executed by the processor 1120, the various processes of the above-mentioned method embodiment of controlling output data are implemented.

[0136] The transceiver 1130 is configured to receive and send data under the control of the processor 1120 .

[0137] In the present application, a bus architecture (represented by bus 1110 ) may include any number of interconnected buses and bridges, and bus 1110 connects various circuits including one or more processors represented by processor 1120 and memory represented by memory 1150 .

[0138] Bus 1110 represents one or more of any of several types of bus structures, including a memory bus and memory controller, a peripheral bus, an accelerated graphics port, a processor, or a local bus using any of a variety of bus architectures. By way of example and not limitation, such architectures include: Industry Standard Architecture bus, Micro Channel Architecture bus, expansion bus, graphics, Video Electronics Standards Association, and Peripheral Component Interconnect bus.

[0139] The processor 1120 can be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method embodiment can be completed by an integrated logic circuit of the hardware in the processor or instructions in the form of software. The above-mentioned processors include: general-purpose processors, central processing units, network processors, digital signal processors, application-specific integrated circuits, field programmable gate arrays, complex programmable logic devices, programmable logic arrays, microcontrollers or other programmable logic devices, discrete gates, transistor logic devices, discrete hardware components. The various methods, steps and logic block diagrams disclosed in this application can be implemented or executed. For example, the processor can be a single-core processor or a multi-core processor, and the processor can be integrated into a single chip or located on multiple different chips.

[0140] Processor 1120 can be a microprocessor or any conventional processor. The method steps disclosed herein can be performed directly by a hardware decoding processor, or by a combination of hardware and software modules within the decoding processor. The software modules can be located in a readable storage medium known in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, erasable programmable read-only memory, or registers. The readable storage medium is located in a memory, and the processor reads the information in the memory and, in conjunction with its hardware, performs the steps of the method described above.

[0141] The bus 1110 may also connect various other circuits, such as peripheral devices, voltage regulators, or power management circuits. The bus interface 1140 provides an interface between the bus 1110 and the transceiver 1130. These are all well known in the art and are therefore not further described in this application.

[0142] The transceiver 1130 can be a single component or multiple components, such as multiple receivers and transmitters, providing a means for communicating with various other devices over a transmission medium. For example, the transceiver 1130 receives external data from other devices and transmits data processed by the processor 1120 to other devices. Depending on the nature of the computer device, a user interface 1160 may also be provided, such as a touch screen, physical keyboard, display, mouse, speaker, microphone, trackball, joystick, or stylus. An image acquisition unit 1170 is provided for capturing images and videos of human bony landmark data.

[0143] It should be understood that in the present application, the memory 1150 may further include a memory remotely located relative to the processor 1120, and these remotely located memories may be connected to the server via a network. One or more portions of the aforementioned network may be an ad hoc network, an intranet, an extranet, a virtual private network, a local area network, a wireless local area network, a wide area network, a wireless wide area network, a metropolitan area network, the Internet, a public switched telephone network, a plain old telephone service network, a cellular telephone network, a wireless network, a wireless fidelity network, or a combination of two or more of the aforementioned networks. For example, the cellular telephone network and the wireless network may be a global mobile communications device, a code division multiple access device, a global interoperability for microwave access device, a general packet radio service device, a wideband code division multiple access device, a long term evolution device, an LTE frequency division duplex device, an LTE time division duplex device, an advanced long term evolution device, a universal mobile communications device, an enhanced mobile broadband device, a massive machine type communication device, an ultra-reliable low latency communication device, etc.

[0144] It should be understood that the memory 1150 in the present application may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory includes: read-only memory, programmable read-only memory, erasable programmable read-only memory, electrically erasable programmable read-only memory, or flash memory.

[0145] Volatile memory includes random access memory (RAM), which serves as an external cache memory. By way of example and not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate SDRAM, enhanced SDRAM, synchronous linked dynamic random access memory (SRAM), and direct memory bus (DMA) random access memory (DMA). Memory 1150 of the electronic device described herein includes, but is not limited to, the aforementioned and any other suitable types of memory.

[0146] In the present application, the memory 1150 stores the following elements of the operating system 1151 and the application 1152 : executable modules, data structures, or subsets thereof, or extended sets thereof.

[0147] Specifically, operating system 1151 includes various device programs, such as a framework layer, a core library layer, and a driver layer, for implementing various basic services and handling hardware-based tasks. Application programs 1152 include various application programs, such as media players and browsers, for implementing various application services. Programs implementing the methods of the present application may be included in application programs 1152. Application programs 1152 include applets, objects, components, logic, data structures, and other computer-executable instructions that perform specific tasks or implement specific abstract data types.

[0148] In addition, the present application also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the various processes of the above-mentioned method embodiment for controlling output data are implemented and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0149] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. An intelligent human body bony landmark detection method, characterized in that: The method includes the following steps: identifying the bony landmarks of a first user through an intelligent image acquisition device to obtain a set of static bony landmarks of the user, and dynamically marking the set of static bony landmarks of the user; determining a set of posture and body shape evaluation indicators based on the set of static bony landmarks of the user; obtaining a first evaluation picture and video information through an intelligent image acquisition device, wherein the first evaluation picture and video information include a static picture and video information of the first user; determining a predetermined convolution kernel based on the set of posture and body shape evaluation indicators, performing a traversal convolution calculation on each frame of image information of the first evaluation picture and video information according to the predetermined convolution kernel, and obtaining a first convolution calculation result; obtaining first static posture feature information based on the first convolution calculation result, and generating a posture skeleton simulation diagram based on the first static posture feature information for dynamic demonstration; inputting the first static posture feature information and the posture skeleton simulation diagram into a first static posture analysis model for multi-dimensional analysis to obtain a first posture analysis result; generating a first posture accuracy coefficient based on the first posture analysis result; and generating a first reminder instruction if the first posture accuracy coefficient is lower than a preset accuracy coefficient, wherein the first reminder instruction is used to remind the first user to correct the posture. The method further comprises the following steps: obtaining, based on the first posture analysis result, an evaluation indicator that does not meet the standard in the posture and body evaluation indicator set; performing a correlation analysis based on the evaluation indicator that does not meet the standard and the posture and body evaluation indicator set to obtain an indicator correlation set; screening indicators in the indicator correlation set that exceed a predetermined correlation threshold to obtain a correlation correction indicator; and supplementing the first posture analysis result based on the correlation correction indicator to obtain a second posture analysis result; The first static posture feature information and the posture skeletal simulation diagram are input into a first static posture analysis model for multi-dimensional analysis to obtain a first posture analysis result, including: obtaining an initial hidden layer value of a recurrent neural network, and obtaining a first input weight matrix based on the initial hidden layer value; using the historical static posture feature information and the historical posture skeletal simulation diagram as input layer information, and training the recurrent neural network according to the input layer information and the first input weight matrix; using the input layer information and the initial hidden layer value as the next hidden layer value, and using the historical posture analysis result as identification information for iterative training in sequence to construct the first static posture analysis model; The method of supplementing the first posture analysis result according to the correlation correction index to obtain a second posture analysis result includes: obtaining static physical sign information of the first user through the intelligent image acquisition device; obtaining a first human body depth image and gradient data based on the static physical sign information; performing statistics on the values of the first human body depth image and gradient data, and obtaining a first posture evaluation index based on the statistical results; and correcting the second posture analysis result according to the first posture evaluation index.

2. The method for detecting human bony landmarks according to claim 1, wherein: If the first posture accuracy coefficient is lower than the preset accuracy coefficient, a first reminder instruction is generated, and the first reminder instruction is used to remind the first user to correct the posture, including: obtaining a first posture correction coefficient based on the difference between the first posture accuracy coefficient and the preset accuracy coefficient; constructing a standard posture database, inputting the first posture correction coefficient and the standard posture database into a posture correction analysis model to obtain a first posture correction plan; and correcting the posture of the first user according to the first posture correction plan.

3. The method for detecting human bony landmarks according to claim 2, wherein: The method of constructing a standard posture database and inputting the first posture correction coefficient and the standard posture database into a posture correction analysis model to obtain a first posture correction scheme includes: obtaining a standard posture image dataset based on the standard posture database; determining a standard posture image variation coefficient based on a static application scenario; performing data amplification on the standard posture image dataset based on an image processing algorithm, wherein the data amplification performs a variation output according to the standard posture image variation coefficient to obtain an amplified standard posture image dataset; and performing data expansion on the standard posture database based on the amplified standard posture image dataset.

4. An intelligent whole-body posture assessment system for implementing the intelligent human bony landmark detection method according to any one of claims 1 to 3, characterized in that: The evaluation system includes: a first obtaining unit, the first obtaining unit is used to identify the bone landmarks of the first user through an intelligent image acquisition device, obtain a set of static bone landmarks of the user, and dynamically mark the set of static bone landmarks of the user; a first determining unit, the first determining unit is used to determine a set of posture and body evaluation indicators based on the set of static bone landmarks of the user; a second obtaining unit, the second obtaining unit is used to obtain first evaluation pictures and video information through an image acquisition device, the first evaluation pictures and video information include static pictures and video information of the first user; a third obtaining unit, the third obtaining unit is used to determine a predetermined convolution kernel based on the set of posture and body evaluation indicators, and perform convolution on each frame of image information of the first evaluation pictures and video information according to the predetermined convolution kernel. A convolution calculation is performed to obtain a first convolution calculation result; a first processing unit is used to obtain first static posture feature information according to the first convolution calculation result, and generate a posture skeleton simulation diagram for dynamic demonstration according to the first static posture feature information; a fourth obtaining unit is used to input the first static posture feature information and the posture skeleton simulation diagram into a first static posture analysis model for multi-dimensional analysis to obtain a first posture analysis result; a first generating unit is used to generate a first posture accuracy coefficient based on the first posture analysis result; a second processing unit is used to generate a first reminder instruction if the first posture accuracy coefficient is lower than a preset accuracy coefficient, and the first reminder instruction is used to remind the first user to correct the posture.

5. The intelligent whole-body posture assessment system according to claim 4, characterized in that: The posture evaluation system further includes: a fifth obtaining unit, the fifth obtaining unit being used to obtain, based on the first posture analysis result, an evaluation indicator that does not meet the standard in the posture and body evaluation indicator set; a sixth obtaining unit, the sixth obtaining unit being used to perform correlation analysis based on the evaluation indicator that does not meet the standard and the posture and body evaluation indicator set, and obtain an indicator correlation set; a seventh obtaining unit, the seventh obtaining unit being used to screen the indicators in the indicator correlation set that exceed a predetermined correlation threshold, and obtain a correlation correction indicator; an eighth obtaining unit, the eighth obtaining unit being used to supplement the first posture analysis result according to the correlation correction indicator, and obtain a second posture analysis result; and a ninth obtaining unit being used to perform correlation analysis based on the evaluation indicator that does not meet the standard and the posture and body evaluation indicator set, and obtain an indicator correlation set. The ninth obtaining unit is used to obtain the static vital sign information of the first user through the intelligent image collector device; the tenth obtaining unit is used to obtain the first human body depth image and gradient data according to the static vital sign information; the eleventh obtaining unit is used to perform statistics on the values of the first human body depth image and gradient data, and obtain a first posture evaluation index according to the statistical results; the first correction unit is used to correct the second posture analysis result according to the first posture evaluation index; the twelfth obtaining unit is used to obtain a first posture accuracy coefficient according to the difference between the first posture accuracy coefficient and the preset accuracy coefficient. a posture correction coefficient; a third processing unit, the third processing unit being used to construct a standard posture database, inputting the first posture correction coefficient and the standard posture database into a posture correction analysis model to obtain a first posture correction scheme; a fourth processing unit, the fourth processing unit being used to perform posture correction on the first user according to the first posture correction scheme; a thirteenth obtaining unit, the thirteenth obtaining unit being used to obtain a standard posture image data set according to the standard posture database; a second determining unit, the second determining unit being used to determine a standard posture image variation coefficient according to a static application scenario; a fourteenth obtaining unit, the fourteenth obtaining unit being used to process the standard posture image data set based on an image processing algorithm; a first user performing data augmentation on the standard posture database, wherein the data augmentation is changed and output according to the standard posture image change coefficient to obtain an augmented standard posture image data set; a first expansion unit, wherein the first expansion unit is used to perform data augmentation on the standard posture database according to the augmented standard posture image data set; a fifteenth obtaining unit, wherein the fifteenth obtaining unit is used to obtain basic physiological information of the first user; a first selection unit, wherein the first selection unit is used to select a posture evaluation mode according to the personal training goal of the first user; a first calling unit, wherein the first calling unit is used to call the first static posture analysis model from a static posture analysis model library based on the basic physiological information and the posture evaluation mode;A sixteenth obtaining unit is configured to obtain initial hidden layer values of the recurrent neural network and obtain a first input weight matrix based on the initial hidden layer values; a first correction unit is configured to use historical static posture feature information and historical posture skeletal simulation images as input layer information and train the recurrent neural network based on the input layer information and the first input weight matrix; and a first construction unit is configured to use the input layer information and the initial hidden layer values as the next hidden layer values and use historical posture analysis results as identification information for iterative training to construct the first static posture analysis model.

6. An electronic device, comprising: An image collector, a memory, and a processor, wherein the memory stores a computer program, and wherein the processor implements the steps of the method according to any one of claims 1 to 3 when executing the computer program.

7. A computer-readable storage medium having a computer program stored thereon, 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 3 are implemented.

Images