Human body posture recognition system based on artificial intelligence
By using the motion capture sensor module, key point detection module and three-dimensional model module in the human posture recognition system, combined with artificial intelligence recognition training, the problem of relying on manual marking in the existing technology is solved, and high-precision and real-time human posture recognition and feedback is achieved.
Patent Information
- Application Number
- CN202411803817.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-10
- Publication Date
- 2025-03-28
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing human posture recognition methods rely on manual marking after data transmission of complex sensor equipment, resulting in high cost, complex operation and inability to meet the requirements of real-time and accuracy.
Using an artificial intelligence-based human posture recognition system, the motion capture sensor module is fitted to the key parts of the user's body, and combined with the key point detection module and the three-dimensional model module, the accurate capture and reconstruction of dynamic posture is achieved.
The system can capture and reconstruct the user's dynamic poses with high precision, provide real-time feedback, improve training or rehabilitation efficiency, adapt to different application scenarios, and continuously optimize performance through artificial intelligence identification training.
Smart Images
Figure CN119723663A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of computer vision, and in particular relates to a human body posture recognition system based on artificial intelligence. Background Art
[0002] With the rapid development of artificial intelligence technology, human posture recognition has been widely used in many fields such as video surveillance, motion analysis, and human-computer interaction. Traditional human posture recognition methods often rely on manual labeling after complex sensor equipment data transmission. These methods are not only costly and complicated to operate, but also cannot meet the requirements of real-time and accuracy. Therefore, we propose a human posture recognition system based on artificial intelligence. Summary of the invention
[0003] The purpose of the present invention is to provide a human posture recognition system based on artificial intelligence. By attaching a motion capture sensor module to the key parts of the user's body and combining a key point detection module and a three-dimensional model module, the system solves the problem that existing human posture recognition methods often rely on manual marking after complex sensor device data transmission. The system can accurately capture and reconstruct the user's dynamic posture. These methods are not only costly and complicated to operate, but also cannot meet the requirements of real-time and accuracy.
[0004] In order to solve the above technical problems, the present invention is achieved through the following technical solutions:
[0005] The present invention is a human body posture recognition system based on artificial intelligence, comprising a posture motion capture unit, a posture recognition unit and a user monitoring unit;
[0006] The gesture motion capture unit includes a motion capture sensor module, an earpiece, a data receiving module, a key point detection module, a three-dimensional model module, a key point marking module, an imaging preprocessing module and a feature extraction module; the motion capture sensor module is unidirectionally connected to the earpiece; the motion capture sensor module is unidirectionally connected to the data receiving module; the data receiving module is unidirectionally connected to the key point detection module; the key point detection module is unidirectionally connected to the three-dimensional model module; the three-dimensional model module is bidirectionally connected to the key point marking module; the three-dimensional model module is unidirectionally connected to the imaging preprocessing module; the imaging preprocessing module is unidirectionally connected to the feature extraction module; the gesture recognition unit includes a gesture tracking module, a gesture classification module, a behavior recognition module, a timing analysis module, a gesture evaluation module, a gesture correction module and a real-time feedback module; the feature extraction module is unidirectionally connected to the gesture tracking module ; The posture tracking module is unidirectionally connected to the posture classification module; the posture classification module is unidirectionally connected to the behavior recognition module; the behavior recognition module is unidirectionally connected to the timing analysis module; the timing analysis module is unidirectionally connected to the posture evaluation module; the posture evaluation module is unidirectionally connected to the posture correction module; the posture correction module is unidirectionally connected to the real-time feedback module; the user monitoring unit includes a user interaction interface, a microphone, an alarm, posture capture display information, an abnormal monitoring module, a data storage module and artificial intelligence recognition training; the user interaction interface is unidirectionally connected to the microphone, the alarm, posture capture display information, the abnormal monitoring module, the data storage module and artificial intelligence recognition training respectively; the microphone is unidirectionally connected to the earpiece; the user interaction interface is unidirectionally connected to the posture classification module; the artificial intelligence recognition training is unidirectionally connected to the posture classification module.
[0007] As a preferred technical solution, the motion capture sensor module is responsible for fitting to the body of the monitored user, monitoring the body movement trajectory and uploading the trajectory data; the data receiving module is responsible for receiving the data uploaded by the motion capture sensor module and parsing it; the key point detection module is responsible for recording each body node of the monitored user, such as the head, shoulders, elbows, wrists, fingers, torso, hips, knees, ankles and other joints, so as to record the nodes and identify the key point positions using a convolutional neural network model; the three-dimensional model module is responsible for establishing a three-dimensional model based on the contour data uploaded by the monitored user sensor and updating it in real time; the key point marking module is responsible for marking each body node detected by the key node detection module at the corresponding joints of the three-dimensional modeled body and running them synchronously; the imaging preprocessing module is responsible for preprocessing the collected data such as noise reduction, format conversion, and size standardization to improve the efficiency and accuracy of subsequent processing; the feature extraction module is responsible for extracting key features representing human posture from the preprocessed data, such as joint angles, limb positions, and movement directions.
[0008] As a preferred technical solution, the posture tracking module is responsible for continuously tracking the posture of the human body in motion and identifying posture changes. The module can use an algorithm based on target tracking, such as Kalman filtering; the posture classification module is responsible for determining the direction of posture analysis, and by setting the target group, the specific posture name can be accurately analyzed according to the behavior. They can be divided into fitness personnel, sports personnel, film and television public personnel, dancers and rehabilitation personnel. According to their occupations and groups, the names of various postures can be refined and distinguished; the behavior recognition module is responsible for posture behavior recognition of personnel in the current occupation, and identifies the name of the posture of the currently monitored user; the timing analysis module is responsible for By analyzing continuous posture data, the module can identify the time series patterns and dynamic trends of human body movements, which provides important information for understanding complex movements and predicting future postures; the posture evaluation module is responsible for evaluating the posture identified by the monitored user, thereby automatically detecting and identifying errors or deviations that may occur in the process, and scoring them, and distinguishing the gap from the perfect posture according to the score; the posture correction module is responsible for correcting the erroneous or deviant movement posture identified by the monitored user, and guiding it correctly to make the movement regular; the implementation feedback module is responsible for feeding back the posture recognition results of the current monitored user to the observing user or system in real time, so as to correct and improve the current posture in time.
[0009] As a preferred technical solution, the user interaction interface is responsible for providing the user with an interactive interface, allowing the user to interact with the system, perform settings and operations; the module is designed with an intuitive graphical user interface, and the user can operate through a touch screen or physical buttons, so that the user can understand the current user posture state; the microphone is responsible for receiving the sound emitted by the user, so as to facilitate the communication between the observing user and the monitored user; the alarm is responsible for issuing an alarm prompt after detecting deviations and errors in the user's current posture, so as to prompt the observing user to correct the monitored user; the posture capture display information is responsible for capturing the deviations in the posture of the monitored user, and displaying and marking them, so as to provide an accurate picture of the status of the monitored user. The abnormal monitoring module is responsible for monitoring whether the posture of the currently monitored user does not conform to the normal posture, so as to distinguish whether the recognition distortion is caused by node conflict, disappearance or offset, and then automatically monitor; the data storage module is responsible for storing and managing all raw data, processing data and posture recognition results, and adopts efficient database system and data security measures to ensure data integrity and privacy; the artificial intelligence recognition training is responsible for inputting a large amount of labeled data for training, adjusting its internal parameters to minimize the difference between the predicted value and the actual value, deploying the trained artificial intelligence into practical applications, and monitoring the system's operating status to ensure the stability and reliability of the model in the actual environment.
[0010] The present invention has the following beneficial effects:
[0011] The present invention attaches the motion capture sensor module to the key parts of the user's body, and combines the key point detection module and the three-dimensional model module. There is no need to rely on complex sensor equipment to upload data and then manually mark it. The system can accurately capture and reconstruct the user's dynamic posture. This high-precision posture capture provides a reliable basis for subsequent posture analysis and evaluation.
[0012] The present invention can display the posture recognition results to the user in real time through the real-time feedback module, so that the user can quickly understand whether his or her posture is correct and how to adjust it. This instant feedback mechanism greatly improves the efficiency of training or rehabilitation. The posture evaluation module and the posture correction module can provide personalized evaluation and correction suggestions according to the specific needs of different users. For athletes and rehabilitation patients, the system can provide customized posture analysis and correction solutions according to their specific needs.
[0013] The modular design of the system of the present invention enables it to adapt to different application scenarios, such as fitness guidance, sports analysis, film and television production, dance teaching and rehabilitation therapy. This flexibility enables the system to meet the diverse needs of different users. An easy-to-operate user interaction interface is provided through the user monitoring unit, allowing users to easily interact with the system, set personal parameters, and obtain gesture capture display information. The addition of a microphone and an alarm enhances the interactivity between the user and the system and improves the user experience.
[0014] The present invention provides an artificial intelligence recognition training module to allow the system to improve the accuracy and efficiency of gesture recognition through continuous learning and adjustment. This continuous learning ability enables the system to continuously optimize performance over time and adapt to new user needs and environmental changes.
[0015] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for describing the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.
[0017] Figure 1 The present invention is a flow chart of a human posture recognition system based on artificial intelligence. DETAILED DESCRIPTION
[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0019] See also Figure 1 As shown, the present invention is a human posture recognition system based on artificial intelligence, including a posture motion capture unit, a posture recognition unit and a user monitoring unit;
[0020] The gesture motion capture unit includes a motion capture sensor module, an earpiece, a data receiving module, a key point detection module, a three-dimensional model module, a key point marking module, an imaging preprocessing module and a feature extraction module; the motion capture sensor module is unidirectionally connected to the earpiece; the motion capture sensor module is unidirectionally connected to the data receiving module; the data receiving module is unidirectionally connected to the key point detection module; the key point detection module is unidirectionally connected to the three-dimensional model module; the three-dimensional model module is bidirectionally connected to the key point marking module; the three-dimensional model module is unidirectionally connected to the imaging preprocessing module; the imaging preprocessing module is unidirectionally connected to the feature extraction module; the gesture recognition unit includes a gesture tracking module, a gesture classification module, a behavior recognition module, a timing analysis module, a gesture evaluation module, a gesture correction module and a real-time feedback module; the feature extraction module is unidirectionally connected to the gesture tracking module The posture tracking module is unidirectionally connected to the posture classification module; the posture classification module is unidirectionally connected to the behavior recognition module; the behavior recognition module is unidirectionally connected to the timing analysis module; the timing analysis module is unidirectionally connected to the posture evaluation module; the posture evaluation module is unidirectionally connected to the posture correction module; the posture correction module is unidirectionally connected to the real-time feedback module; the user monitoring unit includes a user interaction interface, a microphone, an alarm, posture capture display information, an abnormal monitoring module, a data storage module and artificial intelligence recognition training; the user interaction interface is unidirectionally connected to the microphone, the alarm, posture capture display information, the abnormal monitoring module, the data storage module and artificial intelligence recognition training respectively; the microphone is unidirectionally connected to the earpiece; the user interaction interface is unidirectionally connected to the posture classification module; the artificial intelligence recognition training is unidirectionally connected to the posture classification module.
[0021] The motion capture sensor module is responsible for fitting onto the body of the monitored user, monitoring the body's motion trajectory and uploading the trajectory data; the data receiving module is responsible for receiving the data uploaded by the motion capture sensor module and parsing it; the key point detection module is responsible for recording each body node of the monitored user, such as the head, shoulders, elbows, wrists, fingers, torso, hips, knees, ankles and other joints, thereby recording the nodes and using the convolutional neural network model to identify the key point positions; the three-dimensional model module is responsible for establishing a three-dimensional model based on the contour data uploaded by the monitored user's sensor and updating it in real time; the key point marking module is responsible for marking each body node detected by the key node detection module at the corresponding joints of the three-dimensional modeled body and running them synchronously; the imaging preprocessing module is responsible for preprocessing the collected data such as noise reduction, format conversion, and size standardization to improve the efficiency and accuracy of subsequent processing; the feature extraction module is responsible for extracting key features that characterize human posture from the preprocessed data, such as joint angles, limb positions, and movement directions.
[0022] The posture tracking module is responsible for continuously tracking the posture of the moving human body and identifying posture changes. This module can use algorithms based on target tracking, such as Kalman filtering; the posture classification module is responsible for determining the direction of posture analysis. By setting the target group, the specific posture name can be accurately analyzed according to the behavior. They can be divided into fitness personnel, sports personnel, film and television public personnel, dancers and rehabilitation personnel. According to their occupations and groups, the names of various postures can be refined and distinguished; the behavior recognition module is responsible for posture behavior recognition of personnel in the current occupation and identifying the name of the posture of the currently monitored user; the timing analysis module is responsible for analyzing continuous Posture data, this module can identify the time series patterns and dynamic trends of human movements, which provides important information for understanding complex movements and predicting future postures; the posture evaluation module is responsible for evaluating the posture identified by the monitored user, so as to automatically detect and identify errors or deviations that may occur in the process, and score them, and distinguish the gap from the perfect posture according to the score; the posture correction module is responsible for correcting the erroneous or deviant posture identified by the monitored user, and guiding it correctly to make the movement regular; the implementation feedback module is responsible for feeding back the posture recognition results of the current monitored user to the observing user or system in real time, so as to correct and improve the current posture in time.
[0023] The user interaction interface is responsible for providing an interactive interface for users to interact with the system, set up and operate. The module is designed with an intuitive graphical user interface, which users can operate through the touch screen or physical buttons, so that users can understand the current user posture status. The microphone is responsible for receiving the sound emitted by the user, so as to facilitate the communication between the observing user and the monitored user. The alarm is responsible for giving an alarm prompt after detecting deviations and errors in the current user posture, so as to prompt the observing user to correct the monitored user. The posture capture display information is responsible for capturing the deviation of the monitored user's posture, displaying and marking it, so as to prompt the observing user. The abnormal monitoring module is responsible for monitoring whether the current monitored user's posture conforms to the normal posture, so as to distinguish whether it is due to node conflict, disappearance or offset, resulting in recognition distortion, so as to perform automatic monitoring. The data storage module is responsible for storing and managing all raw data, processing data and posture recognition results, and adopting efficient database systems and data security measures to ensure data integrity and privacy. The artificial intelligence recognition training is responsible for inputting a large amount of labeled data for training, adjusting its internal parameters to minimize the difference between the predicted value and the actual value, deploying the trained artificial intelligence into practical applications, and monitoring the system's operating status to ensure the stability and reliability of the model in the actual environment.
[0024] A specific application of this embodiment is: by attaching the motion capture sensor module to the key parts of the user's body, and combining the key point detection module and the three-dimensional model module, there is no need to continue to rely on complex sensor equipment to upload data and then manually mark it. The system can accurately capture and reconstruct the user's dynamic posture. This high-precision posture capture provides a reliable basis for subsequent posture analysis and evaluation. By selecting the monitoring posture direction, the monitored action posture is distinguished and identified, so that it can adapt to different application scenarios, such as fitness guidance, sports analysis, film and television production, dance teaching and rehabilitation treatment. This flexibility enables the system to meet the diverse needs of different users, allowing users to quickly understand whether their posture is correct and how to adjust it. This instant feedback mechanism greatly improves the efficiency of training or rehabilitation. An easy-to-operate user interaction interface is provided through the user monitoring unit, allowing users to easily interact with the system, set personal parameters, and obtain posture capture display information. The addition of microphones and alarms enhances the interactivity between users and the system. Artificial intelligence recognition training allows the system to improve the accuracy and efficiency of posture recognition through continuous learning and adjustment. This continuous learning ability enables the system to continuously optimize performance over time, adapt to new user needs and environmental changes, and improve the practicality of the system.
[0025] It is worth noting that in the above system embodiment, the various units included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.
[0026] In addition, those skilled in the art can understand that all or part of the steps in the above-mentioned embodiments can be completed by instructing related hardware through a program, and the corresponding program can be stored in a computer-readable storage medium.
[0027] The preferred embodiments of the present invention disclosed above are only used to help illustrate the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation methods described. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.
Claims
1. A human posture recognition system based on artificial intelligence, comprising a posture motion capture unit, a posture recognition unit and a user monitoring unit; characterized in that: The gesture motion capture unit includes a motion capture sensor module, an earpiece, a data receiving module, a key point detection module, a three-dimensional model module, a key point marking module, an imaging preprocessing module and a feature extraction module; The motion capture sensor module is unidirectionally connected to the earpiece; the motion capture sensor module is unidirectionally connected to the data receiving module; the data receiving module is unidirectionally connected to the key point detection module; the key point detection module is unidirectionally connected to the three-dimensional model module; the three-dimensional model module is bidirectionally connected to the key point marking module; the three-dimensional model module is unidirectionally connected to the imaging preprocessing module; the imaging preprocessing module is unidirectionally connected to the feature extraction module; The posture recognition unit includes a posture tracking module, a posture classification module, a behavior recognition module, a timing analysis module, a posture evaluation module, a posture correction module and a real-time feedback module; the feature extraction module is unidirectionally connected to the posture tracking module; the posture tracking module is unidirectionally connected to the posture classification module; the posture classification module is unidirectionally connected to the behavior recognition module; the behavior recognition module is unidirectionally connected to the timing analysis module; the timing analysis module is unidirectionally connected to the posture evaluation module; the posture evaluation module is unidirectionally connected to the posture correction module; the posture correction module is unidirectionally connected to the real-time feedback module; The user monitoring unit includes a user interaction interface, a microphone, an alarm, gesture capture display information, an abnormality monitoring module, a data storage module and artificial intelligence recognition training; the user interaction interface is unidirectionally connected to the microphone, the alarm, gesture capture display information, the abnormality monitoring module, the data storage module and the artificial intelligence recognition training respectively; the microphone is unidirectionally connected to the earpiece; the user interaction interface is unidirectionally connected to the gesture classification module; the artificial intelligence recognition training is unidirectionally connected to the gesture classification module.
2. The human body posture recognition system based on artificial intelligence according to claim 1, characterized in that: The motion capture sensor module is responsible for fitting onto the body of the monitored user, monitoring the body movement trajectory and uploading the trajectory data; The data receiving module is responsible for receiving the data uploaded by the motion capture sensor module and analyzing it; The key point detection module is responsible for recording various body nodes of the monitored user, such as the head, shoulders, elbows, wrists, fingers, torso, hips, knees, ankles and other joints, so as to record the nodes and identify the key point positions using a convolutional neural network model; The three-dimensional model module is responsible for establishing a three-dimensional model based on the profile data uploaded by the monitored user sensor and updating it in real time; The key point marking module is responsible for marking each body node detected by the key node detection module at the corresponding joint of the three-dimensional modeled body and performing synchronous operation; The imaging preprocessing module is responsible for preprocessing the collected data such as noise reduction, format conversion, and size standardization to improve the efficiency and accuracy of subsequent processing; The feature extraction module is responsible for extracting key features representing human posture, such as joint angles, limb positions and movement directions, from the preprocessed data.
3. The human body posture recognition system based on artificial intelligence according to claim 2, characterized in that: The posture tracking module is responsible for continuously tracking the posture of the moving human body and identifying posture changes. The module can adopt an algorithm based on target tracking, such as Kalman filtering; The posture classification module is responsible for determining the direction of posture analysis. By setting the target group, the specific posture name can be accurately analyzed according to the behavior. They can be divided into fitness personnel, sports personnel, film and television public personnel, dancers and rehabilitation personnel. According to their occupations and groups, the names of various postures can be refined and the postures can be distinguished; The behavior recognition module is responsible for performing posture behavior recognition on the personnel of the current occupation and identifying the name of the posture of the currently monitored user; The timing analysis module is responsible for analyzing continuous posture data. This module can identify the time series pattern and dynamic trend of human body movements, which provides important information for understanding complex movements and predicting future postures. The posture evaluation module is responsible for evaluating the posture identified by the monitored user, thereby automatically detecting and identifying errors or deviations that may occur in the process, and scoring them, and distinguishing the gap from the perfect posture according to the score; The posture correction module is responsible for correcting the erroneous or deviant postures of the monitored user and guiding them correctly so that the movements are regular; The implementation feedback module is responsible for feeding back the posture recognition result of the current monitored user to the observing user or the system in real time, so as to correct and improve the current posture in time.
4. The human body posture recognition system based on artificial intelligence according to claim 3 is characterized in that: The user interaction interface is responsible for providing an interactive interface for the user, allowing the user to interact with the system, perform settings and operations; this module is designed with an intuitive graphical user interface, which the user can operate through the touch screen or physical buttons, so that the user can understand the current user posture status; The microphone is responsible for receiving the sound emitted by the user, thereby facilitating the communication between the observing user and the monitored user; The alarm is responsible for issuing an alarm prompt after detecting deviation and error in the user's current posture, thereby prompting the observing user to correct the monitored user; The posture capture display information is responsible for capturing the posture deviation of the monitored user, and displaying and marking it, thereby prompting the observing user; The abnormal monitoring module is responsible for monitoring whether the posture of the currently monitored user does not conform to the normal posture, so as to distinguish whether the recognition distortion is caused by node conflict, disappearance or offset, and then automatically monitor; The data storage module is responsible for storing and managing all raw data, processed data and gesture recognition results, and adopts efficient database systems and data security measures to ensure data integrity and privacy; The artificial intelligence recognition training is responsible for inputting a large amount of labeled data for training, adjusting its internal parameters to minimize the difference between the predicted value and the actual value, deploying the trained artificial intelligence into practical applications, and monitoring the operating status of the system to ensure the stability and reliability of the model in the actual environment.