Dancing motion detection method and system based on human body key points
By collecting and standardizing 3D images and video frames during the dance process, and extracting and comparing dance movement features, the problem of inconsistent recognition results and low detection accuracy in dance teaching is solved, and precise analysis and prompts for dance movement deviations are achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- HUIZHIAN INFORMATION TECH CO LTD
- Filing Date
- 2025-12-31
- Publication Date
- 2026-05-19
AI Technical Summary
In current dance teaching, relying on manual observation and subjective judgment to identify the quality of dance movements results in inconsistencies, low efficiency, and the inability to standardize. Furthermore, existing technologies fail to effectively consider the differences in body shape among students and the complexity of dance movements, leading to large identification biases and low detection accuracy.
By acquiring full-body 3D images of the target object, extracting the 3D position information of key points on the human body, and performing standardized processing, and combining the video frames of the dance process, extracting and comparing the features of actual and standard dance movements, generating dance movement deviation change features and prompt messages.
It achieves accurate recognition and standardized identification of dance movements, reduces recognition bias, and improves the accuracy and reliability of dance movement detection.
Smart Images
Figure CN122067307A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of intelligent motion recognition technology, and in particular to a method and system for detecting dance movements based on key points of the human body. Background Technology
[0002] Current dance instruction relies on teachers observing students' actual dance movements and subjectively judging the quality of their movements. This method is highly subjective, leading to inconsistent assessments of the correctness of the same student's movements by different teachers. It also suffers from inefficiency and a lack of standardized criteria. While existing technologies simplify the representation of the human dance process by using several points within the body as benchmarks, this approach fails to consider the differences in body shape among students and the diversity and complexity of movements during dance. This results in significant errors in movement recognition and a mismatch between the detection results and the actual dance rhythm, reducing the accuracy and reliability of movement detection. Summary of the Invention
[0003] Considering that existing dance movement recognition methods during dance instruction rely on visual observation and subjective evaluation to obtain movement deviation results, the detection results are highly uncertain and lack uniform standards. Furthermore, they cannot adapt to the complexity of dance movements in different scenarios, reducing the accuracy and reliability of movement detection. In view of the above problems, this invention is proposed to provide a dance movement detection method based on human key points that overcomes or at least partially solves the above problems, including:
[0004] Acquire a full-body 3D image of the target object, extract 3D position information of several key human body points from the full-body 3D image; after preprocessing the 3D position information, generate several standardized position information corresponding to the several key human body points;
[0005] The process involves acquiring dance footage of the target object and obtaining all image frames corresponding to the dance footage; extracting actual dance movement features from the image frames based on the standardized position information; and extracting standard dance movement features from standard dance footage based on the standardized position information.
[0006] Based on the actual dance movement characteristics and the standard dance movement characteristics, determine the movement deviation change characteristics of the target object during the dance process; based on the movement deviation change characteristics, generate dance movement prompt messages for the target object.
[0007] Optionally, a full-body 3D image of the target object is acquired, and 3D position information of several key human body points is extracted from the full-body 3D image; after preprocessing the 3D position information, several standardized position information corresponding to the several key human body points is generated, including:
[0008] Acquire a full-body 3D image of the target object in a baseline body posture, identify the body structure contour features of the target object from the full-body 3D image, mark several key body parts of the target object based on the body structure contour features, and determine the 3D position information of key points of the human body based on the spatial position of the key body parts on the target object's body.
[0009] Based on the three-dimensional coordinate data of the three-dimensional position information corresponding to the aforementioned key human body points, the three-dimensional position information is preprocessed by normalization to generate several standardized position information corresponding to the aforementioned key human body points.
[0010] Optionally, the dance process images of the target object are acquired, and all image frames corresponding to the dance process images are obtained; actual dance movement features are extracted from the image frames based on the standardized position information; and standard dance movement features are extracted from standard dance images based on the standardized position information, including:
[0011] Three-dimensional images of the target object's dance process are acquired, and the three-dimensional images of the dance process are processed into frames to obtain all image frames arranged in time sequence; the standardized position information is mapped to the image frames, thereby extracting the pose vectors of several human key points from the image frames.
[0012] The standard dance video is divided into a queue of standard video frames, and the standardized position information is mapped to the standard video frames, thereby extracting the pose vectors of the human body key points from the standard video frames.
[0013] Optionally, based on the actual dance movement characteristics and the standard dance movement characteristics, the movement deviation change characteristics of the target object during the dance process are determined; based on the movement deviation change characteristics, a dance movement prompt message for the target object is generated, including:
[0014] A one-to-one correspondence is established between all image frames in the 3D image of the dance process and all standard image frames in the standard dance image; based on the correspondence, the image frames and the standard image frames are compared to obtain the posture vector deviation, thereby determining the movement deviation change characteristics of the target object in the dance process; wherein the deviation change characteristics include the change characteristics of the movement trajectory deviation in the time dimension;
[0015] Based on the characteristics of the movement deviation change, a threshold comparison of the movement trajectory offset is performed to determine the abnormal segment of the target object's dance movement, thereby generating a dance movement prompt message for the target object.
[0016] As one aspect of the present invention, embodiments of the present invention also provide a dance motion detection system based on human key points, comprising:
[0017] The key point localization module is used to acquire a full-body 3D image of the target object and extract the 3D position information of several key points of the human body from the full-body 3D image.
[0018] The location information preprocessing module is used to preprocess the three-dimensional location information to generate several standardized location information corresponding to the several key human body points;
[0019] The actual movement feature determination module is used to acquire dance process images of the target object, obtain all image frames corresponding to the dance process images, and extract actual dance movement features from the image frames based on the standardized position information.
[0020] The standard movement feature determination module is used to extract standard dance movement features from standard dance images based on the standardized position information.
[0021] The deviation change determination module is used to determine the movement deviation change characteristics of the target object during the dance process based on the actual dance movement characteristics and the standard dance movement characteristics;
[0022] The prompting module is used to generate dance movement prompt messages for the target object based on the movement deviation change characteristics.
[0023] Optionally, the key point localization module is used to acquire a full-body three-dimensional image of the target object, and extract three-dimensional position information of several key points of the human body from the full-body three-dimensional image, including:
[0024] Acquire a full-body 3D image of the target object in a baseline body posture, identify the body structure contour features of the target object from the full-body 3D image, mark several key body parts of the target object based on the body structure contour features, and determine the 3D position information of key points of the human body based on the spatial position of the key body parts on the target object's body.
[0025] The location information preprocessing module is used to preprocess the three-dimensional location information and generate several standardized location information corresponding to the several key human body points, including:
[0026] Based on the three-dimensional coordinate data of the three-dimensional position information corresponding to the aforementioned key human body points, the three-dimensional position information is preprocessed by normalization to generate several standardized position information corresponding to the aforementioned key human body points.
[0027] Optionally, the actual movement feature determination module is used to acquire dance process images of the target object, obtain all image frames corresponding to the dance process images; and extract actual dance movement features from the image frames based on the standardized position information, including:
[0028] Three-dimensional images of the target object's dance process are acquired, and the three-dimensional images of the dance process are processed into frames to obtain all image frames arranged in time sequence; the standardized position information is mapped to the image frames, thereby extracting the pose vectors of several human key points from the image frames.
[0029] The standard movement feature determination module is used to extract standard dance movement features from standard dance images based on the standardized position information, including:
[0030] The standard dance video is divided into a queue of standard video frames, and the standardized position information is mapped to the standard video frames, thereby extracting the pose vectors of the human body key points from the standard video frames.
[0031] Optionally, the deviation change determination module is used to determine the movement deviation change characteristics of the target object during the dance process based on the actual dance movement characteristics and the standard dance movement characteristics, including:
[0032] A one-to-one correspondence is established between all image frames in the 3D image of the dance process and all standard image frames in the standard dance image; based on the correspondence, the image frames and the standard image frames are compared to obtain the posture vector deviation, thereby determining the movement deviation change characteristics of the target object in the dance process; wherein the deviation change characteristics include the change characteristics of the movement trajectory deviation in the time dimension;
[0033] The prompting module is used to generate dance movement prompt messages for the target object based on the movement deviation change characteristics, including:
[0034] Based on the characteristics of the movement deviation change, a threshold comparison of the movement trajectory offset is performed to determine the abnormal segment of the target object's dance movement, thereby generating a dance movement prompt message for the target object.
[0035] The beneficial effects of the above-mentioned technical solutions provided in the embodiments of the present invention include at least the following:
[0036] This invention provides a method and system for detecting dance movements based on human key points. The system acquires a full-body 3D image of a target object, extracts 3D position information of several human key points from the full-body 3D image, preprocesses the 3D position information to generate standardized position information corresponding to the key points, acquires dance process images of the target object, and obtains all image frames corresponding to the dance process images. Based on the standardized position information, actual dance movement features are extracted from the image frames; based on the standardized position information, standard dance movement features are extracted from standard dance images; based on the actual dance movement features and standard dance movement features, the movement deviation changes of the target object during the dance process are determined, thereby generating dance movement prompt messages. By uniformly standardizing the 3D position information of human key points and mapping it to the actual dance process images and standard dance process images for posture comparison, accurate dance movement deviations are obtained.
[0037] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings.
[0038] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description
[0039] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:
[0040] Figure 1 This is a flowchart illustrating the dance motion detection method based on human key points provided in this embodiment of the invention.
[0041] Figure 2 This is a schematic diagram of the structure of the dance motion detection system based on human key points provided in an embodiment of the present invention. Detailed Implementation
[0042] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.
[0043] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," "outer," "far," "near," "front," and "rear," etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the accompanying drawings and are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.
[0044] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.
[0045] Please see Figure 1 As shown, an embodiment of this application provides a dance motion detection method based on human key points. This dance motion detection method based on human key points includes:
[0046] Acquire a full-body 3D image of the target object, extract the 3D position information of several key human body points from the full-body 3D image; after preprocessing the 3D position information, generate several standardized position information corresponding to several key human body points;
[0047] The process involves acquiring dance video footage of the target subject and obtaining all corresponding image frames; extracting actual dance movement features from the image frames based on standardized position information; and extracting standard dance movement features from standard dance video footage based on standardized position information.
[0048] Based on the characteristics of actual dance movements and standard dance movements, determine the characteristics of movement deviation changes of the target object during the dance process; based on the characteristics of movement deviation changes, generate dance movement prompt messages for the target object.
[0049] The beneficial effects of the above embodiments are that the dance movement detection method based on human body key points uniformly and standardizedly processes the three-dimensional position information of human body key points, maps it to the actual dance process image and the standard dance process image for posture comparison, and obtains accurate dance movement deviations.
[0050] In another embodiment, a full-body 3D image of the target object is acquired, and 3D position information of several key human body points is extracted from the full-body 3D image; after preprocessing the 3D position information, several standardized position information corresponding to the several key human body points is generated, including:
[0051] Acquire a full-body 3D image of the target object in a baseline body posture, identify the body structure contour features of the target object from the full-body 3D image; based on the body structure contour features, mark several key body parts of the target object; based on the spatial position of the key body parts on the target object's body, determine the 3D position information of key points of the human body;
[0052] Based on the three-dimensional coordinate data of the three-dimensional position information corresponding to several key points of the human body, the three-dimensional position information is normalized and preprocessed to generate several standardized position information corresponding to several key points of the human body.
[0053] The beneficial effects of the above embodiments, in practical operation, allow the target object to be instructed to stand at attention, and simultaneously acquire a full-body 3D image of the target object in the current posture. The body structure contour features of the target object are then identified from the full-body 3D image. It is understood that human joints are key body parts that support smooth and flexible dance movements, and they directly determine the range and posture of the dance. Therefore, using human joints as body landmarks for the target object reduces the difficulty of recognizing the target object's dance movements, providing a more concise and accurate representation of the target object's dance movements. Specifically, based on the body structure contour features of the target object, several key body parts (i.e., key joints, such as wrist, elbow, shoulder, hip, knee, and ankle joints) are identified. Physiological structural curves are simplified for each key body part, determining the corresponding human key points, and the 3D position information of each human key point on the target object's body is determined. This 3D position information may include x, y, and z coordinate values.
[0054] Different target subjects have different heights and body shapes, resulting in different 3D position information for the same joint on different subjects' bodies. Current technologies do not consider the impact of height and body shape differences on the 3D position of joints. Therefore, this study obtains the 3D coordinate data of several key human body points, and identifies the maximum value x among the three coordinates (x, y, z) of all key human body points. max y max z max and minimum value x min y min z min Then, based on the above maximum value x max y maxz max and minimum value x min y min z min The three-dimensional position information of each human body key point is preprocessed by normalization to generate standardized position information (x) for each human body key point. standard y standard , z standard This provides a reference benchmark for extracting the actual human body movements and postures at each time point from the actual dance process of the target object.
[0055] In another embodiment, the dance process images of the target object are acquired, and all image frames corresponding to the dance process images are obtained; based on standardized position information, actual dance movement features are extracted from the image frames; based on standardized position information, standard dance movement features are extracted from standard dance images, including:
[0056] The process involves acquiring 3D images of the target object's dance process, processing the 3D images into frames to obtain all image frames arranged in chronological order, mapping standardized position information onto the image frames, and extracting pose vectors of several key human body points from the image frames.
[0057] Standard dance footage is divided into a queue of standard image frames, and standardized position information is mapped to the standard image frames to extract the pose vectors of human key points from the standard image frames.
[0058] The beneficial effects of the above embodiments are as follows: In actual operation, the actual dance process of the target object is captured in three dimensions to obtain a three-dimensional dynamic image of the actual dance process. This three-dimensional dynamic image is then processed into frames to obtain all corresponding image frames. It can be understood that each image frame corresponds to a point in time in the actual dance process. The standardized position information is mapped to each image frame, thus marking all key points of the target object on the screen of each image frame, ensuring that each key point has a corresponding spatial position and posture distribution. Furthermore, the posture vectors of all key points in each image frame are determined in the three-dimensional coordinate system of the target object's actual dance space. Similarly, the standard dance image is divided into a queue containing several standard image frames, and the standardized position information is mapped to each standard image frame. This marks all key points of the target object on the screen of each standard image frame, ensuring that each key point has a corresponding spatial position and posture distribution. This determines the posture vectors of all key points in each standard image frame, providing a comparative data basis for subsequently determining deviations in the target object's actual dance movements.
[0059] In another embodiment, based on the characteristics of actual dance movements and the characteristics of standard dance movements, the movement deviation changes of the target object during the dance process are determined; based on the movement deviation changes, a dance movement prompt message for the target object is generated, including:
[0060] The one-to-one correspondence between all image frames in the 3D image of the dance process and all standard image frames in the standard dance image is determined. Based on the correspondence, the image frames and standard image frames are compared to obtain the posture vector deviation, thereby determining the movement deviation change characteristics of the target object in the dance process. The deviation change characteristics include the change characteristics of the movement trajectory deviation in the time dimension.
[0061] Based on the characteristics of movement deviation changes, the threshold of movement trajectory offset is compared to determine the abnormal segment of the target object's dance movement, thereby generating a dance movement prompt message for the target object.
[0062] The beneficial effects of the above embodiments are that the actual dance process of the target object corresponds to the same dance content as the standard dance process. This establishes a one-to-one correspondence between all image frames in the 3D image of the dance process and all standard image frames in the standard dance image. This correspondence refers to the image frames and standard image frames corresponding to the same dance movements within the same dance content. By comparing these image frames with the standard image frames, the posture vector deviation is obtained, thereby determining the time-dimensional variation characteristics of the target object's movement trajectory deviation during the dance process. Based on these variation characteristics, the trajectory deviation value corresponding to each dance movement of the target object is determined. If the trajectory deviation value is greater than a preset deviation threshold, the dance movement is identified as an abnormal segment, and a corresponding prompt message is generated for this abnormal segment, achieving precise location of abnormal dance movements.
[0063] Please see Figure 2 As shown, an embodiment of this application provides a dance motion detection system based on human key points. This dance motion detection system based on human key points includes:
[0064] The key point localization module is used to acquire a full-body 3D image of the target object and extract the 3D position information of several key points of the human body from the full-body 3D image.
[0065] The location information preprocessing module is used to preprocess the three-dimensional location information to generate several standardized location information corresponding to several human body key points;
[0066] The actual movement feature determination module is used to acquire dance process images of the target object, obtain all image frames corresponding to the dance process images, and extract actual dance movement features from the image frames based on standardized position information.
[0067] The standard movement feature determination module is used to extract standard dance movement features from standard dance images based on standardized position information;
[0068] The deviation change determination module is used to determine the movement deviation change characteristics of the target object during the dance process based on the actual dance movement characteristics and the standard dance movement characteristics.
[0069] The prompting module is used to generate dance movement prompt messages for the target object based on the characteristics of movement deviation changes.
[0070] The beneficial effects of the above embodiments are that the dance movement detection system based on human body key points obtains accurate dance movement deviations by uniformly standardizing the three-dimensional position information of human body key points and mapping it to the actual dance process image and the standard dance process image for posture comparison.
[0071] In another embodiment, the key point localization module is used to acquire a full-body 3D image of the target object, and extract the 3D position information of several key points of the human body from the full-body 3D image, including:
[0072] Acquire a full-body 3D image of the target object in a baseline body posture, identify the body structure contour features of the target object from the full-body 3D image; based on the body structure contour features, mark several key body parts of the target object; based on the spatial position of the key body parts on the target object's body, determine the 3D position information of key points of the human body;
[0073] The location information preprocessing module is used to preprocess the 3D location information and generate several standardized location information corresponding to several human body key points, including:
[0074] Based on the three-dimensional coordinate data of the three-dimensional position information corresponding to several key points of the human body, the three-dimensional position information is normalized and preprocessed to generate several standardized position information corresponding to several key points of the human body.
[0075] In another embodiment, the actual motion feature determination module is used to acquire dance process images of the target object, obtain all image frames corresponding to the dance process images, and extract actual dance motion features from the image frames based on standardized position information, including:
[0076] The process involves acquiring 3D images of the target object's dance process, processing the 3D images into frames to obtain all image frames arranged in chronological order, mapping standardized position information onto the image frames, and extracting pose vectors of several key human body points from the image frames.
[0077] The standard movement feature determination module is used to extract standard dance movement features from standard dance images based on standardized position information, including:
[0078] Standard dance footage is divided into a queue of standard image frames, and standardized position information is mapped to the standard image frames to extract the pose vectors of human key points from the standard image frames.
[0079] In another embodiment, the deviation change determination module is used to determine the movement deviation change characteristics of the target object during the dance process based on the actual dance movement characteristics and the standard dance movement characteristics, including:
[0080] The one-to-one correspondence between all image frames in the 3D image of the dance process and all standard image frames in the standard dance image is determined. Based on the correspondence, the image frames and standard image frames are compared to obtain the posture vector deviation, thereby determining the movement deviation change characteristics of the target object in the dance process. The deviation change characteristics include the change characteristics of the movement trajectory deviation in the time dimension.
[0081] The prompting module is used to generate dance movement prompting messages for the target object based on the characteristics of movement deviation changes, including:
[0082] Based on the characteristics of movement deviation changes, the threshold of movement trajectory offset is compared to determine the abnormal segment of the target object's dance movement, thereby generating a dance movement prompt message for the target object.
[0083] The operation and effect of the dance motion detection system based on human body key points of the present invention are consistent with the above-mentioned dance motion detection method based on human body key points, and the description of the dance motion detection system based on human body key points will not be repeated here.
[0084] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. This disclosure is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this disclosure is limited only by the appended claims. Thus, if these modifications and variations of the invention fall within the scope of the claims of the invention and their equivalents, the invention is also intended to include these modifications and variations.
Claims
1. A method for detecting dance movements based on human key points, characterized in that, include: Acquire a full-body 3D image of the target object, and extract the 3D position information of several key human body points from the full-body 3D image; After preprocessing the three-dimensional position information, several standardized position information corresponding to the several key human body points are generated; The process involves acquiring dance footage of the target object and obtaining all image frames corresponding to the dance footage; extracting actual dance movement features from the image frames based on the standardized position information; and extracting standard dance movement features from standard dance footage based on the standardized position information. Based on the actual dance movement characteristics and the standard dance movement characteristics, determine the movement deviation change characteristics of the target object during the dance process; based on the movement deviation change characteristics, generate dance movement prompt messages for the target object.
2. The dance movement detection method based on human key points as described in claim 1, characterized in that: Acquire a full-body 3D image of the target object, and extract the 3D position information of several key human body points from the full-body 3D image; After preprocessing the three-dimensional position information, several standardized position information corresponding to the several key human body points is generated, including: Acquire a full-body 3D image of the target object in a baseline body posture, and identify the body structure contour features of the target object from the full-body 3D image; Based on the body structure contour features, several key body parts of the target object are identified; based on the spatial position of the key body parts on the target object's body, the three-dimensional position information of key points of the human body is determined. Based on the three-dimensional coordinate data of the three-dimensional position information corresponding to the aforementioned key human body points, the three-dimensional position information is preprocessed by normalization to generate several standardized position information corresponding to the aforementioned key human body points.
3. The dance movement detection method based on human key points as described in claim 1, characterized in that: Acquire dance process images of the target object, and obtain all image frames corresponding to the dance process images; extract actual dance movement features from the image frames based on the standardized position information; extract standard dance movement features from standard dance images based on the standardized position information, including: Three-dimensional images of the target object's dance process are acquired, and the three-dimensional images of the dance process are processed into frames to obtain all image frames arranged in time sequence; the standardized position information is mapped to the image frames, thereby extracting the pose vectors of several human key points from the image frames. The standard dance video is divided into a queue of standard video frames, and the standardized position information is mapped to the standard video frames, thereby extracting the pose vectors of the human body key points from the standard video frames.
4. The dance movement detection method based on human key points as described in claim 3, characterized in that: Based on the actual dance movement characteristics and the standard dance movement characteristics, determine the movement deviation change characteristics of the target object during the dance process; Based on the characteristics of the movement deviation changes, a dance movement prompt message for the target object is generated, including: A one-to-one correspondence is established between all image frames in the 3D image of the dance process and all standard image frames in the standard dance image; based on the correspondence, the image frames and the standard image frames are compared to obtain the posture vector deviation, thereby determining the movement deviation change characteristics of the target object in the dance process; wherein the deviation change characteristics include the change characteristics of the movement trajectory deviation in the time dimension; Based on the characteristics of the movement deviation change, a threshold comparison of the movement trajectory offset is performed to determine the abnormal segment of the target object's dance movement, thereby generating a dance movement prompt message for the target object.
5. A dance motion detection system based on human key points, characterized in that, include: The key point localization module is used to acquire a full-body 3D image of the target object and extract the 3D position information of several key points of the human body from the full-body 3D image. The location information preprocessing module is used to preprocess the three-dimensional location information to generate several standardized location information corresponding to the several key human body points; The actual movement feature determination module is used to acquire dance process images of the target object, obtain all image frames corresponding to the dance process images, and extract actual dance movement features from the image frames based on the standardized position information. The standard movement feature determination module is used to extract standard dance movement features from standard dance images based on the standardized position information. The deviation change determination module is used to determine the movement deviation change characteristics of the target object during the dance process based on the actual dance movement characteristics and the standard dance movement characteristics; The prompting module is used to generate dance movement prompt messages for the target object based on the movement deviation change characteristics.
6. The dance motion detection system based on human key points as described in claim 5, characterized in that: The key point localization module is used to acquire a full-body 3D image of the target object and extract the 3D position information of several key points of the human body from the full-body 3D image, including: Acquire a full-body 3D image of the target object in a baseline body posture, identify the body structure contour features of the target object from the full-body 3D image, mark several key body parts of the target object based on the body structure contour features, and determine the 3D position information of key points of the human body based on the spatial position of the key body parts on the target object's body. The location information preprocessing module is used to preprocess the three-dimensional location information and generate several standardized location information corresponding to the several key human body points, including: Based on the three-dimensional coordinate data of the three-dimensional position information corresponding to the aforementioned key human body points, the three-dimensional position information is preprocessed by normalization to generate several standardized position information corresponding to the aforementioned key human body points.
7. The dance motion detection system based on human key points as described in claim 5, characterized in that: The actual motion feature determination module is used to acquire dance process images of the target object and obtain all image frames corresponding to the dance process images. Based on the standardized location information, actual dance movement features are extracted from the image frame, including: Three-dimensional images of the target object's dance process are acquired, and the three-dimensional images of the dance process are processed into frames to obtain all image frames arranged in time sequence; the standardized position information is mapped to the image frames, thereby extracting the pose vectors of several human key points from the image frames. The standard movement feature determination module is used to extract standard dance movement features from standard dance images based on the standardized position information, including: The standard dance video is divided into a queue of standard video frames, and the standardized position information is mapped to the standard video frames, thereby extracting the pose vectors of the human body key points from the standard video frames.
8. The dance motion detection system based on human key points as described in claim 7, characterized in that: The deviation change determination module is used to determine the movement deviation change characteristics of the target object during the dance process based on the actual dance movement characteristics and the standard dance movement characteristics, including: A one-to-one correspondence is established between all image frames in the 3D image of the dance process and all standard image frames in the standard dance image; based on the correspondence, the image frames and the standard image frames are compared to obtain the posture vector deviation, thereby determining the movement deviation change characteristics of the target object in the dance process; wherein the deviation change characteristics include the change characteristics of the movement trajectory deviation in the time dimension; The prompting module is used to generate dance movement prompt messages for the target object based on the movement deviation change characteristics, including: Based on the characteristics of the movement deviation change, a threshold comparison of the movement trajectory offset is performed to determine the abnormal segment of the target object's dance movement, thereby generating a dance movement prompt message for the target object.