Dance posture action recognition and interaction method and system based on visual calculation

Abstract

The invention relates to a dance posture action recognition and interaction method and system based on visual computing, and the method comprises the steps: obtaining the dance action information of a student and the rhythm information of dance music through a camera device when a recognition starting instruction is received; based on the rhythm information of the dance music, extracting posture contour information of the student at a plurality of preset moments from the dance action information; comparing the extracted attitude contour information with pre-stored standard attitude contour information, and outputting a comparison result; based on a comparison result, screening out attitude contour information of which the coincidence degree with the standard attitude contour information is lower than a preset threshold value, extracting the standard attitude contour information corresponding to the attitude contour information, and generating an action correction data packet; and sending the action correction data packet to an interaction terminal. The method has the effects of accurately judging whether the dance posture action made by the student is standard or not and feeding back to the student in time.

Description

Technical Field

[0001] The present application relates to the technical field of dance posture and action recognition, and in particular to a dance posture and action recognition, interaction method and system based on visual computing. Background Art

[0002] At present, dance movement teaching is done by dance teachers directly interacting with students through language and body language, mainly through movement demonstration and correction. Students observe and imitate the dance teacher's movements, and then receive guidance and correction from the dance teacher to achieve the effect of practice. This method requires students to learn face-to-face with teachers, and dance teachers need to spend a lot of time and energy to correct students' postures and movements one by one, which increases the teacher's workload. When there are many students, it may be difficult for dance teachers to take care of each student, resulting in students having difficulty in obtaining timely feedback during practice. Without accurate and timely correction from professional teachers, dance practice is likely to be delayed, which will greatly reduce the efficiency of practice. Summary of the invention

[0003] In order to accurately judge whether the dance postures and movements performed by students are standard and provide timely feedback to students, this application provides a dance posture and movement recognition and interaction method and system based on visual computing.

[0004] The above-mentioned invention objective of the present application is achieved through the following technical solutions:

[0005] A dance posture and action recognition and interaction method based on visual computing, comprising the steps of:

[0006] When receiving the recognition start instruction, the student's dance movement information and the rhythm information of the dance music are obtained through the camera device;

[0007] Based on the rhythm information of dance music, the posture contour information of students at several preset moments is extracted from the dance movement information;

[0008] Compare the extracted posture profile information with the pre-stored standard posture profile information, and output the comparison result;

[0009] Based on the comparison results, the posture contour information whose overlap with the standard posture contour information is lower than a preset threshold is screened out, and the standard posture contour information corresponding to the posture contour information is extracted to generate a motion correction data packet;

[0010] The interactive terminal reads the correction data packet in real time.

[0011] By adopting the above technical scheme, the dance movement information of students and the rhythm information of dance music are obtained through the camera device, and the dance movement information of students can be obtained instantly. By comparing the standard posture contour information, the dance movement deviation of students can be accurately identified. According to the movement characteristics of each student, a personalized correction data packet is provided, which is conducive to teaching students in accordance with their aptitude. The interactive terminal reads the correction data packet in real time, enhances the interactivity with students, and improves the correction experience. The method realizes real-time recognition and interactive guidance of students' dance movements through visual computing technology, which can improve the efficiency and quality of dance teaching in the scenarios of remote teaching and autonomous learning.

[0012] In a preferred example of the present application, when the recognition start instruction is received, the step of obtaining the student's dance movement information and the rhythm information of the dance music through the camera device specifically includes the following steps:

[0013] When receiving the recognition start instruction, sending the dance music screening list to the interactive terminal;

[0014] For the dance music selected by the user, extract the corresponding pre-stored dance music recording and the corresponding standard dance movement information;

[0015] Using an audio device to play a dance music recording and collect rhythm information of the dance music, the rhythm information including beats in the dance music and a timestamp of each beat;

[0016] The student's dance movement information is obtained through a camera device, and the dance movement information includes all dance movement frames of the student.

[0017] By adopting the above technical solution, using audio equipment to play dance music recordings and collect the rhythm information of dance music, the synchronization between dance movements and music rhythm is improved, making the recognition results more accurate. For the dance music selected by the user, corresponding standard movement information is provided, which enhances the pertinence and standardization of teaching content. All dance movement frames are collected to improve the integrity of movement information, which helps to comprehensively analyze students' dance posture performance.

[0018] In a preferred example of the present application, the step of extracting the posture contour information of the student at several preset moments from the dance movement information based on the rhythm information of the dance music specifically includes the following steps:

[0019] Based on the rhythm information of the dance music and the standard dance movement information, mark the timestamp of the beat corresponding to each standard movement frame;

[0020] According to several key postures in the preset dance movements, standard action frames corresponding to the several key postures are obtained, and the timestamps of the corresponding beats are output;

[0021] Based on the output beat timestamp, the dance action frames corresponding to the students at the timestamp are extracted from the dance action information;

[0022] The posture contour information is output based on the extracted dance action frame, and the standard posture contour information is output based on the standard action frame.

[0023] By adopting the above technical solution, standard action frames corresponding to several key postures are obtained, and the timestamps of the corresponding beats are output, thereby realizing standardized analysis of dance movements. Focusing on key postures helps to quickly locate major problems in dance movements, and extracting corresponding dance movement frames based on timestamps improves the accuracy of posture contour information extraction. This method realizes accurate analysis of students' key dance movements by marking and extracting key postures.

[0024] In a preferred example of the present application, the step of comparing the extracted posture profile information with the pre-stored standard posture profile information and outputting the comparison result specifically includes the following steps:

[0025] Preset personalized body proportion models for students and store them in the interactive terminal;

[0026] Extracting the body proportion model of the corresponding student according to the user identity information sent by the interactive terminal;

[0027] Dynamically adjust the standard posture contour information based on the body proportion model, and correct the standard posture contour information by automatic scaling and deformation;

[0028] The extracted posture profile information is compared with the corrected standard posture profile information, and the coincidence data of the two is output.

[0029] By adopting the above technical solution, the standard posture is adjusted according to the student's body proportion model, so that action recognition is more in line with personal characteristics. Through automatic scaling and deformation, the accuracy of matching the standard posture contour information with the actual action is improved. It can adapt to students with different physical characteristics, improve the universal applicability of the action recognition system, and reduce the possibility of misjudgment due to the different body proportions of students and the standard posture contour.

[0030] In a preferred example of the present application, the step of dynamically adjusting the standard posture contour information based on the body proportion model and correcting the standard posture contour information by automatic scaling and deformation specifically includes the following steps:

[0031] Identify key body parts in a scale model of the student's body, including the head, neck, shoulders, elbows, wrists, hips, knees, and ankles;

[0032] Extract the three-dimensional coordinates of each key body part, calculate the distance ratio between each three-dimensional coordinate, and output the proportion correction parameter;

[0033] According to the proportion correction parameter, the standard posture contour information is corrected by automatic scaling and deformation, and the corrected standard posture contour information is output.

[0034] By adopting the above technical solution, the standard posture contour is corrected according to the individual body proportions, so that the standard posture contour information is more in line with the student's actual body shape, reducing the possibility of misjudgment due to the different body proportions between the student and the standard posture contour.

[0035] In a preferred example of the present application, the step of screening out posture contour information whose overlap with the standard posture contour information is lower than a preset threshold based on the comparison result, extracting the standard posture contour information corresponding to the posture contour information, and generating a motion correction data packet specifically includes the following steps:

[0036] When the coincidence data between the posture profile information and the corrected standard posture profile information is lower than a preset threshold, the posture profile information is associated with a correction mark;

[0037] Based on the posture contour information with the correction mark, extract the corresponding standard posture contour information;

[0038] Analyze the difference between posture profile information and standard posture profile information, and output correction prompt information, including deviations in position, angle, movement range, etc.;

[0039] Pack the correction prompt information to generate a correction data packet.

[0040] By adopting the above technical solution, correction is performed on posture contour information with low overlap, which improves the targeted nature of the correction. Through the correction prompt information, the specific aspects that students need to improve, such as position, angle, movement range, etc., are clarified. The generated correction data packet can be directly used to guide students to improve their dance movements, thereby improving the efficiency of correction.

[0041] In a preferred example of the present application: when the coincidence data between the posture profile information and the corrected standard posture profile information is lower than a preset threshold, after the step of associating the posture profile information with the correction mark, the step further includes:

[0042] The speed of generating the correction mark is recorded in real time during the playing of the dance music;

[0043] When the speed of generating the correction mark is greater than the preset threshold, the music playback is automatically stopped and a repeat practice confirmation request is sent to the interactive terminal;

[0044] The user selects whether to restart the dance movement practice based on the repeat practice confirmation request;

[0045] After receiving the repeat practice instruction from the interactive terminal, all correction marks are cleared and the music is played from the beginning.

[0046] By adopting the above technical solution, the speed of generating correction marks is recorded in real time, and the teaching process can be dynamically adjusted. When the speed of generating correction marks is greater than the preset threshold, it means that the student has too many posture problems when the dance reaches the current progress. At this time, there is little meaning in continuing to practice. It is necessary to practice the previous part repeatedly. By sending a repeated practice confirmation request, students are allowed to choose whether to repeat the practice according to the actual situation, which improves the flexibility and effectiveness of teaching.

[0047] The second object of the invention is achieved by the following technical solutions:

[0048] A dance posture and motion recognition and interaction system based on visual computing, comprising:

[0049] The recognition module is used to obtain the student's dance movement information and the rhythm information of the dance music through the camera device when receiving the recognition start instruction;

[0050] An extraction module is used to extract the posture contour information of the student at a number of preset moments from the dance movement information based on the rhythm information of the dance music;

[0051] A comparison module, used to compare the extracted posture profile information with the pre-stored standard posture profile information, and output the comparison result;

[0052] The feedback module is used to screen out posture profile information whose overlap with the standard posture profile information is lower than a preset threshold based on the comparison result, extract the standard posture profile information corresponding to the posture profile information, generate a motion correction data packet, and send the motion correction data packet to the interactive terminal.

[0053] By adopting the above technical solutions, each module works together to form a complete dance posture and movement recognition and interaction system.

[0054] The third objective of the present application is achieved through the following technical solutions:

[0055] A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the above-mentioned dance posture and action recognition and interaction method based on visual computing are implemented.

[0056] The fourth objective of the present application is achieved through the following technical solutions:

[0057] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the above-mentioned dance posture and action recognition and interaction method based on visual computing.

[0058] In summary, the present application includes at least one of the following beneficial technical effects:

[0059] 1. The camera device can obtain the students' dance movement information and the rhythm information of the dance music, and can instantly obtain the students' dance movement information. By comparing the standard posture contour information, the student's dance movement deviation can be accurately identified. According to the movement characteristics of each student, a personalized correction data package is provided, which is conducive to teaching students in accordance with their aptitude. The interactive terminal reads the correction data package in real time, enhances the interactivity with the students, and improves the correction experience. This method realizes real-time recognition and interactive guidance of students' dance movements through visual computing technology, which can improve the efficiency and quality of dance teaching in the scenarios of remote teaching and autonomous learning;

[0060] 2. Obtain standard action frames corresponding to several key postures and output the timestamps of the corresponding beats, realizing the standardized analysis of dance movements. Focusing on key postures helps to quickly locate the main problems in dance movements. Based on the timestamps, the corresponding dance movement frames are extracted, which improves the accuracy of posture contour information extraction. This method realizes the accurate analysis of students' key dance movements by marking and extracting key postures.

[0061] 3. The standard posture is adjusted according to the student's body proportion model to make the action recognition more in line with individual characteristics. Through automatic scaling and deformation, the accuracy of matching the standard posture contour information with the actual action is improved, which can adapt to students with different physical characteristics, improve the universal applicability of the action recognition system, and reduce the possibility of misjudgment due to the different body proportions of students and the standard posture contour;

[0062] 4. Real-time recording of the speed of correction mark generation can dynamically adjust the teaching process. When the speed of correction mark generation is greater than the preset threshold, it means that the student has too many posture problems when the dance reaches the current progress. At this time, there is little point in continuing to practice. It is necessary to practice the previous part repeatedly. By sending a repeat practice confirmation request, students are allowed to choose whether to repeat the practice according to the actual situation, which improves the flexibility and effectiveness of teaching. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 This is a flow chart of a dance posture and action recognition, interaction method and system embodiment based on visual computing of the present application;

[0064] Figure 2This is a flowchart of step S10 in a dance posture and action recognition, interaction method and system embodiment based on visual computing of the present application;

[0065] Figure 3 This is a flowchart of step S20 in a dance posture and action recognition, interaction method and system embodiment based on visual computing of the present application;

[0066] Figure 4 This is a schematic diagram of a dance posture and action recognition and interaction system based on visual computing in this application;

[0067] Figure 5 This is a principle block diagram of a computer device of the present application. DETAILED DESCRIPTION

[0068] The present application is further described in detail below in conjunction with the accompanying drawings.

[0069] In the following embodiments, Figure 1-3 As shown, the present application discloses a dance posture and action recognition and interaction method based on visual computing, which specifically includes the following steps:

[0070] S10: When receiving the recognition start instruction, obtaining the student's dance movement information and the rhythm information of the dance music through the camera device;

[0071] S20: extracting posture profile information of the student at a number of preset moments from the dance movement information based on the rhythm information of the dance music;

[0072] S30: comparing the extracted posture profile information with the pre-stored standard posture profile information, and outputting the comparison result;

[0073] S40: Based on the comparison result, the posture contour information whose overlap with the standard posture contour information is lower than a preset threshold is screened out, and the standard posture contour information corresponding to the posture contour information is extracted to generate a motion correction data packet;

[0074] S50: The interactive terminal reads the correction data packet in real time.

[0075] In this embodiment, the preset moment is usually the key frame of the dance movement, which is particularly important for the accuracy and style of the movement. The standard posture contour information is usually recorded by professional dance teachers to ensure its accuracy and standardization. The movement correction data package includes videos, pictures, and text descriptions of standard movements.

[0076] Specifically, the students' dance movements are captured by a camera, while the audio equipment collects the rhythm information of the dance music. The image processing technology is used to extract the contours of the students' dance movements at a preset moment from the video sequence. The contour information of the movement is obtained through edge detection, skeleton extraction and other methods. The extracted movement contour is compared with the pre-stored standard dance movement contour. The students' dance posture is evaluated by calculating the overlap ratio of the overall contour and the overlap of some important contours. The higher the overlap ratio, the closer the students' movements are to the standard movements. Real-time feedback and correction prompts are provided to students through interactive terminals (such as screens, tablets, etc.). The correction prompts can be in the form of visual marks, text descriptions or audio prompts.

[0077] In one embodiment, step S10 specifically includes the steps of:

[0078] S11: When receiving the recognition start instruction, sending the dance music screening list to the interactive terminal;

[0079] S12: extracting corresponding pre-stored dance music recordings and corresponding standard dance movement information for the dance music selected by the user;

[0080] S13: using an audio device to play a dance music recording and collect rhythm information of the dance music, where the rhythm information includes beats in the dance music and a timestamp of each beat;

[0081] S14: Acquire the student's dance movement information through a camera device, where the dance movement information includes all dance movement frames of the student.

[0082] In this embodiment, the dance music filter table includes various dance music for selection and related information, and the user can select appropriate music according to his or her needs. The timestamp is used to mark the specific time point when each beat occurs during the music playing process.

[0083] Specifically, when the system receives the recognition start instruction, it will send a dance music filter list to the interactive terminal (such as a tablet computer, smart phone, etc.). After the user selects specific dance music on the interactive terminal, the system will extract the corresponding pre-stored dance music recording and the corresponding standard dance movement information according to the user's selection. These standard movement information are usually dance movements performed by professional dancers, which have been recorded in advance and stored in the system as a reference standard for dance movements. In the dance movements captured by the video, the dance movement frames corresponding to each beat are marked according to the timestamp obtained by analyzing the music rhythm. In this way, each dance movement is associated with a specific beat in the music. While the music is playing, the camera device starts to capture the student's dance movements, including all the student's dance movement frames, that is, continuous dance movement pictures or video frames. These movement frames will be used to extract the student's posture contour information at a specific moment for comparison in subsequent steps.

[0084] In one embodiment, step S20 specifically includes the steps of:

[0085] S21: based on the rhythm information of the dance music and the standard dance movement information, marking the timestamp of the beat corresponding to each standard movement frame;

[0086] S22: according to a plurality of key postures in the preset dance movements, obtaining standard action frames corresponding to the plurality of key postures, and outputting the timestamps of the corresponding beats;

[0087] S23: extracting the dance action frame corresponding to the student at the timestamp from the dance action information based on the output beat timestamp;

[0088] S24: Outputting posture contour information based on the extracted dance action frames, and outputting standard posture contour information based on the standard action frames.

[0089] In this embodiment, the key posture is usually a representative movement in the dance, which is crucial to evaluating the overall performance of the dance. The posture contour information can be the key point positions and body curves of the dancer's body.

[0090] Specifically, according to the rhythm information of the dance music and the pre-stored standard dance movement information, a corresponding beat timestamp is marked for each standard action frame, so that each standard action frame is associated with a specific beat in the music, thereby ensuring the synchronization of the dance movement with the music rhythm. According to several key postures in the preset dance movements, the standard action frames corresponding to these key postures are screened out from the standard dance movement information, and the beat timestamps corresponding to these key postures are output. The beat timestamps output in step S22 are used to extract the dance movement frames corresponding to these timestamps from the student dance movement information. These extracted action frames will be used to compare with the standard action frames to evaluate the student's dance performance. Based on the extracted student dance movement frames, the student's posture contour information is output, and the corresponding standard posture contour information is output based on the standard action frames.

[0091] In one embodiment, step S30 specifically includes the following steps:

[0092] S31: Preset a personalized body proportion model for the student and store it in the interactive terminal;

[0093] S32: extracting a body proportion model of a corresponding student according to the user identity information sent by the interactive terminal;

[0094] S33: dynamically adjusting the standard posture contour information based on the body proportion model, and correcting the standard posture contour information by automatic scaling and deformation;

[0095] S34: Compare the extracted posture profile information with the corrected standard posture profile information, and output the coincidence data of the two.

[0096] In this embodiment, the student's body size data is acquired in advance through three-dimensional scanning or manual input.

[0097] Specifically, personalized body proportion models are preset for students to improve the accuracy of posture comparison. When posture comparison is required, the interactive terminal will send out the user's identity information, and extract the personalized body proportion model of the corresponding student based on this information, so as to use a model that matches the student's body shape to compare the posture contour. Based on this model, the standard posture contour information is dynamically adjusted, which includes automatically scaling and deforming the standard posture contour information to match the student's physical characteristics. The extracted student posture contour information is compared with the corrected standard posture contour information. The result of the comparison is to output the coincidence data of the two. These data reflect the similarity between the student's dance movements and the standard movements. The coincidence data can be used to evaluate the student's dance performance and guide subsequent movement correction.

[0098] In one embodiment, step S33 specifically includes the steps of:

[0099] S331: Identify key body parts in a scale model of the student body, including the head, neck, shoulders, elbows, wrists, hips, knees, and ankles;

[0100] S332: extracting the three-dimensional coordinates of each key body part, calculating the distance ratio between each three-dimensional coordinate, and outputting a ratio correction parameter;

[0101] S333: Correcting the standard posture contour information by automatic scaling and deformation according to the proportion correction parameter, and outputting the corrected standard posture contour information.

[0102] In this embodiment, the standard posture is adjusted according to the student's body proportion model to make action recognition more in line with individual characteristics. Through automatic scaling and deformation, the accuracy of matching the standard posture contour information with the actual action is improved. This can adapt to students with different physical characteristics, improve the universal applicability of the action recognition system, and reduce the possibility of misjudgment due to the different body proportions of the students and the standard posture contour.

[0103] Specifically, key body parts in the student's body proportion model are identified, which usually include the head, neck, shoulders, elbows, wrists, hips, knees and ankles, etc. The three-dimensional coordinates of the above key body parts are extracted to represent the position of each key part in the three-dimensional space, and the ratios of the distances between these three-dimensional coordinates are calculated. These ratios reflect the relative proportions between the various parts of the student's body. Based on these ratios, proportion correction parameters are output. The proportion correction parameters will be used to adjust the standard posture contour information to adapt to the student's physical characteristics. According to the calculated proportion correction parameters, the standard posture contour information is corrected by automatic scaling and deformation. This process ensures that the standard posture contour information matches the student's actual body proportions. The corrected standard posture contour information will more accurately reflect the student's proper dance posture, thereby improving the accuracy of posture comparison and movement correction.

[0104] In one embodiment, step S40 specifically includes the steps of:

[0105] S41: when the coincidence data between the posture profile information and the corrected standard posture profile information is lower than a preset threshold, the posture profile information is associated with a correction mark;

[0106] S42: extracting corresponding standard posture contour information based on the posture contour information with the correction mark;

[0107] S43: Analyze the difference between the posture profile information and the standard posture profile information, and output correction prompt information, including deviations in position, angle, movement range, etc.;

[0108] S44: Pack the correction prompt information to generate a correction data packet.

[0109] In this embodiment, the correction prompt information gives specific suggestions for improvement, such as "the knees should be bent more" or "the arms should be stretched more when raised". When the students are dancing, the system monitors in real time and gives instant feedback to help students adjust their movements immediately. After the students finish the dance, a complete feedback report is provided to allow students to review and analyze their performance. A practice mode is provided to allow students to repeatedly practice specific movements based on feedback until they reach the standard.

[0110] Specifically, the posture contour information is compared with the corrected standard posture contour information. When the overlap data between them is lower than a preset threshold, the posture contour information is marked with a correction mark. Based on the posture contour information with the correction mark, the corresponding standard posture contour information is extracted. The differences between the posture contour information with the correction mark and the corresponding standard posture contour information are analyzed. These differences may include deviations in position, angle, movement amplitude, etc. Specific correction prompt information is output based on these differences. These prompt information will guide students on how to adjust their dance movements to meet the standards. All the correction prompt information is packaged to generate a correction data packet. This data packet contains all the information that students need to correct in specific dance movements, which can be provided to teachers and students in real time through interactive terminals for immediate movement correction and training.

[0111] In one embodiment, after step S41, the following steps are further included:

[0112] S411: Recording the speed of generating the correction mark in real time during the playing of the dance music;

[0113] S412: When the speed of generating the correction mark is greater than a preset threshold, the music is automatically stopped and a repeat practice confirmation request is sent to the interactive terminal;

[0114] S413: The user selects whether to restart the dance movement practice based on the repeated practice confirmation request;

[0115] S414: After receiving the repeat practice instruction from the interactive terminal, all correction marks are cleared and the music is played from the beginning.

[0116] In this embodiment, the speed of generating correction marks is recorded in real time, and the teaching process can be dynamically adjusted. When the speed of generating correction marks is greater than a preset threshold, it means that the student has too many posture problems when the dance reaches the current progress. At this time, there is little point in continuing to practice, and it is necessary to practice the previous part repeatedly. By sending a repeated practice confirmation request, students are allowed to choose whether to repeat the practice according to actual conditions, thereby improving the flexibility and effectiveness of teaching.

[0117] Specifically, when the overlap data between the posture contour information and the corrected standard posture contour information is lower than a preset threshold, the posture contour information is associated with a correction mark. During the playing of dance music, the speed of generating the correction mark is recorded in real time. This speed indicator reflects the frequency of incorrect movements of students during the dance. If it is detected that the speed of generating the correction mark is greater than the preset threshold, this means that the student has made multiple incorrect movements in a short period of time. At this time, the music playback is automatically stopped and a repeat practice confirmation request is sent to the interactive terminal, prompting the student that he or she may need to practice again. The user (student or teacher) will choose whether to restart the dance movement practice based on the repeat practice confirmation request received by the interactive terminal. When the repeat practice instruction issued by the interactive terminal is received, all previous correction marks will be cleared and the music will be played from the beginning. These steps together constitute a closed-loop feedback and correction system that can monitor students' performance in real time and provide instant feedback and practice guidance, thereby optimizing the dance learning process.

[0118] In addition, when the interactive terminal receives a request to confirm repeated practice, the user can also choose to re-read the movement correction data packet according to needs, perform intensive practice on non-standard movements, or choose to open the dance music filter list to switch to other dances, or choose to continue the current dance practice.

[0119] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0120] In one embodiment, a dance posture and action recognition and interaction system based on visual computing is provided, and the device corresponds to a dance posture and action recognition and interaction method based on visual computing in the above embodiment. Figure 4 As shown, the system includes:

[0121] The recognition module is used to obtain the student's dance movement information and the rhythm information of the dance music through the camera device when receiving the recognition start instruction;

[0122] An extraction module is used to extract the posture contour information of the student at a number of preset moments from the dance movement information based on the rhythm information of the dance music;

[0123] A comparison module, used to compare the extracted posture profile information with the pre-stored standard posture profile information, and output the comparison result;

[0124] The feedback module is used to screen out posture profile information whose overlap with the standard posture profile information is lower than a preset threshold based on the comparison result, extract the standard posture profile information corresponding to the posture profile information, generate a motion correction data packet, and send the motion correction data packet to the interactive terminal.

[0125] For the specific definition of a dance posture and action recognition and interaction system based on visual computing, please refer to the definition of a dance posture and action recognition and interaction method based on visual computing in the above text, which will not be repeated here. Each module in the above-mentioned dance posture and action recognition and interaction system based on visual computing can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory in the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0126] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 5 As shown. The computer device includes a processor, a memory, a network interface and a database connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a dance posture and action recognition and interaction method based on visual computing is implemented.

[0127] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, a dance posture and action recognition and interaction method based on visual computing is implemented;

[0128] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, a dance posture and action recognition and interaction method based on visual computing is implemented.

[0129] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0130] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by several functional units and modules as needed, that is, the internal structure of the device can be divided into several functional units or modules to complete all or part of the functions described above.

[0131] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A dance posture recognition and interaction method based on visual computing, characterized in that: Includes steps: When receiving the recognition start instruction, the student's dance movement information and the rhythm information of the dance music are obtained through the camera device; Based on the rhythm information of dance music, the posture contour information of students at several preset moments is extracted from the dance movement information; Compare the extracted posture profile information with the pre-stored standard posture profile information, and output the comparison result; Based on the comparison results, the posture contour information whose overlap with the standard posture contour information is lower than a preset threshold is screened out, and the standard posture contour information corresponding to the posture contour information is extracted to generate a motion correction data packet; The interactive terminal reads the correction data packet in real time.

2. The dance posture and action recognition and interaction method based on visual computing according to claim 1 is characterized in that: The step of obtaining the student's dance movement information and the rhythm information of the dance music through the camera device when the recognition start instruction is received specifically includes the steps of: When receiving the recognition start instruction, sending the dance music screening list to the interactive terminal; For the dance music selected by the user, extract the corresponding pre-stored dance music recording and the corresponding standard dance movement information; Using an audio device to play a dance music recording and collect rhythm information of the dance music, the rhythm information including beats in the dance music and a timestamp of each beat; The student's dance movement information is obtained through a camera device, and the dance movement information includes all dance movement frames of the student.

3. The dance posture and action recognition and interaction method based on visual computing according to claim 1 is characterized in that: The step of extracting the posture contour information of the student at a plurality of preset moments from the dance movement information based on the rhythm information of the dance music specifically comprises the steps of: Based on the rhythm information of the dance music and the standard dance movement information, mark the timestamp of the beat corresponding to each standard movement frame; According to several key postures in the preset dance movements, standard action frames corresponding to the several key postures are obtained, and the timestamps of the corresponding beats are output; Based on the output beat timestamp, the dance action frames corresponding to the students at the timestamp are extracted from the dance action information; The posture contour information is output based on the extracted dance action frame, and the standard posture contour information is output based on the standard action frame.

4. The dance posture and action recognition and interaction method based on visual computing according to claim 1 is characterized in that: The step of comparing the extracted plurality of posture profile information with the pre-stored standard posture profile information and outputting the comparison result specifically comprises the steps of: Preset personalized body proportion models for students and store them in the interactive terminal; Extracting the body proportion model of the corresponding student according to the user identity information sent by the interactive terminal; Dynamically adjust the standard posture contour information based on the body proportion model, and correct the standard posture contour information by automatic scaling and deformation; The extracted posture profile information is compared with the corrected standard posture profile information, and the coincidence data of the two is output.

5. The dance posture and action recognition and interaction method based on visual computing according to claim 4 is characterized in that: The step of dynamically adjusting the standard posture contour information based on the body proportion model and correcting the standard posture contour information by automatic scaling and deformation specifically includes the following steps: Identify key body parts in a scale model of the student's body, including the head, neck, shoulders, elbows, wrists, hips, knees, and ankles; Extract the three-dimensional coordinates of each key body part, calculate the distance ratio between each three-dimensional coordinate, and output the proportion correction parameter; According to the proportion correction parameter, the standard posture contour information is corrected by automatic scaling and deformation, and the corrected standard posture contour information is output.

6. The dance posture and action recognition and interaction method based on visual computing according to claim 1 is characterized in that: The step of screening out posture profile information whose overlap with the standard posture profile information is lower than a preset threshold based on the comparison result, extracting the standard posture profile information corresponding to the posture profile information, and generating a motion correction data packet specifically includes the following steps: When the coincidence data between the posture profile information and the corrected standard posture profile information is lower than a preset threshold, the posture profile information is associated with a correction mark; Based on the posture contour information with the correction mark, extract the corresponding standard posture contour information; Analyze the difference between posture profile information and standard posture profile information, and output correction prompt information, including deviations in position, angle, movement range, etc.; Pack the correction prompt information to generate a correction data packet.

7. The dance posture and action recognition and interaction method based on visual computing according to claim 6 is characterized in that: After the step of associating the posture profile information with the correction mark when the coincidence data between the posture profile information and the corrected standard posture profile information is lower than a preset threshold, the step further includes: The speed of generating the correction mark is recorded in real time during the playing of the dance music; When the speed of generating the correction mark is greater than the preset threshold, the music playback is automatically stopped and a repeat practice confirmation request is sent to the interactive terminal; The user selects whether to restart the dance movement practice based on the repeat practice confirmation request; After receiving the repeat practice instruction from the interactive terminal, all correction marks are cleared and the music is played from the beginning.

8. A dance posture recognition and interaction system based on visual computing, characterized in that: include: The recognition module is used to obtain the student's dance movement information and the rhythm information of the dance music through the camera device when receiving the recognition start instruction; An extraction module is used to extract the posture contour information of the student at a number of preset moments from the dance movement information based on the rhythm information of the dance music; A comparison module, used to compare the extracted posture profile information with the pre-stored standard posture profile information, and output the comparison result; The feedback module is used to screen out posture profile information whose overlap with the standard posture profile information is lower than a preset threshold based on the comparison result, extract the standard posture profile information corresponding to the posture profile information, generate a motion correction data packet, and send the motion correction data packet to the interactive terminal.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the dance posture and action recognition and interaction method based on visual computing as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of a dance posture and action recognition and interaction method based on visual computing as described in any one of claims 1 to 7 are implemented.

Images