Teaching auxiliary training system based on visual motion capture

Abstract

The invention relates to the field of motion-aided training, in particular to a teaching-aided training system based on visual motion capture, and the system comprises a motion collection module which uses a visual sensor to capture a motion picture video of a dynamic target; the preprocessing module is used for separating a video only containing a dynamic target, processing a video frame, obtaining an action picture, performing exposure processing, and obtaining the processed action picture; the action analysis module is used for extracting key part actions from the processed action picture and comparing the key part actions with standard actions to determine non-standard actions so as to determine action deviation and provide a calibration method; the teaching guidance module is used for carrying out voice prompt on a calibration method, capturing an adjustment action, reducing the difficulty level and guiding a dynamic target to execute a simplified action if a standard action is continuously not reached; and the difficulty recovery module monitors the action of the dynamic target in executing the simplified action, and if the standard action is met, the difficulty level of the action is recovered in the next execution period. Therefore, a personalized learning path is provided.

Description

Technical Field

[0001] The present invention relates to the field of auxiliary training, and in particular to a teaching auxiliary training system based on visual motion capture. Background Art

[0002] With the continuous advancement and innovation of technology, the application prospects of motion capture technology in teaching and training will be broader, which will not only improve the quality of education, promote scientific research cooperation, but also push education and scientific research into a new era.

[0003] Chinese patent publication number CN117519487A discloses a tunnel boring machine control teaching auxiliary training system based on visual motion capture, including a camera unit, a display unit and a processing unit; the camera unit is used to capture the user's hand and eye gestures in real time; the display unit is used to display the simulated tunnel boring machine operation screen and the simulated environment feedback effect in real time; the processing unit includes a gesture recognition processing unit and a simulated tunnel boring machine operating system; wherein, the gesture recognition processing unit, on the basis of solving the problem of video highlights and user gesture covering, determines the user's gesture command and attention concentration by detecting the key node spatial data of the user's hands and eyes; the simulated tunnel boring machine operating system is used to receive commands to realize control, and transmit the operation and environment feedback screen to the display unit.

[0004] In the prior art, the user's gesture commands and attention concentration are determined by detecting the key node spatial data of the user's hands and eyes, and there is no adjustment or correction process. Summary of the invention

[0005] To this end, the present invention provides a teaching auxiliary training system based on visual motion capture, which can solve the problem of non-standard execution of dynamic targets.

[0006] To achieve the above object, the present invention provides a teaching auxiliary training system based on visual motion capture, comprising:

[0007] The motion acquisition module is used to capture the motion video of dynamic targets using visual sensors;

[0008] A preprocessing module is connected to the action acquisition module, processes the acquired action picture video, separates the video containing only dynamic targets, performs frame processing on the video, obtains a number of action pictures, performs exposure processing on the action pictures, and obtains the processed action pictures;

[0009] An action analysis module, connected to the preprocessing module, extracts key part actions based on the processed action picture, compares the key part actions with standard actions, determines non-standard actions according to the comparison results, determines action deviations based on the non-standard actions, and provides a calibration method for the action deviations;

[0010] a teaching guidance module, connected to the action analysis module, providing voice prompts based on the calibration method, capturing the adjustment actions performed by the dynamic target under the voice prompts, and reducing the difficulty level of the action if the standard action is not achieved for several consecutive times, and guiding the dynamic target to perform a simplified action, wherein the simplified action is the action of the standard action with the difficulty level reduced;

[0011] The difficulty recovery module is connected to the teaching guidance module, and monitors a number of attempted actions of the dynamic target during the execution of the simplified action. If any of the attempted actions meets the standard action, the action difficulty level is restored when the action prompt is performed in the next execution cycle.

[0012] Furthermore, the action acquisition module includes:

[0013] A sensor configuration unit, used to configure the required visual sensors to ensure that the entire action screen area can be covered;

[0014] An action acquisition unit, connected to the sensor configuration unit, for acquiring action picture videos according to the visual sensor;

[0015] The data storage unit is connected to the action acquisition unit and stores the acquired action picture video in a storage device.

[0016] Furthermore, the preprocessing module includes:

[0017] A background segmentation unit, which performs background segmentation according to the action picture video to separate the video containing only dynamic objects;

[0018] The exposure processing unit is connected to the background segmentation unit, performs frame processing on the video containing only dynamic objects, divides it into a plurality of continuous action pictures and performs exposure processing to obtain the action pictures after exposure processing.

[0019] Furthermore, the exposure processing unit includes:

[0020] A frame capture subunit, which captures the video data containing only dynamic objects frame by frame and divides it into a plurality of coherent action pictures;

[0021] An automatic exposure detection subunit, connected to the frame capture subunit, performs analysis based on the action picture, obtains the brightness and contrast of each frame, and determines whether exposure adjustment is required;

[0022] An exposure adjustment subunit is connected to the automatic exposure detection subunit, and for the action picture that needs exposure adjustment, automatically calculates the required exposure compensation value by using frame brightness and contrast, and adjusts the brightness of the action picture;

[0023] The local exposure optimization subunit is connected to the exposure adjustment subunit and performs local exposure optimization on key parts of the dynamic target in the action picture.

[0024] Furthermore, the action analysis module includes:

[0025] A feature extraction unit, used to extract key part movements of the dynamic target using motion capture technology;

[0026] A feature matching unit is connected to the feature extraction unit and uses a feature matching algorithm to match and compare the key part movements with standard key part movements to obtain non-standard movements;

[0027] The deviation analysis unit is connected to the feature matching unit, performs analysis based on the non-standard action, determines the action deviation according to the analysis result, and provides a calibration method based on the action deviation.

[0028] Furthermore, the feature matching unit includes:

[0029] An image segmentation subunit is configured to segment the action picture and the corresponding standard action picture into a plurality of uniform pixel grids;

[0030] a marking subunit, connected to the image segmentation subunit, for continuously marking the key parts in the action picture and the key parts in the standard action picture;

[0031] A pixel difference subunit is connected to the marking unit, determines the corresponding coordinate point based on the marking, and calculates the difference pixels from the key parts in the standard action picture using the feature matching algorithm;

[0032] The conversion subunit is connected to the pixel difference subunit, and performs mapping processing based on the difference pixels to obtain the non-standard action.

[0033] Furthermore, the deviation analysis unit includes:

[0034] a deviation determination subunit, which determines the rotation angle deviation based on the comparison result based on the comparison between the non-standard action and the standard action;

[0035] An improvement subunit, connected to the deviation determination subunit, provides a calibration method based on the rotation angle deviation with reference to a standard rotation angle.

[0036] Furthermore, the teaching guidance module includes:

[0037] A voice prompt unit, based on the calibration method, generates voice prompts using speech synthesis technology;

[0038] A picture capturing unit, connected to the voice prompting unit, for capturing a picture of the adjustment action performed by the dynamic target under the voice prompting;

[0039] A threshold setting unit connected to the picture capturing unit, for setting a continuous non-standard threshold based on the non-standard action, and reducing the action difficulty level when the adjustment action reaches the threshold;

[0040] The action simplification unit is connected to the threshold setting unit and uses voice to guide the dynamic target to perform the action of reducing the action difficulty level.

[0041] Furthermore, the action simplification unit includes:

[0042] A recording subunit, used for recording the adjustment action screen performed under the voice prompt;

[0043] a level analysis subunit, connected to the level analysis subunit, analyzing the adjustment action picture to be executed based on the dynamic target, and determining the level of the adjustment action executed by the dynamic target according to the analysis result;

[0044] The action simplification selection subunit is connected to the level analysis subunit, and is configured to adjust the level of the action based on the dynamic target execution, and select a simplified action matching the level from the action library.

[0045] Furthermore, the difficulty recovery module includes:

[0046] An action monitoring unit, used for real-time monitoring of several attempted actions of the dynamic target during the execution of the simplified action;

[0047] A standard action verification unit, connected to the action monitoring unit, compares the several attempted actions with the standard action, and determines whether any of the attempted actions satisfies the standard action according to the comparison result;

[0048] The level recovery unit is connected to the standard action verification unit, and based on the fact that any attempted action satisfies the standard action, the action difficulty level is restored when the action prompt is performed in the next execution cycle.

[0049] Compared with the prior art, the beneficial effects of the present invention are that, through the action acquisition module, the action picture video of the dynamic target is captured by using a visual sensor, which provides materials for subsequent analysis of the action; the collected action picture video is processed by using a preprocessing module to separate the video containing only the dynamic target, the video is frame processed, a number of action pictures are obtained and exposed, and the processed action pictures are obtained, thereby improving the picture quality so that every detail of the action can be clearly identified; the key part action of the processed action picture is extracted by the action analysis module and compared with the standard action, so that non-standard actions can be accurately identified, and the non-standard action is determined according to the comparison result. action, thereby determining the action deviation and providing a calibration method for the action deviation to improve the training efficiency; voice prompts are given according to the teaching guidance module to capture the adjustment actions performed by the dynamic target under the voice prompts. If the standard action is not achieved for several consecutive times, the action difficulty level is reduced to guide the dynamic target to perform simplified actions, provide personalized calibration methods and instant feedback, and monitor several attempted actions of the dynamic target during the execution of the simplified action through the difficulty recovery module. If any of the attempted actions meets the standard action, the action difficulty level is restored during the next execution cycle when the action prompt is performed, which is helpful to consolidate and deepen the skills, thereby optimizing the learning curve.

[0050] In particular, by configuring sensors to capture dynamic target movements and save them in a storage device for subsequent processing and analysis, materials are provided for subsequent analysis of movements.

[0051] In particular, the background segmentation unit is used to separate the dynamic target from the video to remove background interference, and the separated dynamic target is exposed to reduce the impact of highlights on the feature point matching algorithm to ensure the clarity of the action. The feature point matching algorithm is used to accurately identify and track the key body parts of the dynamic target to provide accurate motion data.

[0052] In particular, by acquiring the video data frame by frame and analyzing the brightness and contrast, a compensation value is calculated based on the analysis results, thereby adjusting the brightness value to optimize the visibility of the dynamic target, and focusing on optimizing the body joints to ensure that the dynamic target movements are well displayed in detail in both highlight and shadow areas.

[0053] In particular, by extracting the key part movements of the dynamic target and matching and comparing them with the standard key part movements, non-standard movements are obtained, thereby accurately imitating standard movements and improving movement skills.

[0054] In particular, the action picture and the corresponding standard action picture are divided into several uniform pixel grids by the image segmentation unit and the key parts of the dynamic target in the picture are continuously marked by the marking unit, thereby providing an accurate reference picture, which helps to more accurately capture and record the key parts and actions of the dynamic target, and then the pixel difference unit is used to calculate the difference pixels according to the feature matching algorithm to find out the non-standard action, so as to refine the correction of the non-standard action.

[0055] In particular, by comparing the non-standard actions with standard actions, specific deviations in the actions are accurately identified, and improvement methods are provided, so that the technology of dynamic targets can be continuously improved.

[0056] In particular, by identifying the deviation between the non-standard action recognition and the standard action, an improvement method is obtained through analysis, and a voice prompt is generated. By continuously correcting the deviation, the accuracy of the action is improved, and by setting a threshold for the number of consecutive non-standard actions, the difficulty level of the action is reduced, thereby providing a personalized learning path.

[0057] In particular, the level of the adjustment action performed by the dynamic target is determined by analyzing the adjustment action picture performed under the voice prompt and a simplified action matching the level is selected from the action library to improve the execution efficiency and accuracy.

[0058] In particular, by real-time monitoring of the dynamic target's action execution and verifying whether the dynamic target has accurately completed the standard action, the level of difficulty of the standard action will be restored and the record will be eliminated, helping the dynamic target to adapt to different exercise difficulties and optimize the learning curve. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 A schematic diagram of the structure of a teaching auxiliary training system based on visual motion capture provided by an embodiment of the present invention;

[0060] Figure 2 A schematic diagram of the structure of a motion acquisition module of a teaching auxiliary training system based on visual motion capture provided by an embodiment of the present invention;

[0061] Figure 3 A schematic diagram of the structure of a preprocessing module of a teaching auxiliary training system based on visual motion capture provided by an embodiment of the present invention;

[0062] Figure 4 A schematic diagram of the structure of a motion analysis module of a teaching auxiliary training system based on visual motion capture provided by an embodiment of the present invention;

[0063] Figure 5 A schematic diagram of the structure of a teaching guidance module of a teaching auxiliary training system based on visual motion capture provided by an embodiment of the present invention;

[0064] Figure 6 A schematic diagram of the difficulty recovery module structure of the visual motion capture-based teaching auxiliary training system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0065] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0066] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0067] It should be noted that, in the description of the present invention, terms such as "up", "down", "left", "right", "inside" and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the drawings. This is merely for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it should not be understood as a limitation on the present invention.

[0068] In addition, it should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected" and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, or it can be the internal communication of two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0069] See also Figure 1 As shown, this embodiment is a teaching auxiliary training system based on visual motion capture, including:

[0070] The motion acquisition module 10 is used to capture the motion picture video of the dynamic target using a visual sensor;

[0071] The preprocessing module 20 is connected to the action acquisition module, processes the acquired action picture video, separates the video containing only dynamic targets, performs frame processing on the video, obtains a number of action pictures, performs exposure processing on the action pictures, and obtains the processed action pictures;

[0072] The motion analysis module 30 is connected to the preprocessing module, extracts key part motions based on the processed motion picture, compares the key part motions with standard motions, determines non-standard motions according to the comparison results, determines motion deviations based on the non-standard motions, and provides a calibration method for the motion deviations;

[0073] The teaching guidance module 40 is connected to the action analysis module, and performs voice prompts based on the calibration method to capture the adjustment actions performed by the dynamic target under the voice prompts. If the standard action is not achieved for several consecutive times, the difficulty level of the action is reduced, and the dynamic target is guided to perform a simplified action, where the simplified action is the action of the standard action with the difficulty level reduced;

[0074] The difficulty recovery module 50 is connected to the teaching guidance module, and monitors a number of attempted actions of the dynamic target during the execution of the simplified action. If any of the attempted actions meets the standard action, the action difficulty level is restored when the action prompt is performed in the next execution cycle.

[0075] Specifically, the present embodiment provides materials for subsequent analysis of actions by using a motion acquisition module to capture action picture videos of dynamic targets using visual sensors; using a preprocessing module, the collected action picture videos are processed to separate videos containing only dynamic targets, the videos are frame processed, a number of action pictures are acquired and exposed, and the processed action pictures are acquired, thereby improving the picture quality so that every detail of the action can be clearly identified; the action analysis module extracts key part actions from the processed action pictures and compares them with standard actions, so that non-standard actions can be accurately identified, and non-standard actions are determined based on the comparison results, thereby determining action deviations and providing a calibration method for the action deviations to improve training efficiency; voice prompts are given according to the teaching guidance module to capture the adjustment actions performed by the dynamic target under the voice prompts, and if the standard actions are not achieved for several consecutive times, the action difficulty level is reduced, the dynamic target is guided to perform simplified actions, a personalized calibration method is provided and instant feedback is provided, and a difficulty recovery module is used to monitor several attempted actions of the dynamic target during the execution of the simplified actions, and if any of the attempted actions meets the standard actions, the action difficulty level is restored during the next execution cycle when the action prompts are performed, which helps to consolidate and deepen the skills, thereby optimizing the learning curve.

[0076] See also Figure 2 As shown, the action acquisition module 10 includes:

[0077] The sensor configuration unit 11 is used to configure the required visual sensors to ensure that the entire action picture area can be covered;

[0078] An action acquisition unit 12, connected to the sensor configuration unit, for acquiring action picture videos according to the visual sensor;

[0079] The data storage unit 13 is connected to the action acquisition unit and stores the acquired action picture video in a storage device.

[0080] Specifically, this embodiment configures sensors to capture dynamic target actions and stores them in a storage device for subsequent processing and analysis, thereby providing materials for subsequent analysis of actions.

[0081] See also Figure 3 As shown, the preprocessing module 20 includes:

[0082] A background segmentation unit 21 performs background segmentation according to the action picture video to separate the video containing only dynamic objects;

[0083] The exposure processing unit 22 is connected to the background segmentation unit, performs frame processing on the video containing only dynamic objects, divides it into a plurality of continuous action pictures and performs exposure processing to obtain the action pictures after exposure processing;

[0084] Specifically, this embodiment separates the dynamic target from the video through the background segmentation unit to remove background interference, performs exposure processing on the separated dynamic target to reduce the impact of highlights on the feature point matching algorithm to ensure the clarity of the action, and uses the feature point matching algorithm to accurately identify and track key body parts of the dynamic target to provide accurate motion data.

[0085] Specifically, the exposure processing unit includes:

[0086] A frame capture subunit, which captures the video data containing only dynamic objects frame by frame and divides it into a plurality of coherent action pictures;

[0087] An automatic exposure detection subunit, connected to the frame capture subunit, performs analysis based on the action picture, obtains the brightness and contrast of each frame, and determines whether exposure adjustment is required;

[0088] An exposure adjustment subunit is connected to the automatic exposure detection subunit, and for the action picture that needs exposure adjustment, automatically calculates the required exposure compensation value by using frame brightness and contrast, and adjusts the brightness of the action picture;

[0089] The local exposure optimization subunit is connected to the exposure adjustment subunit and performs local exposure optimization on key parts of the dynamic target in the action picture.

[0090] Specifically, the embodiment of the present invention acquires the video data frame by frame and analyzes the brightness and contrast, calculates the compensation value according to the analysis result, and adjusts the brightness value to optimize the visibility of the dynamic target, focusing on optimizing the body joints to ensure that the dynamic target movements have good details in both highlight and shadow areas.

[0091] Specifically, the key parts of the dynamic target include: joints, head, shoulders, elbows, wrists, hips, knees and ankles.

[0092] See also Figure 4 As shown, the action analysis module 30 includes:

[0093] A feature extraction unit 31, used to extract key part movements of the dynamic target using motion capture technology;

[0094] A feature matching unit 32 is connected to the feature extraction unit and uses a feature matching algorithm to match and compare the key part movements with standard key part movements to obtain non-standard movements;

[0095] The deviation analysis unit 33 is connected to the feature matching unit, performs analysis based on the non-standard action, determines the action deviation according to the analysis result, and provides a calibration method based on the action deviation.

[0096] Specifically, by extracting the key part movements of the dynamic target and matching and comparing them with the standard key part movements, non-standard movements are obtained, thereby accurately imitating standard movements and improving movement skills.

[0097] Specifically, the feature matching unit includes:

[0098] An image segmentation subunit is configured to segment the action picture and the corresponding standard action picture into a plurality of uniform pixel grids;

[0099] a marking subunit, connected to the image segmentation subunit, for continuously marking the key parts in the action picture and the key parts in the standard action picture;

[0100] A pixel difference subunit is connected to the marking unit, determines the corresponding coordinate point based on the marking, and calculates the difference pixels from the key parts in the standard action picture using the feature matching algorithm;

[0101] a conversion subunit, connected to the pixel difference subunit, performing mapping processing based on the difference pixels to obtain the non-standard action;

[0102] Specifically, the embodiment of the present invention divides the action picture and the corresponding standard action picture into a number of uniform pixel grids through the image segmentation unit and uses the marking unit to continuously mark the key parts of the dynamic target in the picture, thereby providing an accurate reference picture, which helps to more accurately capture and record the key parts and actions of the dynamic target, and then uses the pixel difference unit to calculate the difference pixels according to the feature matching algorithm to find out the non-standard action, thereby refining the correction of the non-standard action.

[0103] Specifically, the set of key parts in the action picture that are continuously marked in the embodiment of the present invention is P={p 1 ,p 2 ,p 3 ......p n} Among them, p 1 ,p 2 ,p 3 .....p n is the marking point of the key part in the action picture, n>3; the set of continuous marking of the key parts in the standard action picture is P′={p′ 1 ,p′ 2 ,p′ 3 ......p′ n}where p′ 1 ,p′ 2 ,p′ 3 .....p′ n is the marking point of the key part in the standard action picture, n>3, each element in P and P' is identified, the difference element is extracted, the corresponding coordinate point is determined based on the difference element, and the square sum of pixel differences is used to calculate

[0104] SSD=[p i (x i ,y i )-p′ i (x′ i , y′ i )] 2

[0105] Among them, p i (x i ,y i ) is the coordinate point of a key part in the action picture, p′ i (x′ i ,y′ i ) are the coordinate points of the key parts in the corresponding standard action picture.

[0106] Specifically, the deviation analysis unit includes:

[0107] a deviation determination subunit, which determines the rotation angle deviation based on the comparison result based on the comparison between the non-standard action and the standard action;

[0108] The improvement subunit is connected to the deviation determination subunit and provides an improvement method based on the rotation angle deviation with reference to a standard rotation angle.

[0109] Specifically, the embodiment of the present invention compares the non-standard action with the standard action, accurately identifies the specific deviation in the action, and provides an improvement method, so that the technology of dynamic targets is continuously improved.

[0110] Specifically, the rotation angle deviation in the embodiment of the present invention is calculated as follows: 1 (x 1 ,y 1 ) 2 (x 2 ,y 2 )

[0111] ,,p 3 (x 3 ,y 3 ) is the key connection point of the non-standard action, and the vector For p 1 and p 2 Key parts connection points at the same level, vector For p 2 and p 3 The same level key parts connection points, through the inverse cosine function cos -1 Determine the rotation in radians Convert the radians to degrees According to the rotation angle θ′ in the standard action, calculate the deviation angle θ′-θ degress .

[0112] See also Figure 5 As shown, the teaching guidance module 40 includes:

[0113] A voice prompt unit 41 generates a voice prompt based on the calibration method and using speech synthesis technology;

[0114] A picture capturing unit 42, connected to the voice prompting unit, for capturing the picture of the adjustment action performed by the dynamic target under the voice prompting;

[0115] A threshold setting unit 43, connected to the picture capturing unit, sets a continuous non-standard threshold based on the non-standard action, and reduces the difficulty level of the action when the adjustment action reaches the threshold;

[0116] The action simplification unit 44 is connected to the threshold setting unit and uses voice to guide the dynamic target to perform the action of reducing the action difficulty level.

[0117] Specifically, the embodiment of the present invention obtains an improvement method by identifying the deviation between the non-standard action recognition and the standard action, and generates a voice prompt. By continuously correcting the deviation, the accuracy of the action is improved, and the difficulty level of the action is reduced by setting a threshold for the number of consecutive non-standard actions, thereby providing a personalized learning path.

[0118] Specifically, the embodiment of the present invention sets the threshold of consecutive non-compliance action targets to 3 times.

[0119] Specifically, the action simplification unit includes:

[0120] A recording subunit, used for recording the adjustment action screen performed under the voice prompt;

[0121] A level analysis subunit, connected to the recording subunit, analyzes the adjustment action screen executed based on the dynamic target, and determines the level of the adjustment action executed by the dynamic target according to the analysis result;

[0122] The action simplification selection subunit is connected to the level analysis subunit, and is configured to adjust the level of the action based on the dynamic target execution, and select a simplified action matching the level from the action library.

[0123] Specifically, the embodiment of the present invention analyzes the adjustment action screen performed under the voice prompt to determine the level of the dynamic target's execution of the adjustment action and selects a simplified action matching the level in the action library to improve execution efficiency and accuracy.

[0124] See also Figure 6 As shown, the difficulty recovery module 50 includes:

[0125] An action monitoring unit 51 is used to monitor in real time a number of attempted actions of the dynamic target during the execution of the simplified action;

[0126] A standard action verification unit 52, connected to the action monitoring unit, compares the several attempted actions with the standard action, and determines whether any of the attempted actions satisfies the standard action according to the comparison result;

[0127] The level recovery unit 53 is connected to the standard action verification unit, and based on whether any attempted action satisfies the standard action, the action difficulty level is recovered when the action prompt is performed in the next execution cycle.

[0128] Specifically, the embodiment of the present invention monitors the action execution of the dynamic target in real time and verifies whether the dynamic target has accurately completed the standard action, thereby restoring the level difficulty of the standard action and eliminating the record, helping the dynamic target to adapt to different practice difficulties and optimizing the learning curve.

[0129] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

[0130] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A teaching-assisted training system based on visual motion capture, It is characterized in that include: The motion acquisition module is used to capture the motion video of dynamic targets using visual sensors; A preprocessing module is connected to the action acquisition module, processes the acquired action picture video, separates the video containing only dynamic targets, performs frame processing on the video, obtains a number of action pictures, performs exposure processing on the action pictures, and obtains the processed action pictures; An action analysis module, connected to the preprocessing module, extracts key part actions based on the processed action picture, compares the key part actions with standard actions, determines non-standard actions according to the comparison results, determines action deviations based on the non-standard actions, and provides a calibration method for the action deviations; a teaching guidance module, connected to the action analysis module, providing voice prompts based on the calibration method, capturing the adjustment actions performed by the dynamic target under the voice prompts, and reducing the difficulty level of the action if the standard action is not achieved for several consecutive times, and guiding the dynamic target to perform a simplified action, wherein the simplified action is the action of the standard action with the difficulty level reduced; The difficulty recovery module is connected to the teaching guidance module, and monitors a number of attempted actions of the dynamic target during the execution of the simplified action. If any of the attempted actions meets the standard action, the action difficulty level is restored when the action prompt is performed in the next execution cycle.

2. The teaching-assisted training system based on visual motion capture according to claim 1, It is characterized in that The motion acquisition module comprises: A sensor configuration unit, used to configure the required visual sensors to ensure that the entire action screen area can be covered; An action acquisition unit, connected to the sensor configuration unit, for acquiring action picture videos according to the visual sensor; The data storage unit is connected to the action acquisition unit and stores the acquired action picture video in a storage device.

3. The teaching auxiliary training system based on visual motion capture according to claim 2, It is characterized in that The preprocessing module comprises: A background segmentation unit, which performs background segmentation according to the action picture video to separate the video containing only dynamic objects; The exposure processing unit is connected to the background segmentation unit, performs frame processing on the video containing only dynamic objects, divides it into a plurality of continuous action pictures and performs exposure processing to obtain the action pictures after exposure processing.

4. The teaching-assisted training system based on visual motion capture according to claim 3, It is characterized in that The exposure processing unit comprises: A frame capture subunit, which captures the video data containing only dynamic objects frame by frame and divides it into a plurality of coherent action pictures; An automatic exposure detection subunit, connected to the frame capture subunit, performs analysis based on the action picture, obtains the brightness and contrast of each frame, and determines whether exposure adjustment is required; An exposure adjustment subunit is connected to the automatic exposure detection subunit, and for the action picture that needs exposure adjustment, automatically calculates the required exposure compensation value by using frame brightness and contrast, and adjusts the brightness of the action picture; The local exposure optimization subunit is connected to the exposure adjustment subunit and performs local exposure optimization on key parts of the dynamic target in the action picture.

5. The teaching-assisted training system based on visual motion capture according to claim 4, It is characterized in that The action analysis module comprises: A feature extraction unit, used to extract key part movements of the dynamic target using motion capture technology; A feature matching unit is connected to the feature extraction unit and uses a feature matching algorithm to match and compare the key part movements with standard key part movements to obtain non-standard movements; The deviation analysis unit is connected to the feature matching unit, performs analysis based on the non-standard action, determines the action deviation according to the analysis result, and provides a calibration method based on the action deviation.

6. The teaching-assisted training system based on visual motion capture according to claim 5, It is characterized in that The feature matching unit comprises: An image segmentation subunit is configured to segment the action picture and the corresponding standard action picture into a plurality of uniform pixel grids; a marking subunit, connected to the image segmentation subunit, for continuously marking the key parts in the action picture and the key parts in the standard action picture; A pixel difference subunit is connected to the marking subunit, determines the corresponding coordinate point based on the marking, and calculates the difference pixels from the key parts in the standard action picture using the feature matching algorithm; The conversion subunit is connected to the pixel difference subunit, and performs mapping processing based on the difference pixels to obtain the non-standard action.

7. The teaching-assisted training system based on visual motion capture according to claim 6, It is characterized in that The deviation analysis unit comprises: a deviation determination subunit, which determines the rotation angle deviation based on the comparison result based on the comparison between the non-standard action and the standard action; The improved subunit provides a calibration method based on the rotation angle deviation with reference to a standard rotation angle.

8. The teaching-assisted training system based on visual motion capture according to claim 7, It is characterized in that The teaching guidance module includes: A voice prompt unit, based on the calibration method, generates voice prompts using speech synthesis technology; A picture capturing unit, connected to the voice prompting unit, for capturing a picture of the adjustment action performed by the dynamic target under the voice prompting; A threshold setting unit connected to the picture capturing unit, for setting a continuous non-standard threshold based on the non-standard action, and reducing the action difficulty level when the adjustment action reaches the threshold; The action simplification unit is connected to the threshold setting unit and uses voice to guide the dynamic target to perform the action of reducing the action difficulty level.

9. The teaching-assisted training system based on visual motion capture according to claim 8, It is characterized in that The action simplification unit comprises: A recording subunit, used for recording the adjustment action picture performed under the voice prompt; A level analysis subunit, connected to the recording subunit, analyzes the adjustment action screen executed based on the dynamic target, and determines the level of the adjustment action executed by the dynamic target according to the analysis result; The action simplification selection subunit is connected to the level analysis subunit, and is configured to adjust the level of the action based on the dynamic target execution, and select a simplified action matching the level from the action library.

10. The teaching-assisted training system based on visual motion capture according to claim 9, It is characterized in that The difficulty recovery module includes: An action monitoring unit, used for real-time monitoring of several attempted actions of the dynamic target during the execution of the simplified action; A standard action verification unit, connected to the action monitoring unit, compares the several attempted actions with the standard action, and determines whether any of the attempted actions satisfies the standard action according to the comparison result; The level recovery unit is connected to the standard action verification unit, and based on the fact that any attempted action satisfies the standard action, the action difficulty level is restored when the action prompt is performed in the next execution cycle.

Images