A leg muscle training analysis method based on image analysis and a training chair

Abstract

The present invention discloses a leg muscle training analysis method and a training chair based on image analysis, and relates to the field of image analysis technology. The method includes obtaining standard training movements for leg muscle training, obtaining a contour image of the leg position, determining the image area where the bone node is located in the contour image based on the standard training movement, obtaining the motion trajectory of the bone node in the image area based on the time series, and obtaining the leg muscle contour image in the image area within a motion trajectory cycle. The present invention obtains the attachment points of the leg muscles during real-time training through standard training movements, and then uses the training movements to determine the range data of effective stretching. Finally, the training range is narrowed by comparing the segments and positions in the leg muscle contour image that meet the effective stretching range data, thereby realizing dynamic tracking and analysis of the muscle status during leg training, and accurately enhancing the accuracy of evaluating the training effect.

Description

Technical Field

[0001] The invention relates to the technical field of image analysis, and in particular to a leg muscle training analysis method based on image analysis and a training chair. Background Art

[0002] Image analysis is a technology that uses cameras to capture human movements. It identifies and tracks various parts of the human body by analyzing image sequences, thereby achieving accurate measurement of human posture and movement. During leg muscle training, the use of image analysis technology can monitor the patient's leg movements in real time to ensure the accuracy and standardization of the training movements.

[0003] After searching, the Chinese invention patent with publication number "CN115272313A" discloses "Muscle balance analysis method, system and equipment based on depth image". The application uses depth images of the front and back of the human body, combined with depth data and bony landmarks, to infer the balance of muscle distribution on the human body surface. Compared with traditional visual touch evaluation methods and judgment methods based on muscle force, this technology is not affected by the subjective judgment of the tester, nor will it affect the analysis results due to improper muscle force or muscle damage of the subject. This technology has the advantages of few interference factors and high accuracy, and can provide quantitative values ​​of muscles, which is conducive to intuitive and accurate analysis of muscle balance.

[0004] The above-mentioned disclosed method and similar methods of analyzing muscles through images, when obtaining the uniformity of muscles, mostly keep the muscles in a static state to reduce the impact on the analysis results. However, the degree of muscle stretching and the muscle contraction state during training directly affect the training effect. Therefore, in the scenario of continuous training, a method that can dynamically track and analyze the muscle movement state is needed to enhance the accuracy of the analysis results. Summary of the invention

[0005] The purpose of the present invention is to provide a leg muscle training analysis method and a training chair based on image analysis to solve the problems raised in the above background technology.

[0006] In a first aspect, to achieve the above-mentioned purpose, the present invention provides the following technical solutions: a leg muscle training analysis method based on image analysis, comprising obtaining standard training movements for leg muscle training,

[0007] Obtaining a contour image of the leg position;

[0008] Determine the image region where the bone nodes are located in the contour image based on the standard training action;

[0009] Obtain the motion trajectory of the skeleton node in the image area based on the time series;

[0010] Acquire a leg muscle contour image at an image region within a motion trajectory cycle;

[0011] Based on standard training movements, determine the effective training range image within the leg muscle contour image;

[0012] Obtaining surface texture values ​​of leg muscles within the effective training range image;

[0013] Judging the effect of leg muscle training based on surface texture values;

[0014] Methods for determining leg muscle contour imaging include:

[0015] Obtaining a first bone node position, where the first bone node is any one of all bone nodes determined in the image region based on a standard training item;

[0016] Obtaining the position of the second bone node, where the second bone node is the one closest to the first bone node in the image area;

[0017] Collecting a first distance, where the first distance is the distance between the first bone node and the second bone node;

[0018] Obtain a second distance, where the second distance is the distance between the first bone node as the starting point and any determined bone node based on the standard training action;

[0019] Based on the time series, the change amplitudes of the first interval and the second interval are compared, and the accuracy of the leg muscle contour image is determined based on the degree of change.

[0020] As a further preferred embodiment of the present technical solution, the method for determining the leg muscle contour image includes:

[0021] Get the position of the third bone node, where the third bone node is the one farthest from the first bone node in the image area;

[0022] Based on the time series, respectively obtain the movement range of the third bone node and the first bone node;

[0023] The accuracy of the leg muscle contour image is determined based on the matching degree of the quadrants in which the two movement ranges are located.

[0024] As a further preferred embodiment of the present technical solution, the method for determining the image area is:

[0025] Obtain training items corresponding to standard training movements based on the Internet;

[0026] Determine the position and number of bone nodes in the leg based on the training project;

[0027] Get the angle variation range of any two non-adjacent bone nodes during a motion trajectory;

[0028] The image region where the bone node is located is determined based on the angle variation range.

[0029] As a further preferred embodiment of the present technical solution, the method for obtaining the motion trajectory includes:

[0030] A two-dimensional coordinate system is constructed based on the center point position of the image area, the abscissa of the two-dimensional coordinate system is used to represent the time series, and the ordinate is used to represent the movement distance of the bone node;

[0031] Collect the positions of the skeleton nodes in the two-dimensional coordinate system at different time series;

[0032] Connect the skeleton nodes of different time series in the two-dimensional coordinate system to form the motion trajectory of the skeleton nodes.

[0033] As a further preferred embodiment of the present technical solution, a method for determining an effective training range image within a leg muscle contour image is as follows:

[0034] Determine muscle attachment points based on the locations of bone nodes within the image region;

[0035] Based on standard training movements, the stretching direction and stretching range of the muscle attachment points during leg muscle training are obtained and effective training indicators are determined;

[0036] Based on the effective training index, the stretching degree value of the corresponding muscle attachment point in the leg muscle contour image is obtained;

[0037] The stretching degree value is compared with the stretching degree value of the standard training movement to determine the effective training range image within the leg muscle contour image.

[0038] As a further preferred embodiment of the present technical solution, the surface texture value is obtained by:

[0039] Capturing the image frequency features of the leg muscle surface area, the image frequency features include: low-frequency features and high-frequency features, the low-frequency features are used to represent smoothness, and the high-frequency features are used to represent texture changes;

[0040] The zebra pattern on the surface area of ​​the leg muscles matches the position of the image frequency features. The zebra pattern appears as a wavy artifact based on the high-frequency features, and the zebra pattern appears as no development based on the low-frequency features.

[0041] The proportion of leg muscle surface area occupied by ripple-like artifacts in the zebra pattern was obtained;

[0042] A numerical value representing the surface texture based on a scaled result.

[0043] As a further preferred embodiment of the present technical solution, a training feedback signal is constructed based on the establishment of an effective training range image;

[0044] Correcting the motion trajectory of the bone nodes in the image area based on the training feedback signal, and obtaining a corrected leg muscle contour image after the training feedback signal;

[0045] Eliminate the area where shadow pixels exist in the corrected leg muscle contour image, and form a corrected leg muscle contour image;

[0046] Obtaining the surface texture value of the corrected leg muscle contour image;

[0047] Surface texture values ​​to discriminate leg muscle training effects.

[0048] As a further preferred embodiment of the present technical solution, the operation method of the training feedback signal includes:

[0049] Real-time monitoring of the surface texture value changes of the leg muscle contour image per unit time. The surface texture value is used to represent the training load.

[0050] Obtain heart rate change data based on leg muscle exercise within a unit time;

[0051] Build correlation formulas based on heart rate variation data and training load;

[0052] The training intensity index is obtained based on the correlation formula, and the training intensity index is used to reflect the intensity and effect of the training;

[0053] Adjust training indicators based on the training intensity index, which includes the number and frequency of completing standard training movements.

[0054] As a further preferred embodiment of the present technical solution, the correlation formula is: , H is used to indicate the change in heart rate, T is used to indicate the change in surface texture value, a and β are used to indicate the weight ratio during training, and the specific values ​​are determined manually, and TII is used to indicate the training intensity.

[0055] In a second aspect, the present invention further proposes a training chair, which uses any one of the above-mentioned leg muscle training analysis methods based on image analysis.

[0056] Compared with the prior art, the present invention has the following beneficial effects:

[0057] This leg muscle training analysis method based on image analysis obtains the attachment points of leg muscles during real-time training through standard training movements, and then uses the training movements to determine the effective stretching range data. Finally, the training range is narrowed by comparing the segments and positions in the leg muscle contour image that meet the effective stretching range data. This achieves dynamic tracking and analysis of the muscle status during leg training, and precisely enhances the accuracy of evaluating the training effect.

[0058] In addition, the design of the training chair combined with the analysis method provides users with a comprehensive training solution that can help users perform scientific leg muscle training. By using the training chair, users can perform leg muscle training more conveniently, and adjust training intensity and movements through real-time feedback to achieve the best training effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 It is a diagram of the operation steps of the method of the present invention;

[0060] Figure 2 A diagram showing the motion trajectory of the present invention;

[0061] Figure 3 A diagram of a method for establishing a training feedback signal disclosed in Embodiment 2 of the present invention;

[0062] Figure 4 This is a structural diagram of the training chair disclosed in the present invention. DETAILED DESCRIPTION

[0063] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0064] Before understanding the implementation methods proposed in the present invention, it should be clear that in the present invention, image analysis technology is applied to leg muscle training. By analyzing the patient's movements on the training chair, the muscle activity and training effect are monitored in real time. In practical application, it can not only provide patients with instant feedback to help them adjust the training intensity and posture, but also provide data support for coaches to formulate more personalized training plans.

[0065] Specifically, refer to Figure 4 It can be seen that three high-speed cameras based on image analysis functions are installed outside the leg training position of the training chair. In addition, a central control display panel is installed on the seat of the training chair to display heart rate changes, training load intensity, real-time images, standard training movement views, exercise indicators, muscle fatigue and other information. It should be added that the high-speed camera can capture the patient's movements, and then analyze the accuracy of the movements and the activity of the muscles through the image processing algorithm carried by the processor in the central control display panel. In addition, the patient can see whether his movements are standard and the activity status of the muscles through the display screen, so as to adjust the training method in time.

[0066] Embodiment 1

[0067] For the rehabilitation training of patients in the hospital, since the patients' legs have been injured, they are unable to accurately complete the standard training movements in the early stage of leg exercise training. In order to increase the amount of leg muscle contour image data obtained, so as to further analyze the patient's leg muscle rehabilitation status, the present invention proposes a leg muscle training analysis method based on image analysis.

[0068] It should be noted that before the actual operation of this method, it is necessary to obtain standard training movements for leg muscle training from a rehabilitation trainer, and input the standard training movements into the training chair through the central control display panel.

[0069] refer to Figure 1 It can be seen that it needs to be supplemented that when the method is running, it includes: step S100-step S700.

[0070] Step S100: Acquire a contour image of the leg position.

[0071] It should be noted that in the present invention, the contour image is acquired through a high-speed camera on the training chair. The camera can capture the patient's leg contour in real time and convert it into a digital signal to display it in the processor in the central control display panel.

[0072] Step S200: determining the image region where the skeleton nodes are located in the contour image based on the standard training action.

[0073] It should be clear that in the present invention, step S200 is used to more accurately analyze whether the patient's movements meet the requirements of standard training movements.

[0074] It should be supplemented that the method for determining the image area in step S200 includes: step S201 to step S204.

[0075] Specifically, step S201: obtaining training items corresponding to standard training movements based on the network.

[0076] It should be explained that in step S201, the standard training movements are provided by sports medicine experts, trainers or rehabilitation institutions. These standard training movements can be common leg training movements such as leg presses and lunges. In the actual acquisition process, the processor in the central control display panel is connected to the Internet, and through accessing professional sports databases, training platforms or online resources (such as fitness teaching websites, rehabilitation research platforms), it queries the training items corresponding to the standard training movements and the surface stretching values ​​of the leg muscles during training.

[0077] Step S202: Determine the position and number of bone nodes in the leg based on the training item.

[0078] It should be noted that, in step S202, since the training item is a known item, the position and number of the skeleton nodes are set.

[0079] Specifically, since each movement included in the training program has its specific bone node positions, these positions are determined based on sports anatomy principles and sports biomechanics analysis. For example, in the leg press movement, the moving bone nodes include the knee, hip joint and ankle.

[0080] Step S203: Obtain the angle variation range of any two known non-adjacent skeletal nodes during a motion trajectory.

[0081] It should be clear that in step S203, for any two non-adjacent bone nodes, such as the knee and hip joints, their angle change range is monitored in real time by a high-precision sensor in a high-speed camera tube. The high-precision sensor captures the subtle changes in the angles between the bone nodes when the patient performs the action, and transmits the data to the processor in the central control display panel. The processor then analyzes the angle change data to ensure the accuracy and safety of the training movements.

[0082] Step S204: determining the image region where the skeleton node is located based on the angle variation range.

[0083] It should be clear that in step S204, when the angle formed between the bone nodes in an image area within the contour image is within the angle variation range, the captured object is reduced from the contour image to the image area within the contour image, thereby achieving accurate tracking and analysis of specific bone nodes, which enhances the accuracy of the analysis to a certain extent.

[0084] Step S300: acquiring the motion trajectory of the skeleton node in the image area based on the time series.

[0085] It should be clear that step S300 in the present invention is used to track the changes in the patient's movements during the training process to ensure the continuity and standardization of the movements. When applied to a training chair, a high-speed camera is used to continuously capture the position changes of the patient's bone nodes in the image area to form a series of coordinate points. These coordinate points are recorded in time series to depict the movement trajectory of the bone nodes and are mapped on the central control display panel. The central control display panel analyzes these trajectory data to determine whether the patient is training according to the standard training movements and whether the movements are executed in place.

[0086] It should be further explained that the method for obtaining the motion trajectory in step S300 includes: step S301 to step S303.

[0087] Specifically, refer to Figure 2It can be seen that in the present invention, step S301: constructing a two-dimensional coordinate system based on the center point position of the image area.

[0088] It should be clear that the horizontal axis of the two-dimensional coordinate system is used to represent the time series, and the vertical axis is used to represent the movement distance of the bone node. It should be supplemented that step S301 is used to determine the movement trajectory of the bone node in the image area. In this way, the movement of the bone node is converted into quantifiable data, which is convenient for subsequent analysis and evaluation.

[0089] Step S302: Collect the positions of the skeleton nodes in the two-dimensional coordinate system in different time series.

[0090] Step S303: Connect the skeleton nodes of different time series in the two-dimensional coordinate system to form a motion trajectory of the skeleton nodes.

[0091] Step S400: Acquire a leg muscle contour image in an image area within a motion trajectory cycle.

[0092] It should be clear that in the present invention, step S400 is used to obtain the leg muscle contour image of the patient during a complete movement trajectory cycle, so as to analyze the muscle activity and strength changes. The specific operation includes: step S401-step S405, wherein step S401-step S405 are images of the leg muscle plane direction.

[0093] Step S401: Obtain the position of a first bone node, where the first bone node is any one of all bone nodes determined within the image area based on a standard training item.

[0094] It should be clear that step S401 is used to determine a reference point to facilitate the subsequent analysis of the muscle contour image. During the specific operation, the processor will select a bone node (knee, hip joint, ankle) in the image area. The selection of this node is usually related to the standard training movement. During the exercise, the position of this bone node is used as a reference, and the subsequent analysis of other bone nodes and muscle contour images will refer to this node.

[0095] Step S402: Obtain the position of a second bone node, where the second bone node is the one closest to the first bone node in the image area.

[0096] It should be noted that the determination of the second bone node is used to assist in determining the boundary of the muscle contour image. Since the boundary of the muscle contour image is affected by the position of the bone nodes, the position information of the second bone node helps to more accurately define the range of the muscle contour. During specific operations, the processor finds the bone node closest to the first bone node in the image area (such as the patella next to the knee joint or other related bones). The selection of this node helps to define the activity area of ​​the leg muscles and the motion trajectory of the training movements.

[0097] Step S403: Collect a first distance, where the first distance is the distance between the first bone node and the second bone node.

[0098] Step S404: obtaining a second distance, where the second distance is the distance between the standard training action based on the first bone node as the starting point and any determined bone node.

[0099] Step S405: Based on the time series, compare the change ranges of the first spacing and the second spacing, and determine the accuracy of the leg muscle contour image based on the degree of change.

[0100] It should be noted that steps S403 to S405 are used in the present invention to obtain leg muscle contour images of the patient during the exercise cycle, and analyze muscle activity, strength changes, and movement accuracy based on the precise positioning of bone nodes and their changes.

[0101] In addition, it should be further explained that when determining the leg muscle contour image, the image of the leg muscle in the plane direction and the image of the leg muscle in the vertical direction show different effects.

[0102] Specifically, for the vertical leg muscle image, it includes steps S410 to S430.

[0103] Step S410: Acquire the position of the third bone node, where the third bone node is the one farthest from the first bone node in the image area.

[0104] It should be clear that step S410 is used to determine another reference point for leg muscle activity in order to more comprehensively analyze the muscle contour image. In actual operation, the processor will select a bone node (such as the toes or buttocks) that is farthest from the first bone node in the image area. The selection of this node helps to evaluate the range of activity and movement trajectory of the leg muscles from different angles. By combining the position information of the first, second and third bone nodes, the dynamic changes of the leg muscles during exercise can be more accurately depicted, thereby providing patients with more detailed training feedback and suggestions.

[0105] Step S420: Obtain the movement ranges of the third skeleton node and the first skeleton node respectively based on the time series.

[0106] It should be clear that step S420 is used to track the position changes of the third skeletal node during the training process, so as to obtain its movement trajectory in the vertical direction. In this way, the activity of the leg muscles in the vertical direction can be evaluated, which is crucial for understanding the force output and movement patterns of muscles in different directions. In actual applications, the processor will record the vertical movement range of the third skeletal node, which helps to analyze the degree of extension and contraction of the patient's leg muscles when performing movements.

[0107] Step S430: Determine the accuracy of the leg muscle contour image based on the matching degree of the quadrants where the two movement ranges are located.

[0108] It should be clear that in step S430, by comparing the quadrants in which the vertical movement ranges of the third skeletal node and the first skeletal node are located, the accuracy of the leg muscle contour image can be evaluated. If the two movement ranges are in the same quadrant of the space, it indicates that the accuracy of the leg muscle contour image is high. Otherwise, the training posture or analysis method may need to be adjusted. In this way, it can be ensured that the patient can obtain accurate muscle activity feedback when using the training chair for leg muscle training, thereby improving the training effect and safety.

[0109] Step S500: Determine the effective training range image within the leg muscle contour image based on the standard training movements.

[0110] It should be clear that in the present invention, step S500 is used to determine which areas in the leg muscle contour image are effectively involved in the training when the patient performs standard training movements. The specific operations include: step S501-step S504.

[0111] Step S501: Determine muscle attachment points based on the positions of bone nodes in the image area.

[0112] It should be clear that in the present invention, step S501 is used to determine the position of the muscle attachment point to facilitate the subsequent analysis of the muscle's range of motion and force output. In actual operation, the processor will identify the specific position of the muscle attachment point based on the position information of the bone nodes in the image area. These points are the sites of muscle force output. By determining these attachment points, more accurate training feedback is provided to patients to help them adjust the training intensity and range to achieve more effective training results.

[0113] Step S502: obtaining the stretching direction and stretching range of the muscle attachment points during leg muscle training based on standard training movements and determining effective training indicators.

[0114] It should be clear that in step S502 of the present invention, the activity level of the muscle during training is determined by measuring the movement of the muscle attachment points, thereby providing the patient with quantitative training indicators. It should be supplemented that the quantifiable training indicators include the degree of muscle stretching, force output and fatigue level.

[0115] Step S503: Based on the effective training index, the stretching degree value of the corresponding muscle attachment point in the leg muscle contour image is obtained.

[0116] It should be clear that in the present invention, effective training indicators include changes in muscle length. Through these parameters, in actual application, the processor measures the pixel distance of the corresponding muscle attachment points in the leg muscle contour image, and converts it into the stretching degree data output of the muscle attachment points at the corresponding positions of the patient in reality.

[0117] Step S504: Compare the stretching degree value with the stretching degree value of the standard training action to determine the effective training range image within the leg muscle contour image.

[0118] It should be clear that when the present invention is actually used, the accuracy and effectiveness of the training movements are ensured by real-time monitoring of muscle activity during the training process. Specifically, when it is detected that the degree of muscle stretching exceeds or does not reach the preset effective training range, the system will automatically adjust the resistance or angle of the training chair to guide the user to adjust the movement, thereby avoiding overtraining or undertraining.

[0119] Step S600: Acquire surface texture values ​​of leg muscles within the effective training range image.

[0120] It should be understood that the method for obtaining the surface texture value includes: step S601 to step S603.

[0121] Step S601: Capturing image frequency characteristics of the leg muscle surface area.

[0122] It should be clear that the image frequency features in step S601 include: low-frequency features and high-frequency features. The low-frequency features are used to represent smoothness, and the high-frequency features are used to represent texture changes.

[0123] It should be added that in the present invention, the acquisition of image frequency characteristics is achieved through the fast Fourier transform (FFT) algorithm in the existing image acquisition technology. The FFT algorithm can efficiently convert the image from the spatial domain to the frequency domain, thereby extracting low-frequency and high-frequency components.

[0124] Specifically, in the present invention, the leg muscle images captured by the high-speed camera are first preprocessed to remove noise and improve image quality. Subsequently, the processor applies the FFT algorithm to perform frequency domain conversion on the preprocessed images, thereby separating low-frequency and high-frequency components. The low-frequency components mainly reflect the smoothness of the muscle surface, while the high-frequency components reveal the texture details of the muscle surface. By analyzing these frequency characteristics, the surface state of the muscles can be accurately evaluated, providing a basis for subsequent adjustments to training movements.

[0125] Step S602: Mapping a zebra pattern that matches the image frequency feature position on the leg muscle surface area.

[0126] It should be noted that in the present invention, the zebra pattern is displayed as a wavy artifact based on the high-frequency feature, and the zebra pattern is displayed as no development based on the low-frequency feature.

[0127] Specifically, in the present invention, after acquiring the frequency characteristics of the leg muscle image, the processor will match these characteristics with a preset zebra pattern template. Through this mapping, the specific positions of the high-frequency characteristics and the low-frequency characteristics are intuitively displayed in the muscle surface area. Through this visual distinction, patients and medical staff can more easily identify and understand the surface state of the muscles, thereby making more precise adjustments to the training movements.

[0128] Step S603: obtaining the ratio of the leg muscle surface area occupied by the wavy artifacts in the zebra pattern, and representing the value of the surface texture based on the ratio result.

[0129] It should be clear that in the present invention, step S603 is used to evaluate the texture condition of the muscle surface. Specifically, the higher the proportion of ripple-like artifacts, the greater the fluctuation of the skin on the muscle surface during leg muscle training, and vice versa. Through this ratio value, the activity state of the muscle is quantitatively analyzed to provide a scientific basis for adjusting the training intensity. In addition, the ratio results can also be used to track changes in muscle status over a long period of time, helping medical staff monitor training effects and muscle recovery.

[0130] Step S700: Determine the leg muscle training effect based on the surface texture value.

[0131] It should be noted that in the present invention, step S700 is used to evaluate the training effect. The specific method is to compare the obtained surface texture value with the stretching value of the leg muscle surface during training obtained by the standard training action. If the surface texture value is lower than the stretching value of the standard training action, it means that the training intensity is insufficient, the muscle activity has not achieved the expected effect, and the training load needs to be increased. Conversely, if the surface texture value is higher than the standard value, it indicates that the training intensity is too high, which will cause excessive muscle fatigue or damage. At this time, the training intensity should be appropriately reduced.

[0132] Embodiment 2

[0133] When the patient is doing leg training on the training chair, the patient's leg training movements are fed back. In Example 2, when the patient starts leg training, the sensor in the high-speed camera captures the patient's movement data and generates a training feedback signal to transmit to the processor. The processor analyzes whether the patient's movements are standard and whether there are deviations based on preset training movement standards, and outputs real-time correction instructions to the patient based on the training feedback signal.

[0134] It should be noted that the reference Figure 3 It can be seen that the method for establishing the training feedback signal includes:

[0135] Step A100-Step A500.

[0136] Step A100: constructing a training feedback signal based on the establishment of the effective training range image.

[0137] Step A200: Correct the motion trajectory of the skeleton nodes in the image area based on the training feedback signal, and obtain a corrected leg muscle contour image after the training feedback signal.

[0138] It should be clear that in the present invention, the data contained in the training feedback signal includes the motion trajectory information of the skeletal nodes, the shape change data of the muscle contour, and the intensity index of the muscle activity. Through these data, the processor can accurately identify whether the patient's training movements meet the standards, and can adjust the training feedback signal in real time to guide the patient to perform correct muscle training.

[0139] Step A300: Eliminate the areas with shadow pixels in the corrected leg muscle contour image, and form a modified leg muscle contour image.

[0140] It should be noted that in the present invention, the shadow pixels in step A300 include shielding objects and training protective gear carried by the patient.

[0141] Specifically, the shadow pixels are identified and distinguished by using an image processing algorithm. Specifically, the processor analyzes the pixel brightness and color information in the image to distinguish those areas with large differences in brightness and color contrast from the surrounding environment. These areas usually correspond to occlusions or training protective gear. Then, by comparing the motion patterns of these areas in consecutive frames, the algorithm can further confirm whether these areas are stationary occlusions or training protective gear that moves with the patient's movements. Once the shadow pixels are determined, the processor will remove these areas from the training feedback signal to ensure that the generated corrected leg muscle contour image can accurately reflect the patient's actual muscle movement.

[0142] Step A400: Obtain the surface texture value of the corrected leg muscle contour image.

[0143] Step A500: Using surface texture values ​​to determine leg muscle training effects.

[0144] It should be noted that step A400 and step A500 match the contents of step S600 and step S700 in embodiment 1, respectively.

[0145] It should be further explained that, in step A100, the operation method of the training feedback signal includes: step A101-step A105.

[0146] Step A101: Real-time monitoring of the surface texture value changes of the leg muscle contour image per unit time, where the surface texture value is used to represent the training load.

[0147] It should be noted that the method for obtaining the surface texture value in step A101 is the same as step S600 in embodiment 1, and the unit time in step A101 is represented as 1 minute.

[0148] Step A102: Obtain heart rate variation data based on leg muscle exercise within a unit time.

[0149] Step A103: Construct a correlation formula based on the heart rate change data and the training load.

[0150] It should be noted that the correlation formula in step A103 is: , H is used to indicate the change in heart rate, T is used to indicate the change in surface texture value, a and β are used to indicate the weight ratio during training, and the specific values ​​are determined manually, and TII is used to indicate the training intensity.

[0151] Step A104: A training intensity index is obtained based on the association formula. The training intensity index is used to reflect the intensity and effect of the training.

[0152] Specifically, when the associated formula is actually running, for example, the heart rate change data obtained is 10 times / minute, and the surface texture value changes by 0.5 units. If a is set to 0.6 and β is set to 0.4, the training intensity TII can be calculated as: TII = 0.6 *10 + 0.4 * 0.5 = 6.2. This value reflects the comprehensive intensity of the current training and can be used as a basis for adjusting the training plan.

[0153] Step A105: adjusting the training index based on the training intensity index, the training index includes the number and frequency of completing the standard training movements. It should be noted that in the present invention, the adjustment of the training index in step A105 is dynamic and is adjusted in real time according to the change of the training intensity index TII. For example, if the TII value shows that the training intensity is low, then the training index will be increased accordingly to ensure the training effect; conversely, if the TII value is too high, it indicates that the training intensity is too high, which will cause excessive load on the muscles. At this time, the training index will be reduced to avoid sports injuries. In this way, the patient can obtain the best training effect while ensuring safety. Specifically, the adjustment of the training index can be to increase the number of standard training movements or to increase the frequency of the movements. For example, the patient completed 20 standard movements in the previous training and the frequency was 10 times per minute, while the current TII value is 6.2, indicating that the training intensity is moderate, but there is still room for improvement. Then the training index can be adjusted to complete 25 standard movements and the frequency is increased to 12 times per minute. Such adjustment is intended to ensure that the patient gradually increases the training intensity without increasing excessive load, thereby achieving better training effects.

[0154] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations can be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is limited by the attached embodiments and their equivalents.

Claims

1. A leg muscle training analysis method based on image analysis, comprising obtaining standard training movements for leg muscle training, characterized in that: Obtaining a contour image of the leg position; Determine the image region where the bone nodes are located in the contour image based on the standard training action; Obtain the motion trajectory of the skeleton node in the image area based on the time series; Acquire a leg muscle contour image at an image region within a motion trajectory cycle; Based on standard training movements, determine the effective training range image within the leg muscle contour image; Obtaining surface texture values ​​of leg muscles within the effective training range image; Judging the effect of leg muscle training based on surface texture values; Method for determining the effective training range image within the leg muscle contour image: Determine muscle attachment points based on the locations of bone nodes within the image region; Based on standard training movements, the stretching direction and stretching range of the muscle attachment points during leg muscle training are obtained and effective training indicators are determined; Based on the effective training index, the stretching degree value of the corresponding muscle attachment point in the leg muscle contour image is obtained; Compare the stretching degree value with the stretching degree value of the standard training action to determine the effective training range image within the leg muscle contour image; Constructing training feedback signals based on the establishment of effective training range images; Correcting the motion trajectory of the bone nodes in the image area based on the training feedback signal, and obtaining a corrected leg muscle contour image after the training feedback signal; Eliminate the area where shadow pixels exist in the corrected leg muscle contour image, and form a corrected leg muscle contour image; Obtaining the surface texture value of the corrected leg muscle contour image; Surface texture values ​​to determine leg muscle training effects; The operation methods of training feedback signals include: Real-time monitoring of the surface texture value changes of the leg muscle contour image per unit time. The surface texture value is used to represent the training load. Obtain heart rate change data based on leg muscle exercise within a unit time; Build correlation formulas based on heart rate variation data and training load; The training intensity index is obtained based on the correlation formula, and the training intensity index is used to reflect the intensity and effect of the training; Adjust training indicators based on the training intensity index, including the number and frequency of completing standard training movements; The association formula is: H is used to indicate changes in heart rate, T is used to indicate changes in surface texture values, a and β are used to indicate weight ratios during training, and the specific values ​​are determined manually. TII is used to indicate training intensity.

2. The leg muscle training analysis method based on image analysis according to claim 1, characterized in that: Methods for determining leg muscle contour imaging include: Obtaining a first bone node position, where the first bone node is any one of all bone nodes determined in the image region based on a standard training item; Obtaining the position of the second bone node, where the second bone node is the one closest to the first bone node in the image area; Collecting a first distance, where the first distance is the distance between the first bone node and the second bone node; Get a second distance, where the second distance is a standard training action based on the first bone node as a starting point and any one to determine the distance between the bone nodes; Based on the time series, the change amplitudes of the first interval and the second interval are compared, and the accuracy of the leg muscle contour image is determined based on the degree of change.

3. The leg muscle training analysis method based on image analysis according to claim 2, characterized in that: Methods for determining leg muscle contour imaging include: Get the position of the third bone node, where the third bone node is the one farthest from the first bone node in the image area; Based on the time series, respectively obtain the movement range of the third bone node and the first bone node; The accuracy of the leg muscle contour image is determined based on the matching degree of the quadrants in which the two movement ranges are located.

4. The leg muscle training analysis method based on image analysis according to claim 1, characterized in that: Method for determining image area: Obtain training items corresponding to standard training movements based on the Internet; Determine the position and number of bone nodes in the leg based on the training project; Get the angle variation range of any two non-adjacent bone nodes during a motion trajectory; The image region where the bone node is located is determined based on the angle variation range.

5. The leg muscle training analysis method based on image analysis according to claim 1, characterized in that: The methods for obtaining motion trajectories include: A two-dimensional coordinate system is constructed based on the center point position of the image area, the abscissa of the two-dimensional coordinate system is used to represent the time series, and the ordinate is used to represent the movement distance of the bone node; Collect the positions of the skeleton nodes in the two-dimensional coordinate system at different time series; Connect the skeleton nodes of different time series in the two-dimensional coordinate system to form the motion trajectory of the skeleton nodes.

6. The leg muscle training analysis method based on image analysis according to claim 1, characterized in that: The methods for obtaining surface texture values ​​include: Capturing the image frequency features of the leg muscle surface area, the image frequency features include: low-frequency features and high-frequency features, the low-frequency features are used to represent smoothness, and the high-frequency features are used to represent texture changes; The zebra pattern on the surface area of ​​the leg muscles matches the position of the image frequency features. The zebra pattern appears as a wavy artifact based on the high-frequency features, and the zebra pattern appears as no development based on the low-frequency features. The proportion of leg muscle surface area occupied by ripple-like artifacts in the zebra pattern was obtained; A numerical value representing the surface texture based on a scaled result.

Images