Medical rehabilitation underwater gait analysis system and method based on skeleton calculation

Abstract

The invention relates to the technical field of medical instruments, and discloses a medical rehabilitation underwater gait analysis system and method based on skeleton calculation. The system comprises a plurality of underwater cameras used for shooting moving images of human bodies in water; the light source equipment is mounted underwater and is used for irradiating the reflective marker ball on the human body; the reflective marker ball is used for being mounted on a preset mark point of a human body so as to reflect light; the image processing module is used for acquiring an underwater human body motion image shot by the underwater camera and carrying out image distortion correction and mark point position detection; and the skeleton resolving module is used for carrying out skeleton model construction and skeleton resolving based on the position change of the mark points so as to carry out gait analysis. The method has the advantage of being capable of achieving efficient and accurate motion capture and gait analysis in a complex underwater environment.

Description

Technical Field

[0001] The present application relates to the technical field of medical devices, and in particular to a medical rehabilitation underwater gait analysis system and method based on skeleton solution. Background Art

[0002] As an important means of acquiring dynamic data, motion capture technology has been widely used in medical rehabilitation, sports training, animation production, robot navigation and other fields. Existing motion capture systems are mainly divided into mechanical, acoustic, electromagnetic and optical types. Among them, the optical motion capture system is particularly suitable for scenes that require accurate dynamic data due to its non-contact capture, high precision and high sampling frequency.

[0003] The underwater optical 3D motion capture system in related technologies reduces the interference of ambient light by using infrared cameras and active light-emitting markers, and improves the capture accuracy to a certain extent. However, water has a significant absorption effect on light, especially long-wavelength light (such as infrared light), which will rapidly weaken as the propagation distance increases. This light attenuation phenomenon directly affects the capture range and accuracy, and it is difficult to fully meet the actual application needs of underwater environments. Summary of the invention

[0004] In order to achieve efficient and accurate motion capture and gait analysis in complex underwater environments, the present application provides a medical rehabilitation underwater gait analysis system and method based on skeleton solution.

[0005] In a first aspect, the present application provides a medical rehabilitation underwater gait analysis system based on skeleton solution, which adopts the following technical solution:

[0006] A medical rehabilitation underwater gait analysis system based on skeleton solution, comprising:

[0007] Multiple underwater cameras for capturing images of human motion in water;

[0008] Light source equipment, installed underwater, used to illuminate the reflective marker ball on the human body;

[0009] Reflective marker balls are used to be installed on pre-set marking points on the human body to reflect light;

[0010] An image processing module is used to obtain the image of human body motion in water taken by an underwater camera, and to perform image distortion correction and position detection of marking points;

[0011] The skeleton solving module is used to construct skeleton models and solve skeletons based on the position changes of marker points for gait analysis.

[0012] Optionally, the reflective marker ball includes:

[0013] A plastic shell, the plastic shell comprising two hemispherical transparent plastic shells, the two hemispherical plastic shells being able to be combined into a closed sphere;

[0014] Reflective stickers are used to stick to the inner side of the plastic shell.

[0015] Optionally, the image processing module includes:

[0016] An image acquisition unit, used to acquire an image of a human body in water captured by an underwater camera;

[0017] A noise reduction unit, used for performing noise reduction processing through a filtering algorithm;

[0018] A distortion correction unit, used for performing distortion correction to correct image distortion problems caused by the camera lens and the underwater environment;

[0019] The marker point positioning unit is used to extract the position information of each marker point in the image using an edge detection algorithm.

[0020] Optionally, the skeleton solving module includes:

[0021] A joint position marking unit, used to detect and mark the positions of various joints of the human body in the preprocessed image;

[0022] An initial skeleton model building unit, used to build an initial skeleton model through the three-dimensional coordinates of the marking points;

[0023] A skeleton solving unit, used for solving the skeleton model based on inverse dynamics and / or forward dynamics to generate three-dimensional gait data;

[0024] The gait analysis unit is used to perform gait analysis on the three-dimensional gait data generated by the skeleton solving unit, wherein the gait analysis includes the calculation and evaluation of the stride, stride frequency and joint angle changes.

[0025] Optionally, the skeleton solving unit is used to perform the following steps:

[0026] Set the target position and orientation of the end effector;

[0027] Starting from the end effector, reversely calculate the rotation angles of the skeleton joints one by one;

[0028] According to the rotation angle, calculate the transformation matrix of the bone;

[0029] Using the transformation matrix, calculate the position and orientation of each joint;

[0030] Improve the accuracy and stability of the solution through iterative optimization.

[0031] Optionally, the joint position marking unit is used to perform the following steps:

[0032] Initialize the human skeleton model according to the position information of the marker points;

[0033] Determine the position and connection relationship of each joint;

[0034] Determine the accuracy of skeleton initialization to avoid solving errors due to marker errors.

[0035] Optionally, the reflective marker ball is attached to four points on the waist, one on each of the left and right thigh roots, one on the inner and outer sides of the knees, one on each of the toes and heels, one on each of the outer sides of the little toes, and one on each of the inner and outer sides of the ankles.

[0036] In the second aspect, the present application provides a method for underwater gait analysis for medical rehabilitation based on skeleton solution, which adopts the following technical solution:

[0037] A method for underwater gait analysis for medical rehabilitation based on skeleton solution, comprising:

[0038] S1. Obtaining an image of a human body moving in water captured by an underwater camera;

[0039] S2. De-noising by filtering algorithm;

[0040] S3. Perform distortion correction to correct image distortion caused by the camera lens and the underwater environment;

[0041] S4. Using edge detection algorithm, extract the position information of each marker point in the image;

[0042] S5. In the preprocessed image, detect and mark the positions of various joints of the human body;

[0043] S6. construct an initial skeleton model through the three-dimensional coordinates of the marking points;

[0044] S7. Solving the skeleton model based on inverse dynamics and / or forward dynamics to generate three-dimensional gait data;

[0045] S8. Perform gait analysis on the three-dimensional gait data generated by the skeleton solving unit, wherein the gait analysis includes calculation and evaluation of stride length, stride frequency, and joint angle changes.

[0046] Optionally, the S7 comprises the following steps:

[0047] S71. Setting the target position and direction of the end effector;

[0048] S72. Starting from the end effector, reversely calculate the rotation angles of the skeletal joints one by one;

[0049] S73. Calculate the transformation matrix of the skeleton according to the rotation angle;

[0050] S74. Using the transformation matrix, calculate the position and orientation of each joint;

[0051] S75. Improve the accuracy and stability of the solution through iterative optimization.

[0052] In summary, the present application includes at least one of the following beneficial technical effects:

[0053] 1. The system uses non-contact optical capture technology to reduce interference to patients and is suitable for patients who are inconvenient to move or sensitive to traditional sensors. Combined with high-precision image processing and bone solution algorithms, it can accurately capture gait characteristics and provide doctors with a scientific basis for formulating personalized rehabilitation plans. The system supports real-time feedback and long-term monitoring, providing strong support for dynamic adjustment and trend analysis during the rehabilitation process, greatly improving rehabilitation effects and efficiency.

[0054] 2. The system realizes accurate quantitative evaluation of athletes' movement techniques, and helps coaches adjust movements and optimize techniques in real time during training through real-time feedback. The long-term monitoring function supports the phased evaluation of training progress and provides a scientific basis for formulating the next stage of training plans. In addition, gait analysis can identify potential injury risks and prevent them in advance, while assisting the scientific rehabilitation of athletes after injury and accelerating their return to the field.

[0055] 3. The system efficiently captures real gait movements and converts them into 3D animation data, reducing the manual workload of animators. Real-time data feedback supports fine-tuning of character movements, making animation more natural and smooth. Gait capture for a variety of scenarios significantly improves the efficiency and visual effects of animation production, especially in sports, dance or complex dynamic scenes, further enhancing the viewing experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 This is a module connection diagram of a medical rehabilitation underwater gait analysis system based on skeleton solution in this application.

[0057] Figure 2 Used to illustrate a gait analysis chart in a certain embodiment of the present application.

[0058] Figure 3 Used to illustrate the joint movement angle diagram in a certain embodiment of the present application. DETAILED DESCRIPTION

[0059] Embodiments of the present application are described in detail below, examples of which are shown in the accompanying drawings.

[0060] In the description of this specification, the description with reference to the terms "certain embodiments", "one embodiment", "some embodiments", "illustrative embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials, or characteristics described in conjunction with the embodiments or examples are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0061] The present application embodiment discloses a medical rehabilitation underwater gait analysis system based on skeleton solution, referring to Figure 1 The medical rehabilitation underwater gait analysis system based on skeleton solution includes multiple underwater cameras, light source equipment, multiple reflective marker balls, an image processing module and a skeleton solution module.

[0062] Multiple underwater cameras are used to capture images of human body motion in water. Light source equipment is installed underwater to illuminate reflective marker balls on the human body. Reflective marker balls are installed on pre-set marking points on the human body to reflect light. An image processing module is used to obtain images of human body motion in water captured by underwater cameras, and to perform image distortion correction and marker point position detection. A skeleton solution module is used to construct a skeleton model and perform skeleton solution based on the position changes of marker points in order to perform gait analysis.

[0063] All underwater cameras and light source equipment are connected to the image processing module and the skeleton solution module through high-strength waterproof cables. These cables not only have good data transmission performance, but also have waterproof and corrosion-resistant properties, and can meet the needs of long-term underwater operations. During the connection process, the signal output end of the underwater camera corresponds one-to-one with the input interface of the image processing module, and a tight connection is achieved through a standard interface to ensure that the image data can be stably transmitted to the image processing module. At the same time, the light source equipment is connected to an independent power supply module through a cable. The power supply module is set on the water platform and has a waterproof and sealed interface with the light source equipment to ensure the continuous and stable operation of the light source.

[0064] In order to improve the safety and reliability of system operation, an electromagnetic shielding layer is added to the cable connection to avoid the impact of electromagnetic interference in water on data transmission. In addition, the system is tested multiple times after connection to ensure that the power supply and data transmission lines of all devices are connected properly and there is no looseness or disconnection at the interface, thus ensuring that the entire system can operate efficiently in complex underwater environments.

[0065] The underwater camera is a high-resolution dedicated device. The camera housing is made of high-strength waterproof material, and the multi-layer sealing structure ensures that it can still work normally at a depth of up to 40 meters. In order to further improve the imaging performance, the camera is equipped with a high-precision active filter, which effectively reduces the attenuation and scattering of light in the water, thereby ensuring image clarity and detail presentation under low light conditions.

[0066] The camera has a resolution of 4K or above, which can capture rich image details, so that even the slightest changes in human movement can be clearly recorded. In addition, it supports a frame rate of up to 200fps, which can achieve smooth image capture even in fast motion, avoiding blur and delay in motion images. In order to cope with possible particle interference in underwater environments, the camera's optical system has been specially designed. Through the combination of filters and image processing technology, background interference is eliminated to the greatest extent, improving the recognition accuracy of markers.

[0067] During installation, the camera is fixed in a suitable position in the underwater experimental area by a special bracket. The bracket is corrosion-resistant and adjustable, and can adjust the angle and shooting range of the camera according to different experimental requirements. Before the experiment, the focus, exposure and filter parameters of the camera are accurately calibrated to ensure that the image quality of the captured area is optimal. During the experiment, the camera maintains a real-time connection with the image processing module, and the captured image data will be immediately transmitted to the image processing module for preprocessing.

[0068] The light source equipment uses high-performance LED fill light, which has the characteristics of stable light intensity, high efficiency and long life. When used underwater, the LED light source effectively reduces the absorption and scattering of light in the water by optimizing the wavelength design, thereby ensuring that the light can evenly cover the entire capture area under different depths and water conditions, improving the image contrast and visibility of the mark point.

[0069] The light source equipment is installed at a specific location on the underwater experimental platform and fixed with a corrosion-resistant and pressure-resistant bracket. The bracket design is adjustable and can adjust the angle and height of the light source according to the specific needs of the experimental area to ensure that the light can be evenly irradiated on the human body marking points. Before the experiment, the light source equipment was precisely calibrated in brightness and angle to adapt to the light requirements of the underwater environment and avoid local overexposure or shadows that affect the recognition of marking points.

[0070] The light source equipment is connected to the independent power supply module through a special cable. The module is located on the surface experimental platform and connected to the light source equipment through a waterproof sealing interface to ensure the stability and safety of the equipment during long-term underwater operation. During the installation and commissioning of the light source, the lighting range and light uniformity are tested to ensure that the light source performs best in the experiment.

[0071] The reflective marker ball consists of two transparent plastic hemispheres that are sealed and bonded to form a complete sphere. The plastic shell is made of lightweight, strong and corrosion-resistant materials, which can withstand underwater pressure and corrosion for a long time while maintaining its structural stability and performance. In order to improve the optical recognition effect, a highly reflective, waterproof and wear-resistant sticker is affixed to the inner side of the plastic shell. This sticker can effectively enhance the reflective performance and ensure that the motion capture system can accurately identify the marker points even in complex underwater environments.

[0072] During the manufacturing process, the plastic shell is prepared by injection molding to ensure the precise size and consistent shape of the hemispheres, which is convenient for subsequent assembly and sealing. The two hemispheres are bonded together by a special adhesive, and the bonding point is strictly sealed to prevent water from penetrating into the sphere, thereby ensuring the long-term waterproof performance of the marker ball. In the process of laminating the reflective sticker, the sticker is evenly applied with glue or adhesive on the inner side of the plastic shell, and then the sticker is flatly pasted on the inner surface, which ensures the firmness and reflective effect of the sticker.

[0073] Reflective marker balls are precisely installed at preset key parts of the human body, including four points on the waist, one on each of the left and right thigh roots, one on the inner and outer sides of the knees, one on the toes and heels, one on the outer side of the little toe, and one on the inner and outer sides of the ankles. This distribution design ensures that the marker points required for gait analysis cover the main joints and movement parts of the lower limbs, and can fully capture the dynamic characteristics of the lower limb movement of the human body.

[0074] The reflective marker ball's performance under water has been rigorously tested to ensure that it can maintain high reflectivity, water resistance and durability after long-term use. This design not only meets the needs of high-precision gait analysis, but also maintains stable operation under harsh underwater conditions, providing reliable data support for subsequent image processing and skeleton solution. Through such an optimized design, the reflective marker ball provides an important guarantee for the efficient operation of the entire system.

[0075] The image processing module is a key functional component in the embodiment of the present application. Its core task is to process and optimize the images collected by the underwater camera to extract high-precision marker point position information and provide reliable data support for subsequent skeleton solution. The image processing module is composed of an image acquisition unit, a noise reduction unit, a distortion correction unit and a marker point positioning unit. The image acquisition unit is used to acquire the underwater human body motion image taken by the underwater camera, and the noise reduction unit is used to perform denoising through a filtering algorithm; the distortion correction unit is used to perform distortion correction to correct the image distortion problem caused by the camera lens and the underwater environment; the marker point positioning unit is used to use an edge detection algorithm to extract the position information of each marker point in the image.

[0076] The images collected by the underwater camera are first transmitted to the image processing module in real time by the image acquisition unit. To ensure data integrity and transmission efficiency, the image acquisition unit adopts a high-performance data interface and supports multi-threaded concurrent processing to process a large number of high-resolution images in a short time.

[0077] Subsequently, the noise reduction unit performs noise processing on the captured images. Due to the common optical interference in underwater environments and random noise caused by particles in the water, the noise reduction unit removes small noises through Gaussian filtering algorithm.

[0078] For example, when there are random bright spots in the image, Gaussian filtering can effectively smooth these bright spots without affecting the overall image structure. At the same time, median filtering is also used to remove salt and pepper noise. This method ensures the preservation of image edge features by selecting the median value in the pixel neighborhood instead of the central pixel value.

[0079] The image after noise reduction enters the distortion correction unit, which is used to correct the image distortion caused by the camera lens and the underwater environment. The radial distortion of the lens will cause straight lines to bend in the image. The correction steps include:

[0080] Distortion coefficient calibration: The intrinsic matrix and distortion coefficient of the camera are obtained in advance through the chessboard calibration method (for example, ).

[0081] Correction formula:

[0082]

[0083] Where r is the radial distance from the pixel to the center of the image, x and y are the horizontal and vertical coordinates before correction, , The above formula can be used to correct the curved straight line into a straight line.

[0084] After the distortion correction is completed, the marker location unit extracts the marker position in the image through the edge detection algorithm. The specific implementation process includes:

[0085] Grayscale: Convert a color image to a grayscale image to reduce computational complexity.

[0086] Edge Enhancement: Enhance contrast using the CLAHE method.

[0087] Edge detection: The Canny edge detection algorithm is used and completed in two steps:

[0088] Gradient calculation: The gradient size and direction of each pixel are calculated using the Sobel operator.

[0089] Dual Threshold Detection: Set high and low thresholds to keep major edges and connect weak edges.

[0090] Mark point extraction: Use contour detection technology to process the edge detection results and extract the precise coordinates of the marker points.

[0091] The skeleton solution module includes a joint position marking unit, an initial skeleton model construction unit, a skeleton solution unit and a gait analysis unit. The joint position marking unit is used to detect and mark the positions of various joints of the human body in the preprocessed image; the initial skeleton model construction unit is used to construct an initial skeleton model through the three-dimensional coordinates of the marking points; the skeleton solution unit is used to solve the skeleton model based on inverse dynamics and / or forward dynamics to generate three-dimensional gait data; the gait analysis unit is used to perform gait analysis through the three-dimensional gait data generated by the skeleton solution unit, wherein the gait analysis includes the calculation and evaluation of stride length, step frequency, and joint angle changes.

[0092] The joint position marking unit detects and marks the positions of various joints of the human body according to the position information of the marking points, and initializes the human skeleton model.

[0093] Specifically, the joint position marking unit is used to perform the following steps:

[0094] Initialize the human skeleton model according to the position information of the marker points;

[0095] Determine the position and connection relationship of each joint;

[0096] Determine the accuracy of skeleton initialization to avoid solving errors due to marker errors.

[0097] As an example, the algorithm steps of the joint position marking unit include:

[0098] Step 1: Group the marker points according to the human body structure to locate the corresponding joints.

[0099] Algorithm process:

[0100] 1. Mark point data Classify according to preset human joint distribution.

[0101] 2. Use the spatial distribution rules of the markers, such as the distance and relative position of the markers in the waist, thigh, knee, ankle, etc., to divide them into different joint groups:

[0102] .

[0103] 3. If the grouping of some markers is unclear, they are automatically assigned based on spatial distance using the KNN algorithm.

[0104] Step 2: Calculate the joint center point and calculate the position of the corresponding joint based on each set of marker points.

[0105] Algorithm process:

[0106] 1. For each set of marked points , calculate its center of gravity as the joint position:

[0107] .

[0108] 2. The calculated center of gravity Marked as the initial position of the joint.

[0109] The initial skeleton model building unit is used to perform the following steps:

[0110] Step 1: Infer the joint connection relationship, establish the connection relationship between the joints, and construct the initial skeleton.

[0111] Algorithm process:

[0112] 1. Define joint connections based on the topological structure of the human anatomical model, for example: the waist connects to the left and right thigh roots; the thigh roots connect to the knees, and the knees connect to the ankles.

[0113] 2. For each pair of adjacent joints, calculate the spatial distance d and direction vector :

[0114]

[0115] 3. Record the information of each connection line, including the number, direction and distance of the connecting joint.

[0116] Step 2: Initialize the skeleton model and build the initial human skeleton model based on joint positions and connection relationships.

[0117] Algorithm process:

[0118] 1. Utilize all joint positions And connection information to generate the initial skeleton model:

[0119] The graph structure G=(V,E) is constructed with joints as nodes and connection relationships as edges, where V is the joint set and E is the connection set.

[0120] 2. Standardize the parameters of the skeleton model to make it conform to the human body proportion model.

[0121] Step 3: Model verification and adjustment: verify the accuracy of the skeleton model and adjust the deviation caused by marker point errors.

[0122] Verification process:

[0123] 1. Compare the skeleton model with the standard human skeleton template to check the deviation of joint positions: .

[0124] 2. If the deviation exceeds the preset threshold, adjust the joint position:

[0125] Kalman filtering is used to optimize the spatiotemporal consistency of markers.

[0126] Perform local iterative adjustments on joints in abnormal positions and recalculate the connection relationships.

[0127] Specifically, the skeleton solving unit is used to execute the following steps S71-S75.

[0128] S71. Set the target position and orientation of the end effector.

[0129] S72. Starting from the end effector, reversely calculate the rotation angles of the skeleton joints one by one.

[0130] S73. Calculate the transformation matrix of the bone according to the rotation angle.

[0131] S74. Use the transformation matrix to calculate the position and orientation of each joint.

[0132] S75. Improve the accuracy and stability of the solution through iterative optimization.

[0133] In one embodiment, the algorithm steps of the skeleton solving unit include:

[0134] Step 1: Set the target position and orientation of the end effector to clarify the target position and orientation of the end joint (such as toe, heel) and provide initial conditions for inverse dynamics.

[0135] Algorithm process:

[0136] Get the end effector position of the current skeleton model from the joint position marker unit .

[0137] According to the goal of gait analysis, set the target position and orientation of the end effector ,in is the target direction vector.

[0138] Step 2: Inverse dynamics solution, starting from the end effector, reversely calculate the rotation angles of the skeleton joints one by one.

[0139] Algorithm process:

[0140] Initialize the position and orientation of the end effector: , where t is the time step.

[0141] Recursively calculate from the end effector to the root joint:

[0142] For each joint , calculate the adjacent joints Rotation angle :

[0143] ,

[0144] in and are the direction vectors of the current joint and the previous joint respectively.

[0145] Update the transformation matrix

[0146] are the rotation matrices around the z, y, and x axes respectively.

[0147] Iterate until the root joint.

[0148] Step 3: Forward dynamics solution, starting from the root joint, forward calculates the position and direction of each joint.

[0149] Algorithm process:

[0150] Initialize the root joint position and orientation:

[0151] For each joint , according to the transformation matrix of the previous joint and joint length Calculate its position:

[0152] Update the direction vector:

[0153] The position information of each joint is calculated iteratively until the end effector.

[0154] Step 4: Iterate and optimize to improve the accuracy and stability of the solution and correct the deviations caused by initial conditions or calculation errors.

[0155] Algorithm process:

[0156] Define the objective function , represents the error between the target position and the actual position of the end effector:

[0157] Minimize the objective function using an optimization algorithm such as gradient descent or Newton's method:

[0158] is the learning rate.

[0159] The optimization is iterated at each time step until the error meets the preset convergence condition.

[0160] Step 5: The solution results are output, and the position and direction of each joint are displayed in real time for gait analysis.

[0161] Output data:

[0162] The set of positions for each joint:

[0163] A collection of direction vectors for each joint:

[0164] Joint rotation angle collection:

[0165] Step 6: Anomaly detection and adjustment, detect abnormal solution results and make timely adjustments.

[0166] Detection method:

[0167] If the end effector deviates from the target position beyond the allowable range, re-optimize the solution parameters.

[0168] If the bone angle changes beyond the physiological range (such as the knee joint exceeds the normal flexion and extension angle), it will be automatically corrected to a reasonable range.

[0169] The present application also discloses a method for underwater gait analysis of medical rehabilitation based on skeleton solution, including:

[0170] S1. Obtaining an image of a human body moving in water captured by an underwater camera;

[0171] S2. De-noising by filtering algorithm;

[0172] S3. Perform distortion correction to correct image distortion caused by the camera lens and the underwater environment;

[0173] S4. Using edge detection algorithm, extract the position information of each marker point in the image;

[0174] S5. In the preprocessed image, detect and mark the positions of various joints of the human body;

[0175] S6. construct an initial skeleton model through the three-dimensional coordinates of the marking points;

[0176] S7. Solving the skeleton model based on inverse dynamics and / or forward dynamics to generate three-dimensional gait data;

[0177] S8. Perform gait analysis on the three-dimensional gait data generated by the skeleton solving unit, wherein the gait analysis includes calculation and evaluation of stride length, stride frequency, and joint angle changes.

[0178] Optionally, S7 includes the following steps:

[0179] S71. Setting the target position and direction of the end effector;

[0180] S72. Starting from the end effector, reversely calculate the rotation angles of the skeletal joints one by one;

[0181] S73. Calculate the transformation matrix of the skeleton according to the rotation angle;

[0182] S74. Using the transformation matrix, calculate the position and orientation of each joint;

[0183] S75. Improve the accuracy and stability of the solution through iterative optimization.

[0184] This application also discloses a specific implementation process, as shown below:

[0185] Preparation:

[0186] Ensure that all equipment has been correctly installed and debugged, and all parameters have been adjusted to the optimal state.

[0187] Prepare the subject for gait analysis and attach active luminous markers to it, making sure the markers are positioned accurately and firmly.

[0188] Image acquisition:

[0189] Turn on all cameras and light sources and start image acquisition. Capture the subject moving according to the predetermined gait movements, ensuring that the entire process is carried out underwater.

[0190] Monitor the image acquisition process in real time to ensure image quality and stability. If any problems are found, adjust the device parameters or retake the image.

[0191] Data Retention:

[0192] The collected image data is transmitted to the image processing module in real time for preprocessing. At the same time, the original image data is saved for subsequent analysis and processing.

[0193] Image preprocessing:

[0194] In the image processing module, the collected images are denoised to reduce the impact of random noise in the images.

[0195] Perform distortion correction to correct image distortion caused by the camera lens and underwater environment. Ensure that the positions of each marker point in the image are accurate.

[0196] Use edge detection algorithms to extract the location information of each marker point in the image, ensuring that the edges of the marker points are clear and the positions are accurate.

[0197] Skeleton initialization:

[0198] Initialize the human skeleton model based on the position information of the marker points, determine the position and connection relationship of each joint, and prepare for the subsequent skeleton solution.

[0199] Ensure the accuracy of skeleton initialization to avoid solving errors caused by marker errors. Figure 2 , How to stick the points on the lower body: four points on the waist, one on each thigh root, one on the inner and outer sides of the knees, one on the toes and heels, one on the outer side of the little toe, and one on the inner and outer sides of the ankles, a total of 20 points.

[0200] Skeleton solving:

[0201] Enter the skeleton solving module and start solving the gait data. Use the combination of inverse dynamics (ID) and forward dynamics (FK) to solve the skeleton model.

[0202] The ID method starts from the end effector and reversely calculates the rotation angle of the skeleton joints one by one; the FK method starts from the root joint and forward calculates the position and direction of each joint. By combining the two methods, efficient and accurate gait analysis can be achieved.

[0203] Display the solution results in real time, observe the dynamic changes of gait data, and check the accuracy and stability of the solution. If any abnormality is found, adjust the solution parameters or re-solve it in time.

[0204] Gait analysis:

[0205] Based on the solved three-dimensional gait data, a detailed gait analysis is performed, including the calculation and evaluation of parameters such as stride length, stride frequency, and joint angle changes.

[0206] Compare the gait analysis results with standard data to evaluate the subject's gait health status and provide a scientific reference.

[0207] Reference Figure 2 and Figure 3 , export gait analysis reports, including charts and data, to facilitate subsequent research and application.

[0208] The application areas and advantages of this application are as follows:

[0209] Medical rehabilitation field

[0210] In the field of medical rehabilitation, gait analysis is an important diagnostic tool that can help doctors evaluate patients' motor function and rehabilitation progress. Traditional gait analysis methods usually need to be performed under laboratory conditions, with complex equipment and cumbersome operations. The underwater optical motion capture camera system provides a new solution with the following advantages:

[0211] Non-contact capture: Optical capture technology eliminates the need to attach sensors or electrodes to patients, reducing interference and discomfort to patients. This is especially useful for patients who are unable to move or are sensitive to traditional sensors.

[0212] High-precision analysis: Using advanced image processing and bone solution algorithms, high-precision gait data capture and analysis can be achieved, which can accurately evaluate the patient's gait characteristics and abnormalities. It helps doctors develop personalized rehabilitation treatment plans.

[0213] Real-time feedback: The system supports real-time data processing and feedback. Doctors can instantly observe the patient's gait changes, make real-time adjustments and interventions, and improve the effectiveness and efficiency of rehabilitation treatment.

[0214] Long-term monitoring: It is suitable for continuous monitoring during long-term rehabilitation, recording the changing trend of the patient's gait and providing a scientific basis for the evaluation of rehabilitation progress. At the same time, it can reduce the workload of doctors and improve the efficiency of diagnosis and treatment.

[0215] Sports training field

[0216] In sports training, gait analysis can help coaches understand athletes' movement techniques and provide scientific training guidance. Traditional training methods often rely on the coach's experience and visual observation, which has certain limitations. The advantages of underwater optical motion capture camera systems in sports training include:

[0217] Accurate movement assessment: Through high-precision gait data capture and analysis, the athlete's movement technique is quantified, helping coaches identify problems and deficiencies in the movement and providing targeted improvement suggestions. For example, swimmers can optimize their swimming style and movement coordination by analyzing underwater gait data.

[0218] Real-time feedback and guidance: The system supports real-time data processing and feedback. Coaches can instantly obtain athletes' gait information during training, provide real-time guidance and correction, and improve training effectiveness and efficiency. At the same time, athletes can also adjust themselves and improve their movement techniques through real-time feedback.

[0219] Training effect evaluation: By continuously monitoring the changes in the athlete's gait, the training effect and progress can be evaluated. By comparing the data at different stages, the athlete's progress and technical improvement can be understood, providing a scientific basis for subsequent training plans.

[0220] Injury prevention and rehabilitation: Through gait analysis, potential injury risks of athletes can be discovered in advance, and preventive measures can be taken to avoid injuries. For athletes who have already been injured, gait analysis can evaluate the progress of rehabilitation, guide scientific rehabilitation training, and speed up their return to the game.

[0221] Animation production field

[0222] In animation production, gait analysis is an important part of character animation. Realistic character gait movements can significantly improve the visual effects of animation. Traditional animation production methods often rely on manual adjustments by animators, which is labor-intensive and inefficient. The advantages of underwater optical motion capture camera systems in animation production include:

[0223] Efficiently capture real gait: Through optical capture technology, the gait movements of real people can be efficiently captured and converted into 3D animation data. Animators can perform secondary creation based on this to improve the realism and naturalness of the animation. For example, when making character animations for underwater scenes, the system can be used to capture real underwater gait movements to make the character movements more realistic.

[0224] Fine motion adjustment: The system supports real-time data processing and feedback, and animators can instantly observe the dynamic changes of the character's gait and make fine motion adjustments and optimizations. For example, parameters such as joint angles, speeds, and accelerations can be adjusted to make the character's movements smoother and more natural.

[0225] Improve production efficiency: Through automated gait capture and data processing, the manual adjustment work of animators is reduced, and the efficiency and output of animation production are improved. At the same time, in the case of multi-person collaboration, data and progress can be shared in real time to improve the effect of team collaboration.

[0226] Diverse application scenarios: Applicable to various animation production scenarios, especially animation works that require fine gait performance. For example, in scenes that require high realism, such as sports competitions, dance performances, and war scenes, gait analysis can significantly improve the visual effects and viewing experience of animations.

[0227] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A medical rehabilitation underwater gait analysis system based on skeleton solution, characterized in that: include: Multiple underwater cameras for capturing images of human motion in water; Light source equipment, installed underwater, used to illuminate the reflective marker ball on the human body; Reflective marker balls are used to be installed on pre-set marking points on the human body to reflect light; An image processing module is used to obtain the image of human body motion in water taken by an underwater camera, and to perform image distortion correction and position detection of marking points; The skeleton solving module is used to construct skeleton models and solve skeletons based on the position changes of marker points for gait analysis.

2. The medical rehabilitation underwater gait analysis system based on skeleton solution according to claim 1 is characterized in that: The reflective marker ball comprises: A plastic shell, the plastic shell comprising two hemispherical transparent plastic shells, the two hemispherical plastic shells being able to be combined into a closed sphere; Reflective stickers are used to stick to the inner side of the plastic shell.

3. The medical rehabilitation underwater gait analysis system based on skeleton solution according to claim 2 is characterized in that: The image processing module comprises: An image acquisition unit, used to acquire an image of a human body in water captured by an underwater camera; A noise reduction unit, used for performing noise reduction processing through a filtering algorithm; A distortion correction unit, used for performing distortion correction to correct image distortion problems caused by the camera lens and the underwater environment; The marker point positioning unit is used to extract the position information of each marker point in the image using an edge detection algorithm.

4. The medical rehabilitation underwater gait analysis system based on skeleton solution according to claim 3 is characterized in that: The skeleton solving module includes: A joint position marking unit, used to detect and mark the positions of various joints of the human body in the preprocessed image; An initial skeleton model building unit, used to build an initial skeleton model through the three-dimensional coordinates of the marking points; A skeleton solving unit, used for solving the skeleton model based on inverse dynamics and / or forward dynamics to generate three-dimensional gait data; The gait analysis unit is used to perform gait analysis on the three-dimensional gait data generated by the skeleton solving unit, wherein the gait analysis includes the calculation and evaluation of the stride, stride frequency and joint angle changes.

5. The medical rehabilitation underwater gait analysis system based on skeleton solution according to claim 4 is characterized in that: The skeleton solving unit is used to perform the following steps: Set the target position and orientation of the end effector; Starting from the end effector, reversely calculate the rotation angles of the skeleton joints one by one; According to the rotation angle, calculate the transformation matrix of the bone; Using the transformation matrix, calculate the position and orientation of each joint; Improve the accuracy and stability of the solution through iterative optimization.

6. The medical rehabilitation underwater gait analysis system based on skeleton solution according to claim 5 is characterized in that: The joint position marking unit is used to perform the following steps: Initialize the human skeleton model according to the position information of the marker points; Determine the position and connection relationship of each joint; Determine the accuracy of skeleton initialization to avoid solving errors due to marker errors.

7. The medical rehabilitation underwater gait analysis system based on skeleton solution according to claim 6 is characterized in that: The reflective marker ball is attached to four points on the waist, one at each of the left and right thigh roots, one at each of the inner and outer sides of the knees, one at each of the toes and heels, one at each of the outer sides of the little toes, and one at each of the inner and outer sides of the ankles.

8. A method for underwater gait analysis in medical rehabilitation based on skeleton solution, characterized in that: include: S1. Obtaining an image of a human body moving in water captured by an underwater camera; S2. De-noising by filtering algorithm; S3. Perform distortion correction to correct image distortion caused by the camera lens and the underwater environment; S4. Using edge detection algorithm, extract the position information of each marker point in the image; S5. In the preprocessed image, detect and mark the positions of various joints of the human body; S6. construct an initial skeleton model through the three-dimensional coordinates of the marking points; S7. Solving the skeleton model based on inverse dynamics and / or forward dynamics to generate three-dimensional gait data; S8. Perform gait analysis on the three-dimensional gait data generated by the skeleton solving unit, wherein the gait analysis includes calculation and evaluation of stride length, stride frequency, and joint angle changes.

9. The method for underwater gait analysis for medical rehabilitation based on skeleton calculation according to claim 8, characterized in that: The S7 comprises the following steps: S71. Setting the target position and direction of the end effector; S72. Starting from the end effector, reversely calculate the rotation angles of the skeletal joints one by one; S73. Calculate the transformation matrix of the skeleton according to the rotation angle; S74. Using the transformation matrix, calculate the position and orientation of each joint; S75. Improve the accuracy and stability of the solution through iterative optimization.

Images