Real-time monitoring method for human body rehabilitation training of rehabilitation equipment

Abstract

The invention belongs to the technical field of image processing, and particularly relates to a real-time monitoring method for human body rehabilitation training of rehabilitation equipment, which comprises the following steps of: calculating a local motion energy factor according to the distribution condition of instantaneous speed and instantaneous acceleration in a local track section at each time point; calculating a root-mean-square error of a local ideal trajectory and a local trajectory segment to obtain local trajectory stability, combining the local ideal trajectory and the local trajectory segment with a local average speed to calculate a dynamic detail significance score for adjusting a global basic threshold value to obtain an adaptive distance threshold value, and processing an original motion trajectory sequence by applying a Douglas-Peucker algorithm to obtain an adaptive distance threshold value; and obtaining an optimized track sequence, and calculating the training stability of the patient according to a root-mean-square error between the ideal motion track sequence and the optimized track sequence to generate a visual rehabilitation report. According to the method, random noise is effectively filtered out, the motion trail is smoothed, and meanwhile pathological detail features influencing rehabilitation evaluation are reserved.

Description

Technical Field

[0001] This invention relates to the field of image processing technology. More specifically, this invention relates to a real-time monitoring method for human rehabilitation training using rehabilitation equipment. Background Technology

[0002] In modern rehabilitation medicine, using rehabilitation equipment to assist patients in conducting scientific and standardized rehabilitation training is an important way to restore physical function. In order to objectively evaluate the rehabilitation effect and guide subsequent training plans, it is necessary to monitor and quantify the patient's performance during the training process in real time.

[0003] Transforming complex and continuous human dynamic data into quantifiable and traceable key performance indicators, such as movement completion, stability, and fluency, is the core of achieving accurate rehabilitation assessment. Currently, a mainstream technical solution involves capturing video streams of patients training with rehabilitation equipment using image acquisition devices such as cameras, extracting key joint points of the human body using posture estimation algorithms in machine vision, and analyzing the data to calculate rehabilitation assessment indicators.

[0004] However, in real environments such as homes or non-professional rehabilitation rooms, factors such as changes in lighting, background interference, and loose clothing of patients inevitably cause the joint coordinate sequence output by the posture estimation algorithm to contain a lot of noise, jitter, and abnormal jumping. If these noisy raw data are used directly to calculate rehabilitation indicators, the reliability of the results will be seriously affected, and the system may misinterpret the technical error of the algorithm itself as the patient's physiological performance.

[0005] To address the aforementioned data noise problem, introducing trajectory simplification algorithms to preprocess the key point coordinate sequences has become a common improvement approach. Among these, the Douglas-Peucker algorithm (DP) is widely used due to its efficiency and shape preservation. The Douglas-Peucker algorithm simplifies and smooths the trajectory by setting a distance threshold and removing points that deviate slightly from the main path, thereby filtering out some noise.

[0006] The traditional Douglas-Puk algorithm uses a globally fixed distance threshold, which cannot distinguish the nature of trajectory details. When filtering out visual noise, it will mistakenly remove microscopic trajectory details with diagnostic value, losing the core information necessary for refined and differentiated rehabilitation assessment, thus failing to accurately quantify the patient's true motion state. Summary of the Invention

[0007] To address the technical problem of existing Douglas-Puk algorithms erroneously removing diagnostically valuable microscopic details of trajectories when filtering visual noise, this invention provides a real-time monitoring method for human rehabilitation training in rehabilitation equipment. The method includes: acquiring monitoring videos of hand patients mimicking hand recovery test videos; extracting multiple joint points from each frame using a human pose estimation algorithm, designating any joint point as a target joint point, combining the spatial coordinates of the target joint point across multiple frames into an original motion trajectory sequence, and calculating the instantaneous velocity and acceleration vectors of the target joint point at each time point; calculating the local motion energy factor based on the distribution of instantaneous velocity and acceleration within the local trajectory segment at each time point; and applying the least squares method to the local motion energy vector. Local path fitting is performed on the trajectory segments, and the root mean square error between the obtained ideal local trajectory and the local trajectory segments is calculated to obtain the local trajectory stability. Based on the local motion energy factor, local trajectory stability index, and local average velocity, the dynamic detail significance score is calculated, and the global base threshold is adjusted to obtain the adaptive distance threshold used to determine whether the coordinates of the target joint at each time point are retained. The Douglas-Puk algorithm is applied to process the original motion trajectory sequence to obtain the optimized trajectory sequence. The least squares method is used to fit the spatial coordinates of the target joint in the hand recovery test video. Based on the root mean square error between the obtained ideal motion trajectory sequence and the optimized trajectory sequence, the patient's training stability is calculated to generate a visualized rehabilitation report.

[0008] This invention constructs a dynamic detail saliency scoring index by comprehensively analyzing the energy, stability, and speed of local movements, and dynamically adjusts the threshold of the filtering algorithm accordingly. This allows the filtering algorithm to effectively filter out random noise and smooth the movement trajectory while accurately preserving the pathological details that are crucial for rehabilitation assessment. As a result, the data quality on which the final rehabilitation assessment report is based is greatly improved, and the assessment results more accurately and realistically reflect the patient's neuromuscular control ability.

[0009] Preferably, the target joint points include, but are not limited to, the right shoulder, right elbow, and right wrist. When the patient is performing rehabilitation exercises with their left hand, the target joint points include, but are not limited to, the left shoulder, left elbow, and left wrist.

[0010] Preferably, the method for obtaining the local trajectory segments at each time point is as follows: for each time point... Obtain its local time window Local time window Includes from a point in time At the appointed time common At which point in time, For local time windows Half width; based on the local time window From the original motion trajectory sequence of the target joint Extract the corresponding local trajectory segment from the middle .

[0011] Preferably, the step of calculating the local motion energy factor based on the distribution of instantaneous velocity and instantaneous acceleration within the local trajectory segment at each time point includes: In the formula, For the target key point at time point The local motion energy factor of the local trajectory segment; , The target key points at time points are respectively The instantaneous velocity vector and instantaneous acceleration vector; Indicates the magnitude of the vector; For local trajectory segments All time points The standard deviation of the magnitude of the instantaneous acceleration vector; For local trajectory segments All time points The mean of the magnitude of the instantaneous acceleration vector.

[0012] This invention combines the amplitude of instantaneous velocity with the dispersion of instantaneous acceleration, which can effectively amplify signals with regular high-frequency oscillations, such as physiological tremors. From the perspective of physical energy, it can initially distinguish meaningful tremors from irregular and isolated noise points, providing a basis for subsequent significance judgment.

[0013] Preferably, the calculation of the root mean square error between the local ideal trajectory and the local trajectory segment to obtain the local trajectory stability includes: calculating the local trajectory segment... From each coordinate point in the local ideal trajectory The Euclidean distances of the corresponding projection points are calculated, and the root mean square of all Euclidean distances is taken as the root mean square error between the local ideal trajectory and the local trajectory segment. The root mean square error is then inversely normalized, and the result is used as the target joint point at time point. Local trajectory stability The inverse proportional normalization is achieved through... To achieve, among which, It is a natural exponential function.

[0014] This invention obtains trajectory stability by comparing the actual local trajectory with an ideal smooth fitted trajectory. It can identify the persistent small-scale trajectory roughness caused by muscle instability. Even if the energy of these fluctuations is not strong, they can be identified as unstable. This provides supplementary information for distinguishing trajectory details of different causes, namely, high-energy tremors and geometric instability.

[0015] Preferably, the step of calculating the dynamic detail saliency score based on the local motion energy factor, local trajectory stability index, and local average velocity includes: In the formula, For the target key point at time point Dynamic detail saliency score; For the target key point at time point The local motion energy factor of the local trajectory segment; For the target key point at time point Local trajectory stability; For the target key point at time point The instantaneous velocity vector; Indicates the magnitude of the vector; For local trajectory segments All time points The mean of the magnitude of the instantaneous velocity vector, representing the time point. The local average velocity at that location.

[0016] This invention combines an energy factor representing the intensity of motion with a stability factor representing geometric smoothness, and uses average speed as a contextual adjustment factor. This enables the scoring model to make complex decisions: tolerating greater fluctuations in high-speed motion, while being more sensitive to minute jitters in low-speed fine control, thereby achieving intelligent and contextualized judgment of the importance of trajectory details.

[0017] Preferably, the step of adjusting the global base threshold to obtain the adaptive distance threshold used to determine whether the coordinates of the target joint at each time point are retained includes: for the target joint at each time point... Coordinates of the corresponding frame image : Calculate the relationship between the number 1 and the target key point at time points Dynamic detail saliency score The difference Calculate the product of the difference and the global basic threshold to obtain the judgment coordinate point. The adaptive distance threshold used when deciding whether to retain a value.

[0018] Preferably, the process of applying the Douglas-Puk algorithm to process the original motion trajectory sequence to obtain an optimized trajectory sequence includes: determining whether to retain each coordinate point by using an adaptive distance threshold; and for each coordinate point... : Determining whether to retain coordinate points When calculating coordinate points To the key points before and after and perpendicular distance of the connecting line segments and coordinate points Corresponding adaptive distance threshold If a comparison is made, Then retain the coordinates. Otherwise, remove the coordinate point; repeat the judgment process until the original motion trajectory sequence is obtained. After all coordinate points in the data have been processed, the optimized trajectory sequence is obtained. Initially, the original motion trajectory sequence is... The first and last coordinate points are used as key points. and The key point that follows is the retained coordinates.

[0019] Preferably, calculating the patient's training stability based on the root mean square error between the obtained ideal motion trajectory sequence and the optimized trajectory sequence includes: for the target joint: calculating the optimized trajectory sequence The root mean square error (RMSE) of the ideal motion trajectory sequence is compared with the root mean square error (RMSE). The RMSE is then inversely normalized, and the result is used as the training stability of the target joint. This inverse normalization is achieved through... To achieve, among which, The natural exponential function is used to calculate the mean of training stability across all joints, which is then used as the patient's training stability.

[0020] The optimized trajectory sequence obtained by this invention not only filters out interfering noise but also retains the true physiological tremors and other manifestations. Therefore, its error with the ideal trajectory can more realistically reflect the patient's actual motor control level. Compared with the evaluation results based on the original noisy data or oversimplified data, the stability index generated by this invention is less susceptible to algorithm artifacts, thereby improving the reliability of the final rehabilitation evaluation index.

[0021] Preferably, the calculation of the instantaneous velocity vector and instantaneous acceleration vector of the target joint at each time point includes: instantaneous velocity vector , , The target key points at time points are respectively , The corresponding coordinates in the frame image; The time interval for acquiring frame images; instantaneous acceleration vector. , , The target key points at time points are respectively , The instantaneous velocity vector.

[0022] The beneficial effects of this invention are as follows: This invention constructs a dynamic detail saliency scoring index by comprehensively analyzing the energy, stability, and speed of local movements, and dynamically adjusts the threshold of the filtering algorithm accordingly. This allows the filtering algorithm to effectively filter out random noise and smooth the movement trajectory while accurately preserving the pathological details that are crucial for rehabilitation assessment. As a result, the data quality on which the final rehabilitation assessment report is based is greatly improved, and the assessment results more accurately and realistically reflect the patient's neuromuscular control ability. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating a real-time monitoring method for human rehabilitation training using rehabilitation equipment according to the present invention; Figure 2 This is a flowchart illustrating step S2; Figure 3 It is a schematic diagram illustrating the curves of the original motion trajectory sequence and the ideal motion trajectory sequence; Figure 4 This schematically illustrates the application of a standard dynamic programming (DP) algorithm based on a global baseline threshold. Figure 3 The result of processing the original motion trajectory sequence; Figure 5 This schematically illustrates the adaptive distance threshold and the adaptive P algorithm for... Figure 3 The result of processing the original motion trajectory sequence. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0026] This invention discloses a real-time monitoring method for human rehabilitation training using rehabilitation equipment, referring to... Figure 1 This includes steps S1 to S4: S1: Collect monitoring videos of patients with hand diseases imitating the movements in the hand recovery test video, use human pose estimation algorithm to extract target joints in each frame of the image, and combine the spatial coordinates of the target joints in multiple consecutive frames of the image to form the original motion trajectory sequence of the target joints.

[0027] It should be noted that during the rehabilitation process of patients with hand diseases, continuous hand exercises are required using specialized hand rehabilitation equipment. To better facilitate hand rehabilitation exercises, patients are asked to watch a hand recovery test video after each exercise session and try their best to imitate the movements shown. By monitoring the patient's hand movements and analyzing the differences between the monitoring video and the hand recovery test video, a better understanding of the patient's hand rehabilitation progress can be achieved.

[0028] Specifically, hand rehabilitation exercises are performed on patients with hand diseases using medical and nursing hand rehabilitation exercise equipment. After each hand rehabilitation exercise, patients are asked to watch a hand recovery test video and try their best to imitate the movements in the video. During this process, the image acquisition unit that is matched with the rehabilitation exercise equipment captures the video of the patient's rehabilitation training in real time at a preset frame rate.

[0029] Furthermore, for consecutive image frames in the rehabilitation training video, multiple joint points and their spatial coordinates in each frame are extracted using existing human pose estimation algorithms; for the target joint points: the target joint points are located at time points... The coordinates of the corresponding frame image are denoted as ,in, , The target key points at time points are respectively The horizontal and vertical coordinates of the corresponding frame image.

[0030] Among them, existing human pose estimation algorithms include, but are not limited to, OpenPose and MediaPipe; when the patient's right hand is used for rehabilitation exercises, the joint points include, but are not limited to, the right shoulder, right elbow, and right wrist; when the patient's left hand is used for rehabilitation exercises, the joint points include, but are not limited to, the left shoulder, left elbow, and left wrist.

[0031] Furthermore, the spatial coordinates of the target joints in all frame images are combined to form an original motion trajectory sequence of the target joints containing multiple data points. .

[0032] It should be noted that this provides the raw data input for subsequent trajectory detail analysis and adaptive simplification, and its output... Although it contained noise, it also completely recorded all the original motion information, laying the foundation for distinguishing noise from effective details later.

[0033] S2: Based on the distribution of instantaneous velocity and instantaneous acceleration within the local trajectory segment at each time point, calculate the local motion energy factor, obtain the root mean square error of the local ideal trajectory and the local trajectory segment to obtain the local trajectory stability, and combine the two with the local average velocity to calculate the dynamic detail significance score.

[0034] The flowchart for step S2 is shown below. Figure 2 The process includes steps S201 to S204, specifically as follows: S201: Calculate the instantaneous velocity vector and instantaneous acceleration vector of the target joint at each time point.

[0035] To calculate the local motion energy factor, it is necessary to obtain the discrete original motion trajectory sequence. The fundamental physical quantities describing the dynamic characteristics of motion are extracted: instantaneous velocity vector and instantaneous acceleration vector; specifically, based on the difference approximation of the target joint's coordinates in time, the target joint's position at time point is calculated. instantaneous velocity vector and instantaneous acceleration vector And instantaneous velocity vector , , The target key points at time points are respectively , The corresponding coordinates in the frame image; The instantaneous acceleration vector is the time interval for acquiring frames of images. , , The target key points at time points are respectively , The instantaneous velocity vector; This is the time interval for acquiring frame images.

[0036] S202: Calculate the local motion energy factor of the target joint at each time point based on the instantaneous velocity and instantaneous acceleration within the local trajectory segment at each time point.

[0037] It should be noted that real physiological tremors or shaking caused by muscle fatigue, although with small displacement, will exhibit a regular high-frequency oscillation in velocity and acceleration, with concentrated energy consumption; while the random noise generated by the algorithm tends to have irregular isolated spikes in velocity and acceleration. Therefore, by comprehensively considering the amplitude of velocity changes and the concentration of acceleration changes, the trajectory details of these two different types can be distinguished.

[0038] Specifically, regarding the time point Obtain its local time window Local time window Includes from a point in time At the appointed time common At which point in time, For local time windows half the width, and The value range is [2, 6]. This invention will... Set to 3; based on local time window From the original motion trajectory sequence of the target joint Extract the corresponding local trajectory segment from the middle .

[0039] Furthermore, based on the time point Local trajectory segment at the location Instantaneous velocity and instantaneous acceleration within the time frame are used to calculate the target joint at time point. Local motion energy factor of local trajectory segment The specific calculation formula is as follows: ; In the formula, For the target key point at time point The local motion energy factor of the local trajectory segment; For local time windows Half the width; , The target key points at time points are respectively The instantaneous velocity vector and instantaneous acceleration vector; Indicates the magnitude of the vector; This represents the target key point at a given time. The instantaneous kinetic energy reflects the intensity of the motion; For local trajectory segments All time points The standard deviation of the magnitude of the instantaneous acceleration vector is used to measure the degree of dispersion of acceleration changes; For local trajectory segments All time points The mean of the magnitude of the instantaneous acceleration vector is used to measure the overall acceleration level.

[0040] Among them, the formula for calculating the local motion energy factor includes the kinetic energy term. Modulation terms composed of the statistical characteristics of acceleration When the trajectory segment exhibits regular, high-frequency tremors, such as physiological tremors, the acceleration will oscillate violently around a non-zero mean, leading to... Larger, and Maintaining a certain level, at which point the ratio of the two is... It will increase significantly, thereby amplifying the local motion energy factor. The value of ; conversely, for isolated noise points in smooth motion, although they may produce a single large acceleration, the statistical average effect within the local time window is not significant, and therefore does not affect the overall smooth motion. and Both are relatively small, and the ratio between the two is relatively small. There will be no abnormal increase; therefore, the local motion energy factor The higher the value, the more likely the local trajectory segment is to contain meaningful physiological details rather than pure algorithmic noise.

[0041] It should be noted that the calculated local motion energy factor provides a basis for judgment in the subsequent adaptive algorithm, enabling the identification of potentially important details in the trajectory based on kinematic principles.

[0042] S203: Local path fitting is performed on the local trajectory segment using the least squares method to obtain the local ideal trajectory of the target joint at each time point. The root mean square distance between the local trajectory segment and the local ideal trajectory is calculated to obtain the local trajectory stability of the target joint at each time point.

[0043] It should be noted that, in addition to tremor, rehabilitation assessment also focuses on the stability of the patient's movements, that is, the degree to which the actual movement trajectory deviates from the ideal smooth path. Therefore, a local trajectory stability index independent of tremor intensity is constructed to quantify the local geometric smoothness of the trajectory.

[0044] Specifically, the least squares method is used to analyze local trajectory segments. All coordinate points Perform local path fitting to obtain a representation of the local time window. A smooth parabolic segment of the local ideal path within the time frame is used as the target key point at time point. Local ideal trajectory .

[0045] Furthermore, calculate the local trajectory segment. From each coordinate point in the local ideal trajectory The Euclidean distances of the corresponding projection points are calculated, and the root mean square of all Euclidean distances is taken as the root mean square error between the local ideal trajectory and the local trajectory segment. The root mean square error is then inversely normalized, and the result is used as the target joint point at time point. Local trajectory stability The inverse proportional normalization is achieved through... To achieve, among which, It is a natural exponential function.

[0046] It should be noted that the obtained local trajectory stability quantifies the trajectory details from a geometric perspective, characterizing the local roughness of the trajectory. Lower local trajectory stability indicates that the local trajectory segment... This could stem from irregular and unstable muscle control, or from significant noise generated by the algorithm, providing information about the deviation of the path morphology for subsequent decisions.

[0047] S204: Calculate the dynamic detail significance score of the target joint at each time point based on the local motion energy factor, local trajectory stability index, and local average velocity.

[0048] It should be noted that whether the details in the trajectory are important and worth preserving depends not only on their own physical characteristics, such as tremor intensity, but also on the motion context in which they occur. Therefore, by integrating the local motion energy factor and local trajectory stability obtained in the preceding steps, and introducing motion speed as a context adjustment factor, a comprehensive detail saliency score is generated.

[0049] Specifically, based on the local motion energy factor, local trajectory stability index, and local average velocity, the target joint at time point is calculated. Dynamic detail saliency score The specific calculation formula is as follows: ; In the formula, For the target key point at time point Dynamic detail saliency score; For the target key point at time point The local motion energy factor of the local trajectory segment; For the target key point at time point Local trajectory stability; For the target key point at time point The instantaneous velocity vector; Indicates the magnitude of the vector; For local trajectory segments All time points The mean of the magnitude of the instantaneous velocity vector, representing the time point. The local average velocity at that location.

[0050] Among them, the local motion energy factor With terms characterizing path stability Multiplication combines the intrinsic physical properties of details with the macroscopic context of motion, and generates a local average velocity. Normalization is performed: the more unstable the path, The smaller the value, The closer a term is to 1, the smaller its impact on the energy factor; when the path is more stable, The larger the value, The closer the term is to 0, the more it suppresses the score value; at the same time, the velocity term in the denominator plays a role in context adjustment. For low-speed motion, the denominator is smaller, which amplifies the dynamic characteristics in the numerator.

[0051] S3: Adjust the global base threshold based on the dynamic detail saliency score to obtain the adaptive distance threshold used to determine whether the coordinates of the target key point at each time point are retained; use the adaptive distance threshold to process the original motion trajectory sequence using the Douglas-Puk algorithm to obtain the optimized trajectory sequence.

[0052] It should be noted that, in order for the Douglas-Puk algorithm to differentiate between trajectory points of different natures, the globally fixed comparison threshold is replaced with a locally adaptive threshold that dynamically changes according to the characteristics of each trajectory point.

[0053] Specifically, the global base threshold is adjusted based on the dynamic detail saliency score to obtain the adaptive distance threshold used to determine whether the coordinates of the target key point at each time point are retained. The specific calculation formula is as follows: ; In the formula, To determine the coordinates of the point Whether to retain the adaptive distance threshold; coordinate points For the target key point at time point The corresponding coordinates in the frame image; The global baseline threshold represents the maximum simplification of the expected trajectory or the baseline noise level without any special details. Therefore, the global baseline threshold... The value range is [2, 10]. This invention will... Set to 3; For the target key point at time point The dynamic detail saliency score.

[0054] in, When the value approaches 1, it indicates that there are important details at this point. Approaching 0, a very small distance threshold means that only when the coordinates point It landed almost precisely and Only when the lines connecting these points are true or false can they be removed, which makes it highly likely that trajectory points containing small oscillations will be preserved; conversely, when... When the value approaches 0, it indicates that the motion is smooth. Approaching the global base threshold The algorithm is simplified in a conventional way, effectively filtering out noise, but coordinate points containing details are difficult to remove.

[0055] It should be noted that the core improvement to the standard Douglas-Puk algorithm is achieved by applying the dynamic detail saliency score calculated in the previous step to the decision-making process of the Douglas-Puk algorithm, constructing an adaptive distance threshold, and incorporating the expectation of preserving details into each step of the algorithm's decision, making the algorithm's simplification behavior more targeted; finally, by applying the adaptive threshold to the standard Douglas-Puk algorithm process, the adaptation of the original trajectory can be completed, resulting in an optimized result that combines purity and information richness.

[0056] Specifically, the Douglas-Puk algorithm is applied to the original motion trajectory sequence of the target joints. The process involves using an adaptive distance threshold to determine whether to retain each coordinate point. : Determining whether to retain coordinate points When calculating coordinate points To the key points before and after and perpendicular distance of the connecting line segments and coordinate points Corresponding adaptive distance threshold If a comparison is made, Then retain the coordinates. Otherwise, remove the coordinate point; repeat the judgment process until the original motion trajectory sequence is obtained. After processing all coordinate points, an optimized trajectory sequence that is both pure and retains key details is obtained. .

[0057] Initially, the original motion trajectory sequence is... The first and last coordinate points are used as key points. and The key point that follows is the retained coordinates.

[0058] It should be noted that the obtained optimized trajectory sequence provides a high-quality data foundation for subsequent accurate rehabilitation assessment, achieving smoothing and noise reduction on a macro level and preserving valuable motion details on a micro level.

[0059] S4: The path of the target joint in the hand recovery test video is fitted by the least squares method. The root mean square error between the obtained ideal motion trajectory sequence and the optimized trajectory sequence is calculated as the training stability and used to generate a visualized rehabilitation report.

[0060] It should be noted that this step is the final application stage of the method of the present invention. Its purpose is to use the optimized trajectory data obtained in the preceding steps to generate a quantitative assessment report that has practical guiding significance for clinical practice, thereby realizing the closed loop of the entire monitoring method.

[0061] First, using existing human pose estimation algorithms, the spatial coordinates of the target joints in each frame of the hand recovery test video are extracted from consecutive test image frames. Then, using the least squares method, the spatial coordinates of the target joints in all test image frames are fitted to obtain a smooth parabolic segment representing the ideal path, which serves as the ideal motion trajectory sequence of the target joints.

[0062] Then, for any target key point: calculate the optimized trajectory sequence. The root mean square error of the ideal motion trajectory sequence is compared with the root mean square error. This root mean square error is then inversely normalized, and the result is used as the training stability of the target joint. The mean training stability of all joints is calculated and used as the patient's training stability. The inverse normalization is achieved through... To achieve, among which, It is a natural exponential function.

[0063] Among them, due to the optimization of trajectory sequence The study preserves the authentic physiological tremor, and the obtained training stability index can more accurately reflect the patient's neuromuscular control ability without being contaminated by algorithmic noise.

[0064] Finally, the training stability achieved in each rehabilitation training session is linked and stored with patient information, training time, etc., and a visualized rehabilitation report is generated.

[0065] For example, the original motion trajectory sequence And a schematic diagram of the curve of the ideal motion trajectory sequence, as shown below. Figure 3 As shown; for Figure 3 Middle original motion trajectory sequence Based on the global baseline threshold, the processing result using the standard DP algorithm is as follows: Figure 4 As shown, all fluctuations are crudely filtered out. Although the overall trajectory becomes cleaner, pathological tremors with diagnostic value are completely erased. Furthermore, due to the large global baseline threshold, some macroscopic path deviation characteristics of the trajectory, i.e., instability, are also oversimplified. Figure 3 Middle original motion trajectory sequence Based on the adaptive distance threshold, the processing result of the adaptive DP algorithm is as follows: Figure 5 As shown, most of the random noise was also filtered out, making the smooth part of the trajectory clean. Most importantly, in the second half of the trajectory, due to the lower adaptive distance threshold in this area, the details of pathological tremors were accurately preserved. At the same time, the macroscopic path deviation features of the trajectory were also better preserved.

[0066] In summary, the adaptive DP algorithm of this invention performs better than the standard DP algorithm in handling complex rehabilitation patient movement trajectories. It achieves the ability to filter out technical artifacts, i.e. noise, while identifying and retaining pathological details (such as tremors) that are crucial to rehabilitation assessment, thus providing a high-quality data foundation for the subsequent generation of accurate and reliable rehabilitation assessment indicators.

Claims

1. A real-time monitoring method for human rehabilitation training using rehabilitation equipment, characterized in that, include: Collect monitoring videos of patients with hand diseases imitating the movements in hand recovery test videos; A human pose estimation algorithm is used to extract multiple joints in each frame of the image. Any joint is taken as the target joint. The spatial coordinates of the target joint in multiple frames of the image are combined into the original motion trajectory sequence. The instantaneous velocity vector and instantaneous acceleration vector of the target joint at each time point are calculated. Based on the distribution of instantaneous velocity and instantaneous acceleration within the local trajectory segments at each time point, the local motion energy factor is calculated. The local trajectory segments are then fitted using the least squares method to obtain the root mean square error between the obtained local ideal trajectory and the local trajectory segments, thus determining the local trajectory stability. Based on the local motion energy factor, local trajectory stability index, and local average velocity, a dynamic detail saliency score is calculated. The global base threshold is adjusted to obtain the adaptive distance threshold used to determine whether the coordinates of the target joint at each time point are retained. Finally, the Douglas-Puk algorithm is applied to process the original motion trajectory sequence to obtain an optimized trajectory sequence. The path of the target joint in the hand recovery test video is fitted by least squares method. The root mean square error of the obtained ideal motion trajectory sequence and the optimized trajectory sequence is used to calculate the patient's training stability and generate a visualized rehabilitation report.

2. The real-time monitoring method for human rehabilitation training using rehabilitation equipment according to claim 1, characterized in that, The target joints include, but are not limited to, the right shoulder, right elbow, and right wrist. When the patient is performing rehabilitation exercises with their left hand, the target joints include, but are not limited to, the left shoulder, left elbow, and left wrist.

3. The real-time monitoring method for human rehabilitation training using rehabilitation equipment according to claim 1, characterized in that, The method for obtaining the local trajectory segments at each time point is as follows: For time points Obtain its local time window Local time window Includes from a point in time At the appointed time common At which point in time, For local time windows Half width; based on the local time window From the original motion trajectory sequence of the target joint Extract the corresponding local trajectory segment from the middle .

4. The real-time monitoring method for human rehabilitation training using rehabilitation equipment according to claim 1, characterized in that, The calculation of the local motion energy factor based on the distribution of instantaneous velocity and instantaneous acceleration within the local trajectory segment at each time point includes: ； In the formula, For the target key point at time point The local motion energy factor of the local trajectory segment; For local time windows Half the width; , The target key points at time points are respectively The instantaneous velocity vector and instantaneous acceleration vector; Indicates the magnitude of the vector; For local trajectory segments All time points The standard deviation of the magnitude of the instantaneous acceleration vector; For local trajectory segments All time points The mean of the magnitude of the instantaneous acceleration vector.

5. A real-time monitoring method for human rehabilitation training using rehabilitation equipment according to claim 1, characterized in that, The calculated root mean square error between the local ideal trajectory and the local trajectory segment is used to obtain the local trajectory stability, including: Calculate local trajectory segments From each coordinate point in the local ideal trajectory The Euclidean distances of the corresponding projection points are calculated, and the root mean square of all Euclidean distances is taken as the root mean square error between the local ideal trajectory and the local trajectory segment. The root mean square error is then inversely normalized, and the result is used as the target joint point at time point. Local trajectory stability The inverse proportional normalization is achieved through... To achieve, among which, It is a natural exponential function.

6. A real-time monitoring method for human rehabilitation training using rehabilitation equipment according to claim 1, characterized in that, The calculation of dynamic detail significance score based on local motion energy factor, local trajectory stability index, and local average velocity includes: ； In the formula, For the target key point at time point Dynamic detail saliency score; For the target key point at time point The local motion energy factor of the local trajectory segment; For the target key point at time point Local trajectory stability; For the target key point at time point The instantaneous velocity vector; Indicates the magnitude of the vector; For local trajectory segments All time points The mean of the magnitude of the instantaneous velocity vector, representing the time point. The local average velocity at that location.

7. A real-time monitoring method for human rehabilitation training using rehabilitation equipment according to claim 1, characterized in that, The adjustment of the global base threshold to obtain the adaptive distance threshold used to determine whether the coordinates of the target key point at each time point are retained includes: Targeting the key points at time points Coordinates of the corresponding frame image : Calculate the relationship between the number 1 and the target key point at time points Dynamic detail saliency score The difference Calculate the product of the difference and the global basic threshold to obtain the judgment coordinate point. The adaptive distance threshold used when deciding whether to retain a value.

8. A real-time monitoring method for human rehabilitation training using rehabilitation equipment according to claim 1, characterized in that, The application of the Douglas-Puk algorithm to process the original motion trajectory sequence to obtain an optimized trajectory sequence includes: An adaptive distance threshold is used to determine whether each coordinate point should be retained. : Determining whether to retain coordinate points When calculating coordinate points To the key points before and after and perpendicular distance of the connecting line segments and coordinate points Corresponding adaptive distance threshold If a comparison is made, Then retain the coordinates. Otherwise, remove the coordinate point; repeat the judgment process until the original motion trajectory sequence is obtained. After all coordinate points in the data have been processed, the optimized trajectory sequence is obtained. ; Initially, the original motion trajectory sequence is... The first and last coordinate points are used as key points. and The key point that follows is the retained coordinates.

9. A real-time monitoring method for human rehabilitation training using rehabilitation equipment according to claim 1, characterized in that, The step of calculating the patient's training stability based on the root mean square error between the obtained ideal trajectory sequence and the optimized trajectory sequence includes: For the target key points: Calculate the optimized trajectory sequence The root mean square error (RMSE) of the ideal motion trajectory sequence is compared with the root mean square error (RMSE). The RMSE is then inversely normalized, and the result is used as the training stability of the target joint. This inverse normalization is achieved through... To achieve, among which, It is a natural exponential function; The mean of training stability for all joints is calculated as the patient's training stability.

10. A real-time monitoring method for human rehabilitation training using rehabilitation equipment according to claim 1, characterized in that, The calculation of the instantaneous velocity vector and instantaneous acceleration vector of the target joint at each time point includes: Instantaneous velocity vector , , The target key points at time points are respectively , The corresponding coordinates in the frame image; The time interval for acquiring frame images; Instantaneous acceleration vector , , The target key points at time points are respectively , The instantaneous velocity vector.

Images