Long jump data supervision method and system based on cloud computing
The long jump data supervision system uses cloud computing to analyze historical logs, calculate attention indices, and visualize center of gravity trajectories, addressing the limitations of existing systems by providing intelligent and efficient training feedback for performance optimization.
Patent Information
- Application Number
- CN202411219446.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-02
- Publication Date
- 2024-12-24
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Existing long jump data monitoring systems lack the ability to provide adaptive learning based on training data, fail to analyze indicators that can be optimized, and cannot distinguish performance improvements, leading to rigid and ineffective data display and comparison.
A long jump data supervision method and system utilizing cloud computing to analyze historical logs, calculate attention indices of indicators, and visualize center of gravity trajectories to identify abnormal indicators and provide targeted training feedback.
Enables intelligent indicator screening and efficient data feedback, allowing athletes to understand their movements better and make targeted improvements, optimizing long jump performance.
Smart Images
Figure CN119181136A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of data processing, and in particular to a long jump data supervision method and system based on cloud computing. Background Art
[0002] The development of modern technology has made data collection more efficient and accurate. Using high-speed cameras, sensors, GPS and other monitoring equipment, various data of athletes in training and competition can be collected in real time, such as take-off speed, flight time, landing distance, etc. These data provide a rich basis for the analysis and research of long jump.
[0003] At present, for the data-based application in long jump, high-speed cameras are usually used to record long jump images, and various index data are collected through image processing algorithms for reference display. This method has some problems. On the one hand, the existing technology only objectively displays the numerical values of various indicators or selects fixed values for mechanical comparison. These display data or comparison data are often more rigid. The display data cannot clearly reflect the differences between athletes in each long jump. The comparison data cannot truly reflect the weaknesses of each long jump and the true level of athletes due to the influence of fixed values. It is impossible to enable athletes to quickly perceive the long jump situation at a subjective level through data display. On the other hand, the existing technology cannot perform adaptive learning based on training data, analyze the indicators that can be optimized, and distinguish the improvement in performance that can be brought about by the optimization of various indicators, so as to recommend training directions. Therefore, at this stage, a more intelligent long jump data monitoring and management technology solution is needed to solve the above problems. Summary of the invention
[0004] The purpose of the present invention is to provide a long jump data supervision method and system based on cloud computing to solve the problems raised in the above background technology.
[0005] In order to solve the above technical problems, the present invention provides a long jump data supervision method based on cloud computing, which comprises the following steps:
[0006] S100, collecting historical logs of all objects, collecting image information of the long jump area in real time through a high-speed camera, and analyzing all indicators contained in the historical logs.
[0007] S200, analyzing the influence of various indicators on long jump performance and the stability of various indicators under each object through historical logs, calculating the attention index of each indicator and setting key indicators.
[0008] S300: After the object completes the long jump, the cloud computing end divides different action areas for key indicators, draws a center of gravity trajectory for each action area based on image information, analyzes the trajectory information and marks abnormal indicators.
[0009] S400. The visual screen playback displays the long jump image of this time, focusing on selecting the image information of the corresponding action area of each abnormal indicator, and the trajectory comparison chart is provided for reference.
[0010] In S100, the object refers to an athlete participating in long jump training, the historical record refers to the athlete's long jump record, each long jump record includes a video image during a single long jump, specific values of various indicators and a jump distance, and the indicator refers to various quantitative standards used to evaluate the athlete's performance in long jump training. The long jump area refers to a place used for long jump training, and a high-speed camera is used to collect video images of the entire long jump area in real time.
[0011] The indicators are set in advance by the management personnel according to the long jump training goals. All indicators can basically reflect the overall performance of the long jump process. Each indicator can quantify the performance of part of the long jump process. The specific value of the indicator is a numerical type. The indicators usually include take-off speed, take-off angle, take-off height, run-up distance and aerial flip angle.
[0012] In S200, the specific steps are as follows:
[0013] S201, obtain the long jump records in all object history logs, count the number N of these long jump records and the number M of all indicators contained in each long jump record, and the number of indicators in each long jump record is the same. The specific values of each indicator in the N long jump records are summed up and the average value is calculated, and the calculated average value is used as the standard value T of the corresponding indicator.
[0014] S202. Analyze the specific values and jump distances of various indicators in the long jump record, divide the specific values of each indicator by their respective standard values T, and use the calculation results as independent variables, and the jump distance as the dependent variable. M independent variables and one dependent variable are packaged into a sample, and each long jump record generates a sample. These N samples are packaged into a training set.
[0015] S203, establish a multiple linear regression model R, take all the independent variables in each sample in the training set as the input value of R, calculate the difference between the output result and the dependent variable of the corresponding sample, set the error threshold, and when the difference is less than the error threshold, mark the corresponding sample and calculate the next sample. In other cases, adjust the parameters of the regression coefficient and recalculate until all samples are marked. Among them, the expression of the multiple linear regression model is:
[0016] R=L 0 +L 1 d 1 +L 2 d 2 +...+L e d e
[0017] Where, L e is the e-th regression coefficient, d e is the e-th independent variable.
[0018] Compared with directly using the specific value as the independent variable, using the ratio of the specific value to the standard value as the independent variable can minimize the impact of differences in the numerical range of indicators on subsequent weight analysis.
[0019] S204, obtain the regression coefficients in the multivariate linear regression model expression, and sum up all the regression coefficients to obtain the total regression coefficient L sum , divide the regression coefficient corresponding to indicator K by L sum Get the influence coefficient INC K . Get object G h All long jump records in the historical log, count the number of these long jump records j, and calculate the average value of the specific value of indicator K in these j long jump records Substitute into the formula to calculate the object G h The attention index PTI of the next indicator K K The formula is as follows:
[0020]
[0021] In the formula, α is a constant greater than 1, For object G h The specific value of the indicator K in the next i-th long jump record, T K is the standard value of indicator K.
[0022] The impact coefficient indicates the degree of influence of the indicator on the jump distance. The larger the impact coefficient, the greater the influence of the indicator on the jump distance. According to the impact coefficient of the indicator and the stability change analysis of the specific value of the indicator under the same object, and the calculation of the attention index, it is possible to quantify the completion of each indicator under each object, so as to quickly find the most urgent indicator that will have the greatest improvement on the jump distance after optimization.
[0023] The standard value is obtained by summing up the specific values of the same indicator in all long jump records of all subjects. Each indicator has only one standard value. The average value refers to the sum of the specific values of the same indicator in all long jump records of a single subject. Each indicator has different average values for different subjects.
[0024] S205: Calculate the attention index of each indicator under each object respectively, and set the attention index threshold PTI td , for each object, the attention index is greater than PTI td The indicators are set as key indicators.
[0025] In S300, the specific steps are as follows:
[0026] S301, the cloud computing end obtains the video image of the long jump area, uses OpenCV technology to decompose the video into single-frame images, screens all single-frame images, and uses edge detection to remove blurred single-frame images. The YOLOv5 human body detection algorithm is used to perform target recognition on the remaining single-frame images, and the recognized human body target is used as the object. c After completing a long jump, the cloud computing end uses the Alphapose algorithm to evaluate the object's human posture, analyze the changes in the center of gravity position, and measure the specific values of various indicators after matching the key frames in the remaining single-frame images.
[0027] Using cloud computing to analyze and process images can effectively reduce the computing power consumption of on-site equipment and save timeouts caused by massive calculations. Existing image processing algorithms can analyze video images frame by frame, locate key frames that can be used to measure indicators, and automatically collect specific values of indicators.
[0028] S302, obtain the key frame that matches when collecting the key indicator P, and mark the position coordinates (X P ,Y P ), set an influence distance U, with (X P ,Y P ) is the center of the circle and U is the radius. A circular area is drawn as the action area of the key indicator P. Each key indicator corresponds to an action area. c The center of gravity position changes within the action area and a center of gravity trajectory is drawn. There is a center of gravity trajectory in each action area.
[0029] S303, obtain the object G c The attention index of indicator P for all other objects except GZD is marked as less than GZD td For other objects, divide the action area for the indicator P and draw the center of gravity trajectory according to the video image in the history record of the marked object according to step S302, fit all the center of gravity trajectories of the indicator P under all the marked objects, calculate the average trajectory through the fitting algorithm, and the average trajectory represents the common characteristics of these trajectories; establish an average trajectory for each key coordinate under each object.
[0030] The average trajectory refers to a representative trajectory obtained by averaging the position information at each time point in a set of center of gravity trajectories. It can provide a clear view to help identify and understand the overall trend of the center of gravity movement.
[0031] The generation of the average trajectory requires that all centroid trajectories are aligned in time, that is, each time point has corresponding position information. The position information of each time point is then averaged, usually by taking the arithmetic mean of the coordinates of all trajectories at that time point to obtain the coordinates of the average trajectory.
[0032] S304, establish two coordinate sets for each key coordinate, obtain the center of gravity trajectory VZ and the corresponding average trajectory VP of the key coordinate, evenly intercept s trajectory coordinates on VZ and VP and put them into different coordinate sets respectively, substitute them into the formula to calculate the abnormal index ANI of each key coordinate respectively, the formula is as follows:
[0033]
[0034] In the formula, is the oth trajectory coordinate in the coordinate set corresponding to the center of gravity trajectory, is the oth trajectory coordinate in the coordinate set corresponding to the average trajectory; set the anomaly index ANI td , the abnormal index is less than ANI td The key coordinates of are used as anomaly indicators.
[0035] In S400, the video image of the long jump is played back through the visualization screen, and the corresponding action areas of various abnormal indicators are framed in the video image. When the user clicks the framed area, a comparison chart of the center of gravity trajectory and the average trajectory in the action area is displayed to provide a training reference.
[0036] After each long jump, athletes can combine dynamic analysis to find out the reasons for the difference between the center of gravity trajectory of this long jump and the average trajectory, such as the athlete's take-off angle, flight action, landing posture, etc. According to the analysis results, the athlete's long jump action technique is adjusted in a targeted manner, such as optimizing the take-off angle and power output, improving the body posture when flying, and optimizing the body balance when landing. By continuously monitoring the adjusted center of gravity trajectory, observe whether it is more consistent with the average trajectory, and check whether the jump distance is improved.
[0037] Through this kind of analysis and feedback based on the center of gravity trajectory, athletes can better understand their long jump movements and make targeted improvements. This not only helps to optimize the long jump results, but also helps athletes develop the ability to perceive their own movements, thereby continuously improving their long jump level. The analysis of the center of gravity trajectory can provide long jump athletes with a technical feedback channel, which can be applied in long jump training.
[0038] A long jump data supervision system based on cloud computing comprises a data acquisition module, an index analysis module, a data feedback module and a visualization module.
[0039] The data collection module is used to collect the historical logs of all objects and the image information of the long jump area, and analyze all the indicators contained in the historical logs. The indicator analysis module is used to analyze the influence of various indicators on the long jump results, as well as the stability of each indicator of each object, calculate the attention index of each indicator under each object and set key indicators. After the object completes the long jump, the data feedback module divides the action area for the key indicators through the cloud computing end, draws the center of gravity trajectory for each action area according to the image information, and marks the abnormal indicators. The visualization module is used to display the long jump image and abnormal indicator related information for reference.
[0040] The data acquisition module includes an image information acquisition unit and a history log acquisition unit.
[0041] The image information acquisition unit acquires video images of the entire long jump area in real time through a high-speed camera. The long jump area refers to a place used for long jump training.
[0042] The historical log collection unit is used to collect long jump records of all objects. Objects refer to athletes participating in long jump training. Each long jump record includes video images, specific values of various indicators and jumping distance during a single long jump. Indicators refer to various quantitative standards used to evaluate athlete performance in long jump training.
[0043] The indicator analysis module includes an impact analysis unit and a focus planning unit.
[0044] The impact analysis unit is used to analyze the impact of various indicators on long jump results.
[0045] First, the number N of long jump records in the historical logs of all objects is obtained and counted, and the specific values of each indicator in these N long jump records are summed up respectively, and the average value is calculated as the standard value of the corresponding indicator.
[0046] Secondly, the specific values of each indicator in the long jump record were divided by their respective standard values to obtain the calculation results as independent variables, and the jumping distance as the dependent variable. All independent variables and one dependent variable in each long jump record were packaged into a sample, and these N samples were packaged into a training set.
[0047] Finally, a multiple linear regression model is established, and all independent variables in each sample in the training set are used as input values. The difference between the output result and the dependent variable of the corresponding sample is calculated, and the parameters of the regression coefficient are adjusted according to the difference until all samples meet the expression calculation of the multiple linear regression model; all regression coefficients in the multiple linear regression model expression are summed to obtain the total regression coefficient L sum , the regression coefficient corresponding to indicator K divided by L sum Get the influence coefficient INC K .
[0048] The attention planning unit is used to calculate the attention index of each indicator under each object and set key indicators.
[0049] Get and count object G h The number of all long jump records in the historical log is j, and the average value of the specific value of indicator K in these j long jump records is calculated Substituting into the formula: In the calculation, the object G is obtained h The attention index PTI of the next indicator K K . Where α is a constant greater than 1, For object G h The specific value of the indicator K in the next i-th long jump record, T K is the standard value of indicator K. Calculate the attention index of each indicator under each object and set the attention index threshold PTI td , for each object, the attention index is greater than PTI td The indicators are set as key indicators.
[0050] The data feedback module includes a region division unit and an abnormality identification unit.
[0051] The area division unit is used to divide the action area.
[0052] First, the cloud computing end uses OpenCV technology to decompose the video image of the long jump area into single-frame images. After removing the blurred single-frame images, the YOLOv5 human detection algorithm is used to identify the human target as the object.
[0053] Secondly, when the object G c After completing a long jump, the cloud computing end uses the Alphapose algorithm to evaluate the subject's human posture, analyze the changes in the center of gravity position, and measure the specific values of various indicators after matching the key frames in the remaining single-frame images.
[0054] Finally, obtain the key frame that matches the key indicator P, and mark the position coordinates (X P ,Y P ), set an influence distance U, with (X P ,Y P ) is the center and U is the radius, a circular area is drawn as the action area of the key indicator P, and each key indicator corresponds to an action area.
[0055] The anomaly identification unit is used to identify abnormal indicators.
[0056] First, analyze the object G c The change of the center of gravity position in each action area and the trajectory of the center of gravity are drawn. cThe attention index of indicator P for all other objects except GZD is marked as less than GZD td For other objects, the center of gravity trajectory is drawn for the indicator P according to the video images in the history of the marked object.
[0057] Secondly, all the centroid trajectories of the index P under all the marked objects are fitted, and the average trajectory is calculated through the fitting algorithm. The average trajectory represents the common characteristics of these trajectories. An average trajectory is established for each key coordinate under each object.
[0058] Finally, two coordinate sets are established for each key coordinate, and the center of gravity trajectory VZ of the key coordinate and the corresponding average trajectory VP are obtained. S trajectory coordinates are uniformly intercepted on VZ and VP and put into different coordinate sets respectively, and substituted into the formula: Calculate the anomaly index ANI for each key coordinate. Set the anomaly index ANI td , the abnormal index is less than ANI td The key coordinates of are used as abnormal indicators. is the oth trajectory coordinate in the coordinate set corresponding to the center of gravity trajectory, is the oth trajectory coordinate in the coordinate set corresponding to the average trajectory.
[0059] The visualization module is used to display the long jump image, focusing on selecting the image information of the corresponding action area of each abnormal indicator, and the trajectory comparison chart to provide training reference.
[0060] Compared with the prior art, the beneficial effects achieved by the present invention are:
[0061] 1. Intelligent index screening: This application analyzes the long jump records of all subjects, quantifies the influence of various indicators on long jump results, analyzes the specific values of each indicator under each object, quantifies the stability of each indicator under each object, analyzes the attention index of each indicator under each object based on the comprehensive influence and stability, and screens out key indicators for targeted analysis. Compared with the full indicator analysis and mechanical comparison of traditional technologies, this application is more accurate and intelligent.
[0062] 2. Efficient data feedback: This application determines the key frames according to the key indicators, thereby dividing the action area and drawing the center of gravity trajectory. By obtaining the high-quality data of other objects on this indicator to fit the average trajectory, the difference between the center of gravity trajectory and the average trajectory of each long jump is analyzed to determine the abnormal indicators. The image display of the action area corresponding to the abnormal indicator and the trajectory comparison chart provide intuitive training optimization direction, which is more efficient and easier to understand than the numerical stereotyped display of the prior art.
[0063] In summary, compared with traditional technologies, the present invention has the advantages of intelligent indicator screening and efficient data feedback, and can improve the efficiency of long jump data supervision. BRIEF DESCRIPTION OF THE DRAWINGS
[0064] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:
[0065] Figure 1 It is a flow chart of a long jump data supervision method based on cloud computing of the present invention;
[0066] Figure 2 It is a structural schematic diagram of a long jump data monitoring system based on cloud computing of the present invention. DETAILED DESCRIPTION
[0067] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0068] See also Figure 1 The present invention provides a long jump data supervision method based on cloud computing, which comprises the following steps:
[0069] S100, collecting historical logs of all objects, collecting image information of the long jump area in real time through a high-speed camera, and analyzing all indicators contained in the historical logs.
[0070] S200, analyzing the influence of various indicators on long jump performance and the stability of various indicators under each object through historical logs, calculating the attention index of each indicator and setting key indicators.
[0071] S300: After the object completes the long jump, the cloud computing end divides different action areas for key indicators, draws a center of gravity trajectory for each action area based on image information, analyzes the trajectory information and marks abnormal indicators.
[0072] S400. The visual screen playback displays the long jump image of this time, focusing on selecting the image information of the corresponding action area of each abnormal indicator, and the trajectory comparison chart is provided for reference.
[0073] In S100, the object refers to an athlete participating in long jump training, the historical record refers to the athlete's long jump record, each long jump record includes a video image during a single long jump, specific values of various indicators and a jump distance, and the indicator refers to various quantitative standards used to evaluate the athlete's performance in long jump training. The long jump area refers to a place used for long jump training, and a high-speed camera is used to collect video images of the entire long jump area in real time.
[0074] The indicators are set in advance by the management personnel according to the long jump training goals. All indicators can basically reflect the overall performance of the long jump process. Each indicator can quantify the performance of part of the long jump process. The specific value of the indicator is a numerical type. The indicators usually include take-off speed, take-off angle, take-off height, run-up distance and aerial flip angle.
[0075] In S200, the specific steps are as follows:
[0076] S201, obtain the long jump records in all object history logs, count the number N of these long jump records and the number M of all indicators contained in each long jump record, and the number of indicators in each long jump record is the same. The specific values of each indicator in the N long jump records are summed up and the average value is calculated, and the calculated average value is used as the standard value T of the corresponding indicator.
[0077] S202. Analyze the specific values and jump distances of various indicators in the long jump record, divide the specific values of each indicator by their respective standard values T, and use the calculation results as independent variables, and the jump distance as the dependent variable. M independent variables and one dependent variable are packaged into a sample, and each long jump record generates a sample. These N samples are packaged into a training set.
[0078] S203, establish a multiple linear regression model R, take all the independent variables in each sample in the training set as the input value of R, calculate the difference between the output result and the dependent variable of the corresponding sample, set the error threshold, and when the difference is less than the error threshold, mark the corresponding sample and calculate the next sample. In other cases, adjust the parameters of the regression coefficient and recalculate until all samples are marked. Among them, the expression of the multiple linear regression model is:
[0079] R=L 0 +L 1 d 1 +L 2 d 2 +...+L e d e
[0080] Where, L e is the e-th regression coefficient, d e is the e-th independent variable.
[0081] Compared with directly using the specific value as the independent variable, using the ratio of the specific value to the standard value as the independent variable can minimize the impact of differences in the numerical range of indicators on subsequent weight analysis.
[0082] S204, obtain the regression coefficients in the multivariate linear regression model expression, and sum up all the regression coefficients to obtain the total regression coefficient L sum , divide the regression coefficient corresponding to indicator K by L sum Get the influence coefficient INC K . Get object G h All long jump records in the historical log, count the number of these long jump records j, and calculate the average value of the specific value of indicator K in these j long jump records Substitute into the formula to calculate the object G h The attention index PTI of the next indicator K K The formula is as follows:
[0083]
[0084] In the formula, α is a constant greater than 1, For object G h The specific value of the indicator K in the next i-th long jump record, T K is the standard value of indicator K.
[0085] The impact coefficient indicates the degree of influence of the indicator on the jump distance. The larger the impact coefficient, the greater the influence of the indicator on the jump distance. According to the impact coefficient of the indicator and the stability change analysis of the specific value of the indicator under the same object, and the calculation of the attention index, it is possible to quantify the completion of each indicator under each object, so as to quickly find the most urgent indicator that will have the greatest improvement on the jump distance after optimization.
[0086] The standard value is obtained by summing up the specific values of the same indicator in all long jump records of all subjects. Each indicator has only one standard value. The average value refers to the sum of the specific values of the same indicator in all long jump records of a single subject. Each indicator has different average values for different subjects.
[0087] S205: Calculate the attention index of each indicator under each object respectively, and set the attention index threshold PTI td , for each object, the attention index is greater than PTI td The indicators are set as key indicators.
[0088] In S300, the specific steps are as follows:
[0089] S301, the cloud computing end obtains the video image of the long jump area, uses OpenCV technology to decompose the video into single-frame images, screens all single-frame images, and uses edge detection to remove blurred single-frame images. The YOLOv5 human body detection algorithm is used to perform target recognition on the remaining single-frame images, and the recognized human body target is used as the object. c After completing a long jump, the cloud computing end uses the Alphapose algorithm to evaluate the object's human posture, analyze the changes in the center of gravity position, and measure the specific values of various indicators after matching the key frames in the remaining single-frame images.
[0090] Using cloud computing to analyze and process images can effectively reduce the computing power consumption of on-site equipment and save timeouts caused by massive calculations. Existing image processing algorithms can analyze video images frame by frame, locate key frames that can be used to measure indicators, and automatically collect specific values of indicators.
[0091] S302, obtain the key frame that matches when collecting the key indicator P, and mark the position coordinates (X P ,Y P ), set an influence distance U, with (X P ,Y P ) is the center of the circle and U is the radius. A circular area is drawn as the action area of the key indicator P. Each key indicator corresponds to an action area. c The center of gravity position changes within the action area and a center of gravity trajectory is drawn. There is a center of gravity trajectory in each action area.
[0092] S303, obtain the object G c The attention index of indicator P for all other objects except GZD is marked as less than GZD td For other objects, divide the action area for the indicator P and draw the center of gravity trajectory according to the video image in the history record of the marked object according to step S302, fit all the center of gravity trajectories of the indicator P under all the marked objects, calculate the average trajectory through the fitting algorithm, and the average trajectory represents the common characteristics of these trajectories; establish an average trajectory for each key coordinate under each object.
[0093] The average trajectory refers to a representative trajectory obtained by averaging the position information at each time point in a set of center of gravity trajectories. It can provide a clear view to help identify and understand the overall trend of the center of gravity movement.
[0094] The generation of the average trajectory requires that all centroid trajectories are aligned in time, that is, each time point has corresponding position information. The position information of each time point is then averaged, usually by taking the arithmetic mean of the coordinates of all trajectories at that time point to obtain the coordinates of the average trajectory.
[0095] S304, establish two coordinate sets for each key coordinate, obtain the center of gravity trajectory VZ and the corresponding average trajectory VP of the key coordinate, evenly intercept s trajectory coordinates on VZ and VP and put them into different coordinate sets respectively, substitute them into the formula to calculate the abnormal index ANI of each key coordinate respectively, the formula is as follows:
[0096]
[0097] In the formula, is the oth trajectory coordinate in the coordinate set corresponding to the center of gravity trajectory, The oth trajectory coordinate in the coordinate set corresponding to the average trajectory; set the anomaly index ANI td , the abnormal index is less than ANI td The key coordinates of are used as anomaly indicators.
[0098] In S400, the video image of the long jump is played back through the visualization screen, and the corresponding action areas of various abnormal indicators are framed in the video image. When the user clicks the framed area, a comparison chart of the center of gravity trajectory and the average trajectory in the action area is displayed to provide a training reference.
[0099] After each long jump, athletes can combine dynamic analysis to find out the reasons for the difference between the center of gravity trajectory of this long jump and the average trajectory, such as the athlete's take-off angle, flight action, landing posture, etc. According to the analysis results, the athlete's long jump action technique is adjusted in a targeted manner, such as optimizing the take-off angle and power output, improving the body posture when flying, and optimizing the body balance when landing. By continuously monitoring the adjusted center of gravity trajectory, observe whether it is more consistent with the average trajectory, and check whether the jump distance is improved.
[0100] Through this kind of analysis and feedback based on the center of gravity trajectory, athletes can better understand their long jump movements and make targeted improvements. This not only helps to optimize the long jump results, but also helps athletes develop the ability to perceive their own movements, thereby continuously improving their long jump level. The analysis of the center of gravity trajectory can provide long jump athletes with a technical feedback channel, which can be applied in long jump training.
[0101] See also Figure 2 The present invention provides a long jump data supervision system based on cloud computing, which includes a data acquisition module, an indicator analysis module, a data feedback module and a visualization module.
[0102] The data collection module is used to collect the historical logs of all objects and the image information of the long jump area, and analyze all the indicators contained in the historical logs. The indicator analysis module is used to analyze the influence of various indicators on the long jump results, as well as the stability of each indicator of each object, calculate the attention index of each indicator under each object and set key indicators. After the object completes the long jump, the data feedback module divides the action area for the key indicators through the cloud computing end, draws the center of gravity trajectory for each action area according to the image information, and marks the abnormal indicators. The visualization module is used to display the long jump image and abnormal indicator related information for reference.
[0103] The data acquisition module includes an image information acquisition unit and a history log acquisition unit.
[0104] The image information acquisition unit acquires video images of the entire long jump area in real time through a high-speed camera. The long jump area refers to a place used for long jump training.
[0105] The historical log collection unit is used to collect long jump records of all objects. Objects refer to athletes participating in long jump training. Each long jump record includes video images, specific values of various indicators and jumping distance during a single long jump. Indicators refer to various quantitative standards used to evaluate athlete performance in long jump training.
[0106] The indicator analysis module includes an impact analysis unit and a focus planning unit.
[0107] The impact analysis unit is used to analyze the impact of various indicators on long jump results.
[0108] First, the number N of long jump records in the historical logs of all objects is obtained and counted, and the specific values of each indicator in these N long jump records are summed up respectively, and the average value is calculated as the standard value of the corresponding indicator.
[0109] Secondly, the specific values of each indicator in the long jump record were divided by their respective standard values to obtain the calculation results as independent variables, and the jumping distance as the dependent variable. All independent variables and one dependent variable in each long jump record were packaged into a sample, and these N samples were packaged into a training set.
[0110] Finally, a multiple linear regression model is established, and all independent variables in each sample in the training set are used as input values. The difference between the output result and the dependent variable of the corresponding sample is calculated, and the parameters of the regression coefficient are adjusted according to the difference until all samples meet the expression calculation of the multiple linear regression model; all regression coefficients in the multiple linear regression model expression are summed to obtain the total regression coefficient L sum , the regression coefficient corresponding to indicator K divided by L sum Get the influence coefficient INC K .
[0111] The attention planning unit is used to calculate the attention index of each indicator under each object and set key indicators.
[0112] Get and count object G h The number of all long jump records in the historical log is j, and the average value of the specific value of indicator K in these j long jump records is calculated Substituting into the formula: In the calculation, the object G is obtained h The attention index PTI of the next indicator K K . Where α is a constant greater than 1, For object G h The specific value of the indicator K in the next i-th long jump record, T K is the standard value of indicator K. Calculate the attention index of each indicator under each object and set the attention index threshold PTI td , for each object, the attention index is greater than PTI td The indicators are set as key indicators.
[0113] The data feedback module includes a region division unit and an abnormality identification unit.
[0114] The area division unit is used to divide the action area.
[0115] First, the cloud computing end uses OpenCV technology to decompose the video image of the long jump area into single-frame images. After removing the blurred single-frame images, the YOLOv5 human detection algorithm is used to identify the human target as the object.
[0116] Secondly, when the object G c After completing a long jump, the cloud computing end uses the Alphapose algorithm to evaluate the subject's human posture, analyze the changes in the center of gravity position, and measure the specific values of various indicators after matching the key frames in the remaining single-frame images.
[0117] Finally, obtain the key frame that matches the key indicator P, and mark the position coordinates (X P ,Y P ), set an influence distance U, with (X P ,Y P ) is the center and U is the radius, a circular area is drawn as the action area of the key indicator P, and each key indicator corresponds to an action area.
[0118] The anomaly identification unit is used to identify abnormal indicators.
[0119] First, analyze the object G c The change of the center of gravity position in each action area and the trajectory of the center of gravity are drawn. cThe attention index of indicator P for all other objects except GZD is marked as less than GZD td For other objects, the center of gravity trajectory is drawn for the indicator P according to the video images in the history of the marked object.
[0120] Secondly, all the centroid trajectories of the index P under all the marked objects are fitted, and the average trajectory is calculated through the fitting algorithm. The average trajectory represents the common characteristics of these trajectories. An average trajectory is established for each key coordinate under each object.
[0121] Finally, two coordinate sets are established for each key coordinate, and the center of gravity trajectory VZ of the key coordinate and the corresponding average trajectory VP are obtained. S trajectory coordinates are uniformly intercepted on VZ and VP and put into different coordinate sets respectively, and substituted into the formula: Calculate the anomaly index ANI for each key coordinate. Set the anomaly index ANI td , the abnormal index is less than ANI td The key coordinates of are used as abnormal indicators. is the oth trajectory coordinate in the coordinate set corresponding to the center of gravity trajectory, is the oth trajectory coordinate in the coordinate set corresponding to the average trajectory.
[0122] The visualization module is used to display the long jump image, focusing on selecting the image information of the corresponding action area of each abnormal indicator, and the trajectory comparison chart to provide training reference.
[0123] Embodiment 1:
[0124] Assume that there are indicators K1 and K2 under a certain object, their influence coefficients are 0.4 and 0.6 respectively, and their standard values are 60 and 20 respectively. The indicators and specific values in the long jump record of the object are as follows:
[0125] The first long jump record: K1: 65°; K2: 25°;
[0126] The second long jump record: K1: 60°; K2: 27°;
[0127] The third long jump record: K1: 55°; K2: 29°;
[0128] When the constant α is 1.2, the attention index of indicators K1 and K2 is calculated by using the formula:
[0129]
[0130] The attention indexes of K1 and K2 are 0.175 and 0.263 respectively.
[0131] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.
[0132] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.
Claims
1. A long jump data supervision method based on cloud computing, characterized by: The method comprises the following steps: S100, collecting historical logs of all objects, collecting image information of the long jump area in real time through a high-speed camera, and analyzing all indicators contained in the historical logs; S200, analyzing the influence of various indicators on long jump performance and the stability of various indicators under each object through historical logs, calculating the attention index of each indicator to set key indicators; S300, when the object completes the long jump, the cloud computing end divides different action areas for key indicators, draws the center of gravity trajectory for each action area according to the image information, analyzes the trajectory information and marks abnormal indicators; S400. The visual screen playback displays the long jump image of this time, focusing on selecting the image information of the corresponding action area of each abnormal indicator, and the trajectory comparison chart is provided for reference.
2. The long jump data supervision method based on cloud computing according to claim 1, characterized in that: In S100, the object refers to the athlete participating in the long jump training, the historical record refers to the athlete's long jump record, each long jump record includes a video image during a single long jump, the specific values of various indicators and the jump distance, the indicator refers to various quantitative standards used to evaluate the athlete's performance in the long jump training; the long jump area refers to the place used for long jump training, and the video image of the entire long jump area is collected in real time by a high-speed camera.
3. The long jump data supervision method based on cloud computing according to claim 2 is characterized in that: In S200, the specific steps are as follows: S201, obtaining long jump records in all object history logs, counting the number N of these long jump records and the number M of all indicators contained in each long jump record, summing up the specific values of each indicator in these N long jump records and calculating the average value, and using the calculated average value as the standard value T of the corresponding indicator; S202, analyzing the specific values and jump distances of various indicators in the long jump record, dividing the specific value of each indicator by the respective standard value T to obtain the calculation results as independent variables, and the jump distance as the dependent variable, M independent variables and one dependent variable are packaged into a sample, each long jump record generates a sample, and these N samples are packaged into a training set; S203, establish a multiple linear regression model R, use all independent variables in each sample in the training set as the input value of R, calculate the difference between the output result and the dependent variable of the corresponding sample, set the error threshold, and when the difference is less than the error threshold, mark the corresponding sample and calculate the next sample; in other cases, adjust the parameters of the regression coefficient and recalculate until all samples are marked; wherein, the expression of the multiple linear regression model is: R=L0+L1d1+L2d2+...+L e d e Where, L e is the e-th regression coefficient, d e is the e-th independent variable; S204, obtain the regression coefficients in the multivariate linear regression model expression, and sum up all the regression coefficients to obtain the total regression coefficient L sum , divide the regression coefficient corresponding to indicator K by L sum Get the influence coefficient INC K ; Get object G h All long jump records in the historical log, count the number of these long jump records j, and calculate the average value of the specific value of indicator K in these j long jump records Substitute into the formula to calculate the object G h The attention index PTI of the next indicator K K ; The formula is as follows: In the formula, α is a constant greater than 1, For object G h The specific value of the indicator K in the next i-th long jump record, T K is the standard value of indicator K; S205: Calculate the attention index of each indicator under each object respectively, and set the attention index threshold PTI td , for each object, the attention index is greater than PTI td The indicators are set as key indicators.
4. The long jump data supervision method based on cloud computing according to claim 3 is characterized in that: In S300, the specific steps are as follows: S301, the cloud computing end obtains the video image of the long jump area, uses OpenCV technology to decompose the video into single-frame images, screens all single-frame images, and uses edge detection method to remove blurred single-frame images; uses YOLOv5 human body detection algorithm to perform target recognition on the remaining single-frame images, and uses the recognized human body target as the object; when the object G c After completing a long jump, the cloud computing end uses the Alphapose algorithm to evaluate the subject's human posture, analyze the change in the center of gravity position, and measure the specific values of various indicators after matching the key frames in the remaining single-frame images; S302, obtain the key frame that matches when collecting the key indicator P, and mark the position coordinates (X P ,Y P ), set an influence distance U, with (X P ,Y P ) is the center of the circle and U is the radius, a circular area is drawn as the action area of the key indicator P, and each key indicator corresponds to an action area; the analysis object G c The changes of the center of gravity position in the action area are analyzed and the center of gravity trajectory is drawn. There is a center of gravity trajectory in each action area. S303, obtain the object G c The attention index of indicator P for all other objects except GZD is marked as less than GZD td For other objects, according to the video image in the history record of the marked object, the action area is divided for the indicator P and the center of gravity trajectory is drawn according to step S302, all the center of gravity trajectories of the indicator P under all the marked objects are fitted, and the average trajectory is calculated by the fitting algorithm, and the average trajectory represents the common characteristics of these trajectories; Establish an average trajectory for each key coordinate under each object; S304, establish two coordinate sets for each key coordinate, obtain the center of gravity trajectory VZ and the corresponding average trajectory VP of the key coordinate, evenly intercept s trajectory coordinates on VZ and VP and put them into different coordinate sets respectively, substitute them into the formula to calculate the abnormal index ANI of each key coordinate respectively, the formula is as follows: In the formula, is the oth trajectory coordinate in the coordinate set corresponding to the center of gravity trajectory, The oth trajectory coordinate in the coordinate set corresponding to the average trajectory; set the anomaly index ANI td , the abnormal index is less than ANI td The key coordinates of are used as anomaly indicators.
5. The long jump data supervision method based on cloud computing according to claim 4 is characterized in that: In S400, the video image of the long jump is played back through the visualization screen, and the corresponding action areas of various abnormal indicators are framed in the video image. When the user clicks the framed area, a comparison chart of the center of gravity trajectory and the average trajectory in the action area is displayed to provide a training reference.
6. A long jump data monitoring system based on cloud computing, characterized in that: The system includes a data acquisition module, an indicator analysis module, a data feedback module and a visualization module; The data acquisition module is used to collect the historical logs of all objects and the image information of the long jump area, and to parse all the indicators contained in the historical logs; the indicator analysis module is used to analyze the influence of various indicators on the long jump results, as well as the stability of each indicator of each object, calculate the attention index of each indicator under each object and set key indicators; after the object completes the long jump, the data feedback module divides the action area for the key indicators through the cloud computing end, draws the center of gravity trajectory for each action area according to the image information, and thus marks the abnormal indicators; the visualization module is used to display the long jump image and related information of the abnormal indicators for reference.
7. The long jump data monitoring system based on cloud computing according to claim 6 is characterized in that: The data acquisition module includes an image information acquisition unit and a history log acquisition unit; The image information acquisition unit acquires video images of the entire long jump area in real time through a high-speed camera. The long jump area refers to a place used for long jump training. The historical log collection unit is used to collect long jump records of all objects. Objects refer to athletes participating in long jump training. Each long jump record includes video images, specific values of various indicators and jumping distance during a single long jump. Indicators refer to various quantitative standards used to evaluate athlete performance in long jump training.
8. The long jump data monitoring system based on cloud computing according to claim 7, characterized in that: The indicator analysis module includes the impact analysis unit and the focus planning unit; The impact analysis unit is used to analyze the impact of various indicators on long jump performance; First, obtain and count the number of long jump records N in the historical logs of all objects, sum up the specific values of each indicator in these N long jump records, and calculate the average value as the standard value of the corresponding indicator; Secondly, the specific values of each indicator in the long jump record are divided by their respective standard values to obtain the calculated results as independent variables, and the jump distance is used as the dependent variable. All independent variables and one dependent variable in each long jump record are packaged into a sample, and these N samples are packaged as a training set; Finally, a multiple linear regression model is established, and all independent variables in each sample in the training set are used as input values. The difference between the output result and the dependent variable of the corresponding sample is calculated, and the parameters of the regression coefficient are adjusted according to the difference until all samples meet the expression calculation of the multiple linear regression model; all regression coefficients in the multiple linear regression model expression are summed to obtain the total regression coefficient L sum , the regression coefficient corresponding to indicator K divided by L sum Get the influence coefficient INC K ; The attention planning unit is used to calculate the attention index of each indicator under each object and set key indicators; Get and count object G h The number of all long jump records in the historical log is j, and the average value of the specific value of indicator K in these j long jump records is calculated Substituting into the formula: In the calculation, the object G is obtained h The attention index PTI of the next indicator K K ; where α is a constant greater than 1, For object G h The specific value of the indicator K in the next i-th long jump record, T K is the standard value of indicator K; calculate the attention index of each indicator under each object respectively, and set the attention index threshold PTI td , for each object, the attention index is greater than PTI td The indicators are set as key indicators.
9. The long jump data monitoring system based on cloud computing according to claim 8, characterized in that: The data feedback module includes a region division unit and an abnormality identification unit; The area division unit is used to divide the action area; First, the cloud computing end uses OpenCV technology to decompose the video image of the long jump area into single-frame images. After removing the blurred single-frame images, the YOLOv5 human detection algorithm is used to identify the human target as the object; Secondly, when the object G c After completing a long jump, the cloud computing end uses the Alphapose algorithm to evaluate the subject's human posture, analyze the change in the center of gravity position, and measure the specific values of various indicators after matching the key frames in the remaining single-frame images; Finally, obtain the key frame that matches the key indicator P, and mark the position coordinates (X P ,Y P ), set an influence distance U, with (X P ,Y P ) is the center of the circle and U is the radius, a circular area is drawn as the action area of the key indicator P, and each key indicator corresponds to an action area; The anomaly identification unit is used to identify abnormal indicators; First, analyze the object G c The change of the center of gravity position in each action area and draw the center of gravity trajectory; obtain the object G c The attention index of indicator P for all other objects except GZD td For other objects, the center of gravity trajectory is drawn for the index P according to the video image in the history record of the marked object; Secondly, all the centroid trajectories of the index P under all the marked objects are fitted, and the average trajectory is calculated through the fitting algorithm. The average trajectory represents the common characteristics of these trajectories; Establish an average trajectory for each key coordinate under each object; Finally, two coordinate sets are established for each key coordinate, and the center of gravity trajectory VZ of the key coordinate and the corresponding average trajectory VP are obtained. S trajectory coordinates are uniformly intercepted on VZ and VP and put into different coordinate sets respectively, and substituted into the formula: In the above example, the anomaly index ANI of each key coordinate is calculated and set td , the abnormal index is less than ANI td The key coordinates of are used as abnormal indicators; among them, is the oth trajectory coordinate in the coordinate set corresponding to the center of gravity trajectory, is the oth trajectory coordinate in the coordinate set corresponding to the average trajectory.
10. The long jump data monitoring system based on cloud computing according to claim 9, characterized in that: The visualization module is used to display the long jump image, focusing on selecting the image information of the corresponding action area of each abnormal indicator, and the trajectory comparison chart to provide training reference.