Long jump evaluation method, device and equipment and storage medium
By first determining the candidate area and then determining the target position in the long jump evaluation system, the problems of low measurement efficiency and insufficient accuracy in the prior art are solved, and a fast and accurate long jump evaluation is achieved, suitable for physical exercise and physical fitness in multi-person scenarios test.
Patent Information
- Application Number
- CN202411169170.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-23
- Publication Date
- 2024-12-20
AI Technical Summary
The existing long jump evaluation system is inefficient and inaccurate in measuring, and cannot quickly and accurately evaluate athletes' long jump ability.
By acquiring several target images of the target evaluation object, first determine the area containing the target from the entire image as the candidate area, and then determine the target position from the candidate area, thereby improving detection efficiency and reducing interference, thereby improving the accuracy of the evaluation results.
It effectively improves the efficiency and accuracy of long jump assessment, and can accurately collect data in scenarios where multiple people jump at the same time, and is suitable for physical exercise and physical fitness testing.
Smart Images
Figure CN119169249A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of computer vision technology, and in particular to a long jump evaluation method, device, equipment and storage medium. Background Art
[0002] Standing long jump is a basic sports event widely used in physical training and testing, mainly used to evaluate athletes' explosive power, coordination and lower limb strength. At present, although there is a long jump evaluation system based on computer vision, the current long jump evaluation system has the problems of low measurement efficiency or inaccurate measurement, and fast and accurate long jump evaluation is of great significance to the long jump test project. Summary of the invention
[0003] The main technical problem solved by the present application is to provide a long jump evaluation method, device, equipment and storage medium, which can improve the accuracy of evaluation results while improving evaluation efficiency.
[0004] In order to solve the above technical problems, a technical solution adopted in the present application is: to provide a long jump evaluation method, the method comprising: obtaining several first target images of the target evaluation object; respectively determining an area containing at least the target evaluation object from each of the first target images, and using it as a candidate area; respectively determining a first target position where the target evaluation object is located from the candidate area of each of the first target images; and determining an evaluation result of the target evaluation object based on the first target position corresponding to each of the first target images.
[0005] In order to solve the above technical problems, a technical solution adopted by the present application is: to provide a long jump evaluation device, including an acquisition module, a first determination module, a second determination module and a result determination module. Among them, the acquisition module is used to acquire a number of first target images of the target evaluation object; the first determination module is used to determine the area at least containing the target evaluation object from each first target image, and use it as a candidate area; the second determination module is used to determine the first target position where the target evaluation object is located from the candidate area of each first target image; the result determination module is used to determine the evaluation result of the target evaluation object based on the first target position corresponding to each first target image.
[0006] To solve the above technical problems, another technical solution adopted in the present application is: to provide an electronic device, comprising a memory and a processor coupled to each other, the memory storing program instructions; the processor is used to execute the program instructions stored in the memory to implement the above method.
[0007] In order to solve the above technical problem, another technical solution adopted by the present application is: providing a computer-readable storage medium for storing program instructions, which can be executed to implement the above method.
[0008] In the above scheme, before determining the first target position of the target evaluation object in each first target image, first determine the area at least containing the target evaluation object from each first target image and use it as a candidate area, and then determine the first target position where the target evaluation object is located from the candidate area of each first target image. It can be seen that the present application first determines an approximate area containing the target evaluation object from the entire first target image, and uses the approximate area as a candidate area, and then determines the first target position where the target evaluation object is located from the candidate area. Compared with the method of directly determining the first target position from the entire first target image, the above method of the present application can effectively reduce the search area of the target evaluation object, so it can improve the detection efficiency of the subsequent target evaluation object position. Furthermore, on the basis of reducing the search area, the present application can effectively reduce the interference of information in some areas that are not related to the target evaluation object, so it can improve the position determination result of the target evaluation object, and then can improve the subsequent evaluation results. BRIEF DESCRIPTION OF THE DRAWINGS
[0009] Figure 1 It is a flow chart of an embodiment of a long jump evaluation method provided by the present application;
[0010] Figure 2 yes Figure 1 The flowchart of step S13 is shown as an embodiment;
[0011] Figure 3 yes Figure 1 The flowchart of step S14 is shown as an embodiment;
[0012] Figure 4 yes Figure 3 The flowchart of step S32 is shown as an embodiment;
[0013] Figure 5 yes Figure 3 The flowchart of step S33 is shown as an embodiment;
[0014] Figure 6 A schematic diagram of the framework of an embodiment of a long jump evaluation device provided by the present application;
[0015] Figure 7 It is a schematic diagram of a framework of an embodiment of an electronic device provided by the present application;
[0016] Figure 8 It is a schematic diagram of the framework of the computer-readable storage medium provided by this application. DETAILED DESCRIPTION
[0017] In order to make the purpose, technical solution and effect of the present application clearer and more specific, the present application is further described in detail below with reference to the accompanying drawings and examples.
[0018] In addition, if there are descriptions involving "first", "second", etc. in the embodiments of the present application, the descriptions of "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or suggesting their relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such combination of technical solutions does not exist and is not within the scope of protection required by this application.
[0019] It should be noted that the long jump evaluation method provided by the present application can be widely used in any place with conditions such as sports competitions, sports training and physical fitness tests, such as schools, sports clubs or sports teams, etc., and can also be used to evaluate multiple people standing long jump at the same time. For example, the image acquisition unit used to collect images of each target evaluation object during the long jump process is set at the rear of the long jump evaluation site, and the image acquisition range of the image acquisition unit covers the entire long jump evaluation site; wherein, the long jump evaluation site can be provided with multiple long jump evaluation units, and each target evaluation object completes the long jump evaluation in one of the long jump evaluation units. Since the image acquisition range of the image acquisition unit covers the entire long jump evaluation site, even in a scene where multiple people are long jumping at the same time, the image information of each long jump evaluation personnel during the long jump process can be collected, and the evaluation method provided by the present application is used to realize the standing long jump evaluation of multiple people at the same time.
[0020] See also Figure 1 , Figure 1 It is a flow chart of an embodiment of the long jump evaluation method provided by the present application. It should be noted that if there are substantially the same results, this embodiment does not Figure 1 The process sequence shown is limited. Figure 1 As shown, this embodiment includes:
[0021] S11: Acquire a plurality of first target images of the target evaluation object.
[0022] Among them, in the long jump evaluation, it is necessary to track the same target evaluation object in each first target image to determine the movement of the target evaluation object from take-off to landing, to ensure that the target is not lost during the entire long jump evaluation process of the target evaluation object.
[0023] This embodiment is used to determine an approximate area (candidate area) that at least includes the target evaluation object from each collected first target image of the target evaluation object, and then determine the position of the target evaluation object from the approximate area of each first target image, so as to quickly and accurately track the target evaluation object during the long jump process, and then use the tracked position of the target evaluation object in each first target image (first target position) to determine the final evaluation result.
[0024] The target evaluation object is any evaluation object to be evaluated for long jump, which may be, but is not limited to, an athlete, a student, or a physical fitness tester, etc. The first target image is an image acquired at any time during the entire evaluation process from the start of the evaluation to the end of the evaluation.
[0025] S12: Determine, from each first target image, a region that at least includes the target evaluation object and use it as a candidate region.
[0026] In this embodiment, trajectory prediction can be performed based on the position of the target evaluation object at historical moments, and the approximate position of the target evaluation object at the current moment can be predicted, and the approximate position can be used as a candidate area.
[0027] Specifically, at the current tracking moment, the first target image acquired at the current moment can be used as the current image, and one of the first target images acquired at the historical moment can be used as the reference image, and then the trajectory prediction is performed based on the second target position of the target evaluation object in the reference image to obtain the trajectory prediction result, and then based on the trajectory prediction result, the candidate area in the current image that at least contains the target evaluation object is determined. Exemplarily, after determining the trajectory prediction result of the target evaluation object, the area of the predicted trajectory can be expanded so that the determined candidate area contains the target evaluation object.
[0028] Among them, the reference image is the first target image of the target evaluation object at any historical moment between the take-off moment and the current moment, and the second target position of the target evaluation object in the reference image is the first target position of the target evaluation object in the corresponding first target image. In simple terms, the second target position of the target evaluation object in the reference image and the first target position of the target evaluation object in the first target image both refer to the position of the target evaluation object in the image, that is, for the same image, the second target position of the target evaluation object in the image and the first target position of the target evaluation object in the image are the same position.
[0029] In addition, the candidate areas determined from each first target image include at least one candidate object, and one of the at least one candidate object is a target evaluation object that needs to be tracked.
[0030] S13: Determine the first target position where the target evaluation object is located from the candidate areas of each first target image.
[0031] In an optional embodiment, taking into account the differences between different objects and that, generally, during a long jump test, the tester needs to wait in the take-off area for accurate take-off instructions, the target test object located in the take-off area can be used as a reference object, and in the process of determining the first target position of the target evaluation object from the candidate area, the person with the highest feature similarity to the reference object is selected as the target tester.
[0032] Exemplarily, human figure detection can be performed on the take-off area. If an object is detected in the take-off area and has not left within a certain period of time, the object in the take-off area can be determined as a target evaluation object, and the first target image captured when the target evaluation object is in the take-off area can be used as a reference object.
[0033] In another optional embodiment, considering that during the long jump process, the posture difference of the same person in a short time range is small, the first target image obtained at a historical moment close to the current moment can be used as a reference image. In the process of determining the first target position where the target evaluation object is located from the candidate area, the target evaluation object in the current image is determined based on the difference between the posture of the target evaluation object in the current image and the posture in the reference image.
[0034] Specifically, see Figure 2 , Figure 2 yes Figure 1 The flowchart of step S13 of an embodiment is shown. In this embodiment, determining the first target position where the target evaluation object is located from the candidate areas of each first target image respectively includes:
[0035] S21: For the candidate area corresponding to the current image, the posture evaluation result of the target evaluation object in the reference image is used as the reference posture, and the posture evaluation results corresponding to each candidate object in the candidate area are used as the candidate posture.
[0036] Among them, the current image is the first target image obtained at the current moment, and the reference image is the first target image obtained at a historical moment. Preferably, the time difference between the historical moment and the current moment is less than a preset difference threshold. The specific preset difference threshold can be determined according to the tracking effect of the target evaluation object.
[0037] S22: Selecting a target pose from each candidate pose based on a first difference between each candidate pose and the reference pose.
[0038] In this embodiment, since the posture difference of the same target evaluator during the long jump process should be the smallest, the smallest of the first differences between the posture and the reference posture can be used as the target posture, and the target difference between the target posture and the reference posture should be less than a first preset threshold. The preset threshold is to avoid mistaking two different objects that are not similar as the same object.
[0039] S23: Determine the candidate object whose posture evaluation result is the target posture as the target evaluation object, and use the position of the target evaluation object in the candidate area as the first target position corresponding to the current image.
[0040] In summary, Figure 2 The illustrated embodiment is for determining a target evaluation object for target tracking by utilizing the difference between the postures of objects in two frames of images whose time difference is within a preset range.
[0041] S14: Determine the evaluation result of the target evaluation object based on the first target position corresponding to each first target image.
[0042] In this embodiment, the evaluation result of the target evaluation object may be determined in combination with the distance between the take-off position and the landing position of the target evaluation object or the distance between the landing position and the take-off line.
[0043] In the case of determining the take-off position, it is necessary to predetermine the take-off time of the target evaluation object during the entire evaluation process. In the case of determining the landing position, it is necessary to predetermine the landing time of the target evaluation object during the entire evaluation process. Then, based on the posture of the target evaluation object at the take-off time and / or the landing time, determine the evaluation result of the target evaluation object. The specific implementation method of determining the evaluation result of the target evaluation object can be found in the following Figure 3 Related description of the illustrated embodiment.
[0044] In the above scheme, before determining the first target position of the target evaluation object in each first target image, first determine the area at least containing the target evaluation object from each first target image and use it as a candidate area, and then determine the first target position where the target evaluation object is located from the candidate area of each first target image. It can be seen that the present application first determines an approximate area containing the target evaluation object from the entire first target image, and uses the approximate area as a candidate area, and then determines the first target position where the target evaluation object is located from the candidate area. Compared with the method of directly determining the first target position from the entire first target image, the above method of the present application can effectively reduce the search area of the target evaluation object, so it can improve the detection efficiency of the subsequent target evaluation object position. Furthermore, on the basis of reducing the search area, the present application can effectively reduce the interference of information in some areas that are not related to the target evaluation object, so it can improve the position determination result of the target evaluation object, and then can improve the subsequent evaluation results.
[0045] See also Figure 3 , Figure 3 yes Figure 1 The flowchart of step S14 is shown as an embodiment. Figure 3 As shown, this embodiment includes:
[0046] S31: Determine the first posture of the target evaluation object at each first target position.
[0047] This embodiment is used to determine the first posture of the target evaluation object at each first target position, and then determine the second target image corresponding to the target evaluation moment from a plurality of first target images based on each first posture, and then determine the evaluation result of the target evaluation object based on the second posture of the target evaluation object in the second target image. The first target image is an image captured at any time during the process of the target evaluation object from the take-off area to the completion of the long jump evaluation.
[0048] Among them, the above-mentioned postures and the postures recorded in this embodiment can be obtained by processing the target evaluation object at the first target position in each image using a trained posture estimation model. In a specific embodiment, the above-mentioned posture estimation model is a SimDR single-person posture estimation framework based on a one-dimensional heat map paradigm, which is obtained by mixed training using the MS COCO public data set (17 points) and the labeled long jump scene human skeleton point data (30 points). The posture evaluation model obtained can output the data of 30 human skeleton points.
[0049] In one embodiment, the target evaluation time includes at least one of the take-off time and the landing time of the target evaluation subject; of course, in some embodiments, the target evaluation time also includes the highest point time, so that the highest point can be used to analyze the height change of the target evaluation subject during the long jump process and evaluate the technical level of the long jump.
[0050] S32: Based on each first posture, determine a second target image corresponding to the target evaluation moment from a plurality of first target images.
[0051] In one embodiment, considering that there is a certain regularity between the postures of the target evaluation object before and after different target evaluation moments during the long jump process, the corresponding evaluation conditions can be pre-set based on the content included in the target evaluation moment, and then the corresponding target evaluation moment can be determined based on the difference in the posture of the target evaluation object at any adjacent moments during the long jump process.
[0052] Specifically, see Figure 4 , Figure 4 yes Figure 3 The flowchart of step S32 is shown as an embodiment. Figure 4 As shown, this embodiment includes:
[0053] S41: Determine a second difference between first postures corresponding to first target images acquired at adjacent moments.
[0054] S42: Based on the second difference between each adjacent moment, determine the target evaluation moment and the corresponding second target image.
[0055] In this embodiment, when the target evaluation moment includes the take-off moment, the first image pair corresponding to the first adjacent moment whose second difference satisfies the take-off judgment condition can be found from the first target images of each adjacent moment, and the first moment in front of the first adjacent moment is taken as the take-off moment, and the first target image captured at the take-off moment is taken as the corresponding second target image.
[0056] In one embodiment, the first posture includes the posture of the foot bone point of the target evaluation object, and the take-off judgment conditions include: the second difference is greater than the second preset threshold, and the foot bone point of the target evaluation object is located on the side of the take-off line close to the take-off area at the first moment, and is located on the side of the take-off line away from the take-off area at the third moment after the first adjacent moment.
[0057] When the target evaluation moment includes the landing moment, find the second image pair corresponding to the second adjacent moment whose second difference meets the landing judgment condition from the first target images of each adjacent moment, take the front second moment of the second adjacent moments as the landing moment, and take the first target image collected at the landing moment as the corresponding second target image.
[0058] In a specific embodiment, the second difference includes a first sub-difference and a second sub-difference, and the first sub-difference and the second sub-difference are respectively used to characterize the posture difference between the left foot bone points and the right foot bone points of the target evaluation object at the second adjacent moment; the landing judgment condition includes: at least one of the first sub-difference and the second sub-difference is less than the third preset threshold.
[0059] Among them, the second preset threshold is greater than the third preset threshold.
[0060] In simple terms, the logic for determining the take-off time is as follows: determine whether to take off based on the degree of posture deviation of the target evaluation object at the adjacent time and whether the adjacent time has crossed the take-off line. In order to improve the accuracy of subsequent evaluation results, the earlier time among the adjacent times determined to correspond to the take-off is determined as the take-off time.
[0061] The judgment logic of the landing time is as follows: whether the posture difference (position difference) of the same bone point of the same foot of the target evaluation object at adjacent moments is less than a certain threshold (third preset threshold). That is, when it is judged that a certain foot is no longer moving significantly, it can be judged that the target evaluation person has landed.
[0062] In one implementation scenario, the acquisition unit of the first target image is located behind the long jump evaluation site; the second preset threshold is negatively correlated with the first distance, and the third preset threshold is negatively correlated with the second distance; the first distance is the distance between the take-off position of the target evaluation object at the take-off moment and the acquisition unit, and the second distance is the distance between the landing point position of the target evaluation object at the landing moment and the acquisition unit.
[0063] In short, when setting the second preset threshold and the third preset threshold, it is based on the distance between the target evaluation object and the acquisition unit. It can be understood that the farther the target evaluation object is from the acquisition unit, the denser the scale lines in the long jump test area, and the smaller the distance between two adjacent scale lines. Therefore, in this embodiment, different second preset thresholds and third preset thresholds are set according to the distance between the target evaluation object and the acquisition unit, which can effectively improve the accuracy of the evaluation.
[0064] It should be noted that, in one embodiment, each long jump evaluation unit includes a long jump test area consisting of a take-off line and a farthest scale line, and the scale lines between the take-off line and the farthest scale line in the long jump test area are determined by a linear interpolation algorithm. That is, in the present embodiment, only the take-off line and the farthest scale line (e.g., a scale line of 300 cm) are real scale lines in the long jump test area of each long jump evaluation unit, and other scale lines are not visible in the venue. However, when the long jump evaluation mode is turned on, the evaluation system can interpolate and determine the scale lines located in the middle according to the take-off line and the farthest scale line, and the user can see each scale line in the display screen of the evaluation system accordingly.
[0065] For example, in a specific scenario, there are only 0 and 300 cm scale lines in the venue. A linear interpolation algorithm can be used to obtain a scale distribution with an interval of 5 cm. In the subsequent process of determining the touchdown point position (take-off position and landing position) (taking the determination of the landing position as an example), it is possible to first determine that the heel point is within a certain interval, and make an initial large evaluation result interval determination. Then, based on the distance between the heel point and the left and right or front and back scale lines of the interval, the evaluation results in the small interval are calculated according to the strategy of equal proportion division. Finally, the final evaluation result is determined by accumulating the large and small interval evaluation results.
[0066] S33: Determine an evaluation result of the target evaluation object based on a second posture of the target evaluation object in the second target image.
[0067] In this embodiment, the touchdown point position can be determined based on the second posture of the target evaluation object in the second target image; the distance between the take-off and the landing of the target evaluation object can be determined based on the touchdown point position; and the distance between the take-off and the landing can be used as the evaluation result of the target evaluation object. The touchdown point position includes at least one of the take-off position and the landing position.
[0068] Specifically, see Figure 5 , Figure 5 yes Figure 3 The flowchart of step S33 is shown as an embodiment. Figure 5 As shown, this embodiment includes:
[0069] S51: Determine a touchdown point position based on a second posture of the target evaluation object in the second target image, where the touchdown point position includes at least one of a take-off position and a landing position.
[0070] In this embodiment, the second posture includes the foot area of the target evaluation object.
[0071] In one embodiment, the target processing model can be used to process the foot region to obtain at least one of the heel point position and the toe point position of the target evaluation object; the landing position is the heel point position in the second target image corresponding to the landing moment, and the take-off position is the toe point position in the second target image corresponding to the take-off moment. That is, in this embodiment, the heel point position at the landing moment is used as the landing position, and the toe point position at the take-off moment is used as the take-off position.
[0072] Among them, during training, the target processing model needs to mark the heel point and / or toe point according to the prediction, and then uses the marked information to train the target processing model, so that the trained target processing model can directly output the position of the heel point and / or toe point.
[0073] S52: Determine the distance between take-off and landing of the target evaluation subject based on the touchdown point position.
[0074] Among them, when the touchdown point position includes the landing position but does not include the take-off position, the distance between the landing position and the take-off line is obtained as the distance between the take-off and the landing; when the touchdown point position includes the landing position and the take-off position, the distance between the landing position and the take-off position is obtained as the distance between the take-off and the landing.
[0075] S53: The distance between take-off and landing is used as the evaluation result of the target evaluation object.
[0076] In one implementation scenario, the second posture includes the foot area and the hip area of the target evaluation object, and the landing position is the landing position of the heel point in the foot area. In this implementation scenario, before step S53, at least the following two steps are also included:
[0077] First, based on the distance between the foot area and the hip area, determine whether the landing position is blocked.
[0078] If the distance between the foot area and the hip area is less than a preset threshold, it is determined that the landing position is blocked; otherwise, it is determined that the landing position is not blocked.
[0079] Second, in response to the landing position being unobstructed, a step is performed to use the distance between take-off and landing as the evaluation result of the target evaluation object.
[0080] Third, in response to the landing position being blocked, send the user confirmation information of the evaluation result, or continue to execute the second difference based on each adjacent moment, determine the target evaluation moment and the corresponding second target image and subsequent steps, until the landing position is not blocked, and then execute the distance between take-off and landing as the evaluation result of the target evaluation object.
[0081] See also Figure 6 , Figure 6 A schematic diagram of the framework of an embodiment of a long jump evaluation device provided in the present application. In this embodiment, the long jump evaluation device 60 includes an acquisition module 61, a first determination module 62, a second determination module 63 and a result determination module 64. The acquisition module 61 is used to acquire a plurality of first target images of the target evaluation object; the first determination module 62 is used to respectively determine an area containing at least the target evaluation object from each first target image, and use it as a candidate area; the second determination module 63 is used to respectively determine the first target position where the target evaluation object is located from the candidate area of each first target image; the result determination module 64 is used to determine the evaluation result of the target evaluation object based on the first target position corresponding to each first target image.
[0082] In some embodiments, the first determination module 62 determines an area that at least contains the target evaluation object from each first target image, and uses it as a candidate area, including: taking the first target image acquired at a current moment as the current image, and taking one of the first target images acquired at a historical moment as a reference image; performing trajectory prediction based on the second target position of the target evaluation object in the reference image to obtain a trajectory prediction result; based on the trajectory prediction result, determining a candidate area in the current image that at least contains the target evaluation object.
[0083] In some embodiments, the candidate area in the current image includes at least one candidate object; the second determination module 63 determines the first target position where the target evaluation object is located from the candidate areas of each first target image, including: for the candidate area corresponding to the current image, the posture evaluation result of the target evaluation object in the reference image is used as the reference posture, and the posture evaluation results corresponding to each candidate object in the candidate area are used as candidate postures, wherein the current image is the first target image acquired at the current moment, and the reference image is the first target image acquired at a historical moment; based on the first difference between each candidate posture and the reference posture, the target posture is selected from each candidate posture; the candidate object whose posture evaluation result is the target posture is determined as the target evaluation object, and the position of the target evaluation object in the candidate area is used as the first target position corresponding to the current image.
[0084] In some embodiments, the target difference between the target posture and the reference posture is the smallest of the first differences, and the target difference is less than a first preset threshold; and / or, the reference image is the first target image of the target evaluation object at any historical moment between the take-off moment and the current moment; the second target position of the target evaluation object in the reference image is the first target position of the target evaluation object in the corresponding first target image.
[0085] In some embodiments, the result determination module 64 determines the evaluation result of the target evaluation object based on the first target position corresponding to each first target image, including: determining the first posture of the target evaluation object at each first target position; based on each first posture, determining the second target image corresponding to the target evaluation moment from a number of first target images; based on the second posture of the target evaluation object in the second target image, determining the evaluation result of the target evaluation object.
[0086] In some embodiments, the first target image is an image of the target evaluation object captured at any moment in the process from the take-off area to the completion of the long jump evaluation; based on each first posture, the second target image corresponding to the target evaluation moment is determined from a number of first target images, including: determining the second difference between the first postures corresponding to the first target images captured at each adjacent moment; based on the second difference at each adjacent moment, determining the target evaluation moment and the corresponding second target image.
[0087] In some embodiments, the target evaluation moment includes at least one of the take-off moment and the landing moment of the target evaluation object; based on the second difference of each adjacent moment, the target evaluation moment and the corresponding second target image are determined, including: when the target evaluation moment includes the take-off moment, from the first target images of each adjacent moment, the first image pair corresponding to the first adjacent moment whose second difference meets the take-off judgment condition is found, the first moment closer to the front of the first adjacent moment is taken as the take-off moment, and the first target image captured at the take-off moment is taken as the corresponding second target image; and / or, when the target evaluation moment includes the landing moment, from the first target images of each adjacent moment, the second image pair corresponding to the second adjacent moment whose second difference meets the landing judgment condition is found, the second moment closer to the front of the second adjacent moment is taken as the landing moment, and the first target image captured at the landing moment is taken as the corresponding second target image.
[0088] In some embodiments, the first posture includes the posture of the foot bone points of the target evaluation object; the take-off judgment condition includes: the second difference is greater than the second preset threshold, and the foot bone points of the target evaluation object are located on the side of the take-off line close to the take-off area at the first moment, and are located on the side of the take-off line away from the take-off area at the third moment after the first adjacent moment; the second difference includes a first sub-difference and a second sub-difference, and the first sub-difference and the second sub-difference are respectively used to characterize the posture difference between the left foot bone points and the right foot bone points of the target evaluation object at the second adjacent moment; the landing judgment condition includes: at least one of the first sub-difference and the second sub-difference is less than the third preset threshold; wherein the second preset threshold is greater than the third preset threshold.
[0089] In some embodiments, the acquisition unit of the first target image is located behind the long jump evaluation site; the second preset threshold is negatively correlated with the first distance, and the third preset threshold is negatively correlated with the second distance; the first distance is the distance between the take-off position of the target evaluation object at the take-off moment and the acquisition unit, and the second distance is the distance between the landing point position of the target evaluation object at the landing moment and the acquisition unit.
[0090] In some embodiments, the target evaluation moment includes at least one of the take-off moment and the landing moment of the target evaluation object; based on the second posture of the target evaluation object in the second target image, the evaluation result of the target evaluation object is determined, including: based on the second posture of the target evaluation object in the second target image, the touchdown point position is determined, and the touchdown point position includes at least one of the take-off position and the landing position; based on the touchdown point position, the distance between the take-off and the landing of the target evaluation object is determined; and the distance between the take-off and the landing is used as the evaluation result of the target evaluation object.
[0091] In some embodiments, the second posture includes the foot area of the target evaluation object; based on the second posture of the target evaluation object in the second target image, the touchdown point position is determined, including: using the target processing model to process the foot area to obtain at least one of the heel point position and the toe position of the target evaluation object; the landing position is the heel point position in the second target image corresponding to the landing moment, and the take-off position is the toe position in the second target image corresponding to the take-off moment; and / or, based on the touchdown point position, the distance between the take-off and the landing of the target evaluation object is determined, including: when the touchdown point position includes the landing position but does not include the take-off position, obtaining the distance between the landing position and the take-off line as the distance between the take-off and the landing; when the touchdown point position includes the landing position and the take-off position, obtaining the distance between the landing position and the take-off position distance as the distance between take-off and landing; and / or, the second posture includes the foot area and hip area of the target evaluation object, and the landing position is the landing position of the heel point in the foot area; before taking the distance between take-off and landing as the evaluation result of the target evaluation object, the method also includes: based on the distance between the foot area and the hip area, determining whether the landing position is blocked; in response to the landing position being not blocked, executing the step of taking the distance between take-off and landing as the evaluation result of the target evaluation object; in response to the landing position being blocked, sending an evaluation result confirmation message to the user, or continuing to execute the second difference based on each adjacent moment, determining the target evaluation moment and the corresponding second target image and subsequent steps, until in response to the landing position being not blocked, executing the distance between take-off and landing as the evaluation result of the target evaluation object.
[0092] In some embodiments, the acquisition unit of the first target image is arranged at the rear of the long jump evaluation venue, and the image acquisition range of the acquisition unit covers the long jump evaluation venue; the long jump evaluation venue is provided with a plurality of long jump evaluation units, and each target evaluation subject completes the long jump evaluation in one of the long jump evaluation units; each long jump evaluation unit includes a long jump test area consisting of a take-off line and a farthest scale line, and the scale lines in the long jump test area located between the take-off line and the farthest scale line are determined by a linear interpolation algorithm.
[0093] See also Figure 7 , Figure 7 Schematic diagram of the framework of an electronic device provided by the present application. In this embodiment, the electronic device 70 includes a memory 71 and a processor 72 coupled to each other.
[0094] The memory 71 stores program instructions, and the processor 72 is used to execute the program instructions stored in the memory 71 to implement the steps of any of the above method implementations. In a specific implementation scenario, the electronic device 70 may include, but is not limited to: a microcomputer, a server, and in addition, the electronic device 70 may also include a mobile device such as a laptop computer and a tablet computer, which is not limited here.
[0095] Specifically, the processor 72 is used to control itself and the memory 71 to implement the steps of any of the above-mentioned embodiments. The processor 72 can also be called a CPU (Central Processing Unit). The processor 72 may be an integrated circuit chip with signal processing capabilities. The processor 72 can also be a general-purpose processor, a digital signal processor (Digital Signal Processor, DSP), an application-specific integrated circuit (Application Specific Integrated Circuit, ASIC), a field-programmable gate array (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. In addition, the processor 72 can be implemented by an integrated circuit chip.
[0096] See also Figure 8 , Figure 8 It is a schematic diagram of the framework of the computer-readable storage medium provided by the present application. The computer-readable storage medium 80 of the embodiment of the present application stores a program instruction 81, and when the program instruction 81 is executed, the method provided by any embodiment of the above method and any non-conflicting combination is implemented. Among them, the program instruction 81 can form a program file and be stored in the above-mentioned computer-readable storage medium 80 in the form of a software product, so that a computer device (which can be a personal computer, a server, or a network device, etc.) executes all or part of the steps of each implementation method of the present application. The aforementioned computer-readable storage medium 80 includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a disk or an optical disk, or a terminal device such as a computer, a server, a mobile phone, and a tablet.
[0097] In the above scheme, before determining the first target position of the target evaluation object in each first target image, first determine the area at least containing the target evaluation object from each first target image and use it as a candidate area, and then determine the first target position where the target evaluation object is located from the candidate area of each first target image. It can be seen that the present application first determines an approximate area containing the target evaluation object from the entire first target image, and uses the approximate area as a candidate area, and then determines the first target position where the target evaluation object is located from the candidate area. Compared with the method of directly determining the first target position from the entire first target image, the above method of the present application can effectively reduce the search area of the target evaluation object, so it can improve the detection efficiency of the subsequent target evaluation object position. Furthermore, on the basis of reducing the search area, the present application can effectively reduce the interference of information in some areas that are not related to the target evaluation object, so it can improve the position determination result of the target evaluation object, and then can improve the subsequent evaluation results.
[0098] In some embodiments, the functions or modules included in the device provided by the embodiments of the present disclosure can be used to execute the method described in the above method embodiments. The specific implementation can refer to the description of the above method embodiments, and for the sake of brevity, it will not be repeated here.
[0099] The above description of various embodiments tends to emphasize the differences between the various embodiments. The same or similar aspects can be referenced to each other, and for the sake of brevity, they will not be repeated herein.
[0100] In the several embodiments provided in the present application, it should be understood that the disclosed methods and devices can be implemented in other ways. For example, the device implementation described above is only schematic. For example, the division of modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, and the indirect coupling or communication connection of devices or units can be electrical, mechanical or other forms.
[0101] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.
[0102] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.
[0103] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) or a processor (processor) to perform all or part of the steps of each implementation method of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program code.
[0104] The above description is only an implementation method of the present application, and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly used in other related technical fields, are also included in the patent protection scope of the present application.
Claims
1. A long jump evaluation method, It is characterized in that The method comprises: Acquire a plurality of first target images of the target evaluation object; Determine, from each of the first target images, a region at least including the target evaluation object and use it as a candidate region; Determining the first target position where the target evaluation object is located from the candidate areas of each of the first target images; Based on the first target position corresponding to each of the first target images, an evaluation result of the target evaluation object is determined.
2. The method according to claim 1, It is characterized in that The determining of a region at least including the target evaluation object from each of the first target images as a candidate region includes: Taking the first target image acquired at the current moment as the current image, and taking one of the first target images acquired at the historical moment as the reference image; Performing trajectory prediction based on a second target position of the target evaluation object in the reference image to obtain a trajectory prediction result; Based on the trajectory prediction result, determine the candidate area in the current image that at least includes the target evaluation object.
3. The method according to claim 1 or 2, It is characterized in that The candidate area in the current image includes at least one candidate object; The determining, from the candidate areas of each of the first target images, the first target position where the target evaluation object is located includes: For the candidate area corresponding to the current image, the posture evaluation result of the target evaluation object in the reference image is used as the reference posture, and the posture evaluation results corresponding to each candidate object in the candidate area are used as candidate postures, wherein the current image is the first target image acquired at the current moment, and the reference image is the first target image acquired at a historical moment; Selecting a target posture from each of the candidate postures based on a first difference between each of the candidate postures and the reference posture; A candidate object whose posture evaluation result is the target posture is determined as the target evaluation object, and the position of the target evaluation object in the candidate area is used as the first target position corresponding to the current image.
4. The method according to claim 3, It is characterized in that A target difference between the target posture and the reference posture is the smallest of the first differences, and the target difference is less than a first preset threshold; And / or, the reference image is the first target image of the target evaluation object at any historical moment between the take-off moment and the current moment; The second target position of the target evaluation object in the reference image is the first target position of the target evaluation object in the corresponding first target image.
5. The method according to claim 1, It is characterized in that The determining the evaluation result of the target evaluation object based on the first target position corresponding to each of the first target images includes: Determining a first posture of the target evaluation object at each of the first target positions; Based on each of the first postures, determining a second target image corresponding to the target evaluation moment from the plurality of first target images; Based on the second posture of the target evaluation object in the second target image, an evaluation result of the target evaluation object is determined.
6. The method according to claim 5, It is characterized in that The first target image is an image of the target evaluation subject acquired at any time during the process from the take-off area to the completion of the long jump evaluation; The determining, based on each of the first postures, a second target image corresponding to the target evaluation moment from the plurality of first target images comprises: Determine a second difference between first postures corresponding to first target images acquired at adjacent moments; Based on the second difference between each of the adjacent moments, the target evaluation moment and the corresponding second target image are determined.
7. The method according to claim 6, It is characterized in that The target evaluation time includes at least one of the take-off time and the landing time of the target evaluation object; The determining the target evaluation moment and the corresponding second target image based on the second difference between each of the adjacent moments includes: In the case where the target evaluation moment includes the take-off moment, finding the first image pair corresponding to the first adjacent moment at which the second difference meets the take-off judgment condition from the first target images of each adjacent moment, taking the first moment closer to the front of the first adjacent moments as the take-off moment, and taking the first target image captured at the take-off moment as the corresponding second target image; and / or, When the target evaluation moment includes the landing moment, find the second image pair corresponding to the second adjacent moment in which the second difference satisfies the landing judgment condition from the first target images of each adjacent moment, take the front second moment of the second adjacent moments as the landing moment, and take the first target image captured at the landing moment as the corresponding second target image.
8. The method according to claim 7, It is characterized in that The first posture includes the posture of the foot bone points of the target evaluation object; The take-off judgment condition includes: the second difference is greater than a second preset threshold, and the foot bone point of the target evaluation subject is located on the side of the take-off line close to the take-off area at the first moment, and is located on the side of the take-off line away from the take-off area at the third moment after the first adjacent moments; The second difference includes a first sub-difference and a second sub-difference, and the first sub-difference and the second sub-difference are respectively used to characterize the posture difference between the left foot bone points and the right foot bone points of the target evaluation object at the second adjacent moment; the landing judgment condition includes: at least one of the first sub-difference and the second sub-difference is less than a third preset threshold; Wherein, the second preset threshold is greater than the third preset threshold.
9. The method according to claim 8, It is characterized in that The first target image acquisition unit is located at the rear of the long jump evaluation site; The second preset threshold is negatively correlated with the first distance, and the third preset threshold is negatively correlated with the second distance; the first distance is the distance between the take-off position of the target evaluation object at the take-off moment and the collection unit, and the second distance is the distance between the landing point position of the target evaluation object at the landing moment and the collection unit.
10. The method according to claim 5, It is characterized in that The target evaluation time includes at least one of the take-off time and the landing time of the target evaluation object; The determining, based on the second posture of the target evaluation object in the second target image, an evaluation result of the target evaluation object comprises: Determine a touchdown point position based on a second posture of the target evaluation object in the second target image, wherein the touchdown point position includes at least one of a take-off position and a landing position; Determine the distance between the take-off and the landing of the target evaluation subject based on the touchdown point position; The distance between the take-off and the landing is used as the evaluation result of the target evaluation object.
11. The method according to claim 10, It is characterized in that The second posture includes a foot region of the target evaluation object; the determining of the touchdown point position based on the second posture of the target evaluation object in the second target image includes: processing the foot region using a target processing model to obtain at least one of a heel point position and a toe position of the target evaluation object; the landing position is the heel point position in the second target image corresponding to the landing moment, and the take-off position is the toe position in the second target image corresponding to the take-off moment; And / or, determining the distance between take-off and landing of the target evaluation subject based on the touchdown point position includes: when the touchdown point position includes the landing position but does not include the take-off position, obtaining the distance between the landing position and the take-off line as the distance between take-off and landing; when the touchdown point position includes the landing position and the take-off position, obtaining the distance between the landing position and the take-off position as the distance between take-off and landing; And / or, the second posture includes the foot area and hip area of the target evaluation object, and the landing position is the landing position of the heel point in the foot area; before taking the distance between the take-off and the landing as the evaluation result of the target evaluation object, the method further includes: determining whether the landing position is blocked based on the distance between the foot region and the hip region; In response to the landing position being unobstructed, executing the step of using the distance between the take-off and the landing as the evaluation result of the target evaluation object; In response to the landing position being blocked, an evaluation result confirmation message is sent to the user, or the second difference based on adjacent moments is continued to be executed to determine the target evaluation moment and the corresponding second target image and subsequent steps, until in response to the landing position being unblocked, the distance between the take-off and the landing is used as the evaluation result of the target evaluation object.
12. The method according to claim 1, It is characterized in that The first target image acquisition unit is arranged at the rear of the long jump evaluation site, and the image acquisition range of the acquisition unit covers the long jump evaluation site; The long jump evaluation venue is provided with a plurality of long jump evaluation units, and each of the target evaluation subjects completes the long jump evaluation in one of the long jump evaluation units; each of the long jump evaluation units includes a long jump test area consisting of a take-off line and a farthest scale line, and the scale lines in the long jump test area located between the take-off line and the farthest scale line are determined by a linear interpolation algorithm.
13. A long jump evaluation device, It is characterized in that The device comprises: An acquisition module, used to acquire a plurality of first target images of the target evaluation object; A first determination module is used to determine, from each of the first target images, a region at least including the target evaluation object, and use it as a candidate region; A second determination module is used to determine the first target position where the target evaluation object is located from the candidate areas of each of the first target images; A result determination module is used to determine the evaluation result of the target evaluation object based on the first target position corresponding to each of the first target images.
14. An electronic device, It is characterized in that comprising a memory and a processor coupled to each other, The memory stores program instructions; The processor is used to execute the program instructions stored in the memory to implement the method according to any one of claims 1 to 12.
15. A computer-readable storage medium, It is characterized in that The computer-readable storage medium stores program instructions that can be run by a processor, and the program instructions can be executed by the processor to implement the method according to any one of claims 1 to 12.