Object motion analysis method, apparatus, device, and storage medium

Abstract

The application provides a kind of object motion analysis method, device, equipment and storage medium, the method comprises: obtaining the at least two frames of images of court area in preset period, and the object in each frame image is carried out target detection, determines the target detection result corresponding to each frame image;Each target detection result includes the first position of target object in the court area;According to the first position of the target object in the court area, determine the virtual window frame that the target object occupies in the court area in the preset period;Wherein, the court area is divided into at least two virtual window frames in advance in the preset period;According to the number of the window frame of the target object, determine the motion distance of the target object in the preset period.The technical scheme of the application can save the time and manpower cost of player motion analysis, while the accuracy and objectivity of motion analysis result can be guaranteed.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to an object motion analysis method, apparatus, device, and storage medium. Background Technology

[0002] Football is one of the most popular sports today. In football matches, a player's running distance is a crucial indicator of their physical fitness, competitive state, and tactical execution ability. Therefore, accurately analyzing the distance players travel during a match has become an indispensable part of modern football training and tactical planning.

[0003] In related technologies, the analysis of data such as the movement distance of players mainly relies on manual observation and recording, that is, by manually monitoring each player during a football match to obtain the movement distance of each player.

[0004] However, the above methods are not only time-consuming and labor-intensive, but also easily affected by subjective factors, making it difficult to guarantee the accuracy and objectivity of the motion analysis results. Summary of the Invention

[0005] To address the problems existing in the prior art, the present invention provides a method, apparatus, device, and storage medium for object motion analysis, which can save time and manpower costs in player motion analysis, while ensuring the accuracy and objectivity of the motion analysis results.

[0006] This invention provides a method for object motion analysis, comprising:

[0007] Acquire at least two frames of images of the court area within a preset time period, perform target detection on objects in each frame, and determine the target detection result corresponding to each frame; each target detection result includes the first position of the target object in the court area;

[0008] Based on the target object's first position in the court area, determine the virtual panes occupied by the target object in the court area within a preset time period; wherein, the court area is pre-divided into at least two virtual panes within the preset time period;

[0009] Based on the pane distance information of the virtual pane occupied by the target object, the movement distance of the target object within the preset time period is determined.

[0010] According to the object motion analysis method provided by the present invention, the above-mentioned determination of the virtual panes occupied by the target object in the court area within a preset time period based on the first positions of the target object in the court area includes:

[0011] Based on the pre-defined correspondence between the positions and numbers of the virtual panes, the target number corresponding to each first position is determined; each first position corresponds to the position of a virtual pane.

[0012] The virtual pane corresponding to each target number is determined as the virtual pane occupied by the target object.

[0013] According to an object motion analysis method provided by the present invention, determining the motion distance of the target object within the preset time period based on the pane distance information of the virtual pane occupied by the target object includes:

[0014] Based on the historical motion data corresponding to the court area, determine the pane distance information corresponding to the virtual pane; the aforementioned historical motion data includes the historical step distance of each historical object in each team at different times of each game and under different sports types.

[0015] The number of different target numbers in each target number is counted to obtain the number of virtual panes occupied by the target object within a preset time period;

[0016] Based on the number of panes and the distance between panes of the target object, determine the movement distance of the target object within a preset time period.

[0017] According to the object motion analysis method provided by the present invention, the aforementioned pane distance information includes pane width, and the determination of pane distance information corresponding to virtual panes based on historical motion data corresponding to the court area includes:

[0018] For each team, statistical analysis was conducted on the team's historical stride distance at different times and under different sports types in each game to determine the team's average stride distance at different times and under different sports types in each game;

[0019] Within a preset time period for each game, the duration of each team's movement under different sports types is obtained, and the target step distance corresponding to the preset time period is determined based on the proportion of each movement duration within the preset time period and the average step distance of each movement. The preset time period can be any time period among different time periods.

[0020] The pane width corresponding to the preset time period is determined based on the target step distance for the preset time period; among them, different time periods of each game correspond to different pane widths.

[0021] According to the object motion analysis method provided by the present invention, the acquisition of at least two frames of images of a court area within a preset time period includes:

[0022] For each team, the team's movement frequency within a preset time period is determined based on historical movement data corresponding to the court area; the aforementioned movement frequency is used to represent the average time during which the team engages in cross-pane movement within the preset time period.

[0023] The time interval for motion analysis of objects within the team is determined based on the team's movement frequency;

[0024] Acquire at least two frames of images of the court area within a preset time period according to the time interval.

[0025] According to the object motion analysis method provided by the present invention, the above-mentioned at least two frames include the current frame image and the previous frame image of the current frame image; the above-mentioned determination of the virtual panes occupied by the target object in the court area within a preset time period based on the first positions of the target object in the court area includes:

[0026] If the second position of the virtual pane corresponding to the first position of the target object in the current frame image is different from the third position of the virtual pane corresponding to the first position of the target object in the previous frame image, then it is detected whether the virtual panes corresponding to the second position and the third position are consecutive virtual panes.

[0027] If the virtual panes are not consecutive, all frames between the previous frame and the current frame are acquired and analyzed to determine the motion trajectory of the target object.

[0028] Based on the movement trajectory of the target object, determine the virtual panes occupied by the target object in the court area within a preset time period.

[0029] According to the object motion analysis method provided by the present invention, the method further includes:

[0030] Based on the regional images captured by each image acquisition device within the court area, determine at least two candidate regions constituting the court area, as well as the length and width of each candidate region;

[0031] The overlapping areas in each candidate area are removed, and the actual length and width of the court area are determined based on the length and width of each candidate area after removing the overlapping areas.

[0032] Based on the actual length and width of the court area, the court area is divided into at least two virtual panes within a preset time period.

[0033] The present invention also provides an object motion analysis device, comprising:

[0034] The analysis module is used to acquire at least two frames of images of the court area within a preset time period, and to perform target detection on objects in each frame of images to determine the target detection results corresponding to each frame of images; each target detection result includes the first position of the target object in the court area;

[0035] The occupancy pane determination module is used to determine the virtual panes occupied by the target object in the court area within a preset time period based on the target object's first position in the court area; wherein the court area is pre-divided into at least two virtual panes within the preset time period;

[0036] The movement distance determination module is used to determine the movement distance of the target object within the preset time period based on the pane distance information of the virtual pane occupied by the target object.

[0037] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the object motion analysis method as described above.

[0038] The present invention also provides an object motion analysis system, comprising: an image acquisition device and an electronic device interconnected thereto, wherein the image acquisition device is used to acquire images of a court area and transmit them to the electronic device; the electronic device is used to perform the object motion analysis method as described above based on the images transmitted by the image acquisition device.

[0039] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the object motion analysis method as described above.

[0040] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the object motion analysis method as described above.

[0041] The object motion analysis method, apparatus, device, and storage medium provided by this invention acquire at least two frames of images of a sports field area within a preset time period and perform target detection on objects in each frame to determine the target detection results for each frame. Then, based on the first positions of the target objects in the sports field area in each detection result, and considering that the sports field area is pre-divided into at least two virtual panes within the preset time period, the virtual pane occupied by the target object in the sports field area within the preset time period is determined, and the movement distance of the target object within the preset time period is determined based on the distance information of the occupied pane. In this method, since the acquisition and analysis of images of the sports field area can be automated, and the movement distance is obtained through the distance information of the pane occupied by the target object, manual statistics and recording of movement distances can be avoided, thus saving labor and time costs. Furthermore, since the object motion analysis process does not require manual statistics and recording of movement distances, the influence of subjective factors on the object motion analysis process and results can be avoided, thereby ensuring the accuracy and objectivity of the object motion analysis process and results. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0043] Figure 1 This is a flowchart illustrating the object motion analysis method provided in an embodiment of the present invention;

[0044] Figure 2 This is a schematic diagram of setting up an image acquisition device in the field area according to an embodiment of the present invention;

[0045] Figure 3 This is a schematic diagram of gridding the court area provided in an embodiment of the present invention;

[0046] Figure 4 This is a schematic diagram illustrating the numbering of virtual panes in the court area according to an embodiment of the present invention;

[0047] Figure 5 This is a discrete data graph of the historical stride distance of a team under different sports types, provided in an embodiment of the present invention.

[0048] Figure 6 This is a trend graph of the average stride distance of a team under different sports types, provided in an embodiment of the present invention, over time.

[0049] Figure 7 This is a discrete data graph of the historical stride distance of a team in different time periods and under different sports types, provided by an embodiment of the present invention;

[0050] Figure 8 This is a schematic diagram of the virtual panes occupied by the target object over a period of time, as provided in an embodiment of the present invention.

[0051] Figure 9 This is a schematic diagram of the court area and objects therein provided in an embodiment of the present invention;

[0052] Figure 10 This is one of the schematic diagrams provided in the embodiments of the present invention for acquiring and analyzing images of a sports field area;

[0053] Figure 11 This is a schematic diagram of the numbering of the stored virtual panes provided in an embodiment of the present invention;

[0054] Figure 12 This is a second schematic diagram illustrating the acquisition and analysis of images of a sports field area provided in an embodiment of the present invention;

[0055] Figure 13This is a schematic diagram of images captured by various image acquisition devices of the sports field area according to embodiments of the present invention;

[0056] Figure 14 This is a schematic diagram showing the length and width of each court area provided in the embodiments of the present invention;

[0057] Figure 15 This is a schematic diagram of the overlapping area provided in an embodiment of the present invention;

[0058] Figure 16 This is a schematic diagram of the object motion analysis device provided in an embodiment of the present invention;

[0059] Figure 17 This is a schematic diagram of the structure of the electronic device provided in an embodiment of the present invention. Detailed Implementation

[0060] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of the embodiments of the present invention.

[0061] Currently, the analysis of data such as player movement distance mainly relies on manual observation and recording, that is, monitoring each player's movement distance manually during a football match. However, this method is not only time-consuming and labor-intensive, but also easily affected by subjective factors, making it difficult to guarantee the accuracy and objectivity of the motion analysis results.

[0062] With advancements in technology and the development of sports science, an increasing number of high-tech methods are being introduced into football matches, providing more accurate and comprehensive data support for player movement analysis. For example, in one related technology, a positioning receiver placed at a reference node, a positioning transmitter worn by the player, and a host computer can calculate and locate the player's position. However, this method has high implementation requirements and is easily affected by field movement and player movement, leading to receiver and transmitter malfunctions and positioning failures. Another example is a related technology that uses intelligent analysis to calculate player trajectories through real-time analysis of the field footage. This method requires frame-by-frame analysis of the video footage, necessitating a large number of servers for processing, which is resource-intensive and suffers from data real-time issues. Furthermore, due to changes in player movement distance and direction, this method can also introduce significant analysis errors.

[0063] Based on this, embodiments of the present invention provide an object motion analysis method, apparatus, device, and storage medium that can solve the above-mentioned problems.

[0064] The following is combined Figures 1-15 This invention describes an object motion analysis method according to an embodiment of the present invention.

[0065] It should be noted that the execution subject of the embodiments of the present invention may be an electronic device, an object motion analysis device, an object motion analysis system including an electronic device or an object motion analysis device, or a central control device. The following embodiments will use an electronic device as the execution subject to describe the embodiments of the present invention.

[0066] Figure 1 This is a flowchart illustrating the object motion analysis method provided in an embodiment of the present invention. See also... Figure 1 As shown, the method may include the following steps:

[0067] S102, acquire at least two frames of images of the court area within a preset time period, perform target detection on the objects in each frame of images, and determine the target detection result corresponding to each frame of images; each target detection result includes the first position of the target object in the court area.

[0068] The preset time period can be the entire duration of the match, such as 0-90 minutes, or a specific segment of the match, such as 0-45 minutes, 45-75 minutes, or 75-90 minutes. When the preset time period is a specific segment of the match, the specific division of the preset time period can be set according to the actual situation; the above examples are merely illustrations.

[0069] Furthermore, the playing field area refers to the area within which objects can move during a game. These objects can include players, referees, etc. This embodiment primarily analyzes the movement of players within the playing field area; that is, the objects analyzed in this embodiment are mainly players. There are no specific limitations on the ball pairs corresponding to the players.

[0070] Simultaneously, multiple image acquisition devices can be pre-installed at various locations within the stadium area. These devices can be cameras or video cameras, which can monitor the stadium area or provide live video streaming of matches within the stadium area. Each image acquisition device can capture images of a portion of the stadium area, obtaining an image of that specific area. Then, electronic equipment combines the images from multiple devices to create a composite image of the entire stadium area. See also... Figure 2The diagram shows the setup of image acquisition devices in the court area. The left image is a schematic diagram of the actual court area, and the right image is a line drawing of the corresponding court area. P1-P6 represent the locations of the image acquisition devices deployed in the court area. A coordinate system is established with the center of the court as the center point (as shown in the right image), and the coordinate positions corresponding to points P1-P6 are recorded accordingly. Taking cameras as an example, the relevant parameters of cameras P1-P6 are shown in Table 1 below:

[0071] Table 1

[0072] 1 P1 x1 y1 K1 2 P2 x2 y2 K2 3 P3 x3 y3 K3 4 P4 x4 y4 K4 5 P5 x5 y5 K5 6 P6 x6 y6 K6

[0073] Here, K represents the camera type, and different camera types have different viewing angles / field of view ranges.

[0074] It is understandable that the deployment of 6 image acquisition devices in the above-mentioned stadium area is just an example; in reality, more or fewer image acquisition devices can be deployed in the stadium area.

[0075] After setting up multiple image acquisition devices in the stadium area, images of the stadium area can be acquired through each device. Then, by combining the images acquired by each device at each moment, a corresponding frame image of the stadium area at that moment can be obtained. This method can obtain at least two frames of images of the stadium area within a preset time period. Target detection can then be performed on each frame image to obtain the target detection result for each frame. The target object in the target detection result of each frame image can be one or more specified objects, or it can be that each object is a target object in sequence.

[0076] Object detection here could involve using an object detection algorithm to detect objects in each frame of an image, obtaining the object detection result for each frame. This result could include the detection bounding box containing each object, the position of the bounding box, the dimensions of the bounding box, and the object's identifier (such as a player's jersey number or the player's team). The position of the bounding box containing the target object can be considered the target object's primary location within the field area.

[0077] In addition, the image acquisition device for the aforementioned court area can continuously acquire images of the court area and store them in an electronic device. The electronic device can then retrieve these images and perform motion analysis (specifically, target detection, determination of occupied panes, and calculation of movement distance). Specifically, when acquiring at least two frames of images of the court area within a preset time period and performing motion analysis, it can be done by acquiring all images from the electronic device, or by extracting or acquiring a portion of the images from the electronic device at certain time intervals, or by combining the extraction / acquisition of images at certain time intervals with the acquisition of all images for motion analysis, or other methods. This embodiment does not impose specific limitations.

[0078] S104, based on the first positions of the target object in the court area, determine the virtual panes occupied by the target object in the court area within a preset time period; wherein the court area is pre-divided into at least two virtual panes within the preset time period.

[0079] In this step, after obtaining the image of the court area as described above, a line drawing of the court area can be obtained from the image (as shown above). Figure 2 (As shown in the right-hand image), the court area can then be pre-gridded, dividing it into at least two virtual panes. When dividing the court area into at least two virtual panes, the size of the virtual panes can be adjusted according to different time periods of the game; that is, the size of the virtual panes will change as the game progresses. Alternatively, the court area can be divided with the same virtual pane size, meaning the virtual panes are a fixed size and will not change with the game progress. It is understandable that the size of each virtual pane obtained after gridding the court area is generally the same.

[0080] Specifically, after analyzing each frame of the image within the preset time period to obtain the first position of the target object in the court area, the corresponding virtual grid position of the target object in the virtual grid in each frame of the image can also be obtained according to the gridded court area, thereby obtaining the virtual pane occupied by the target object in the court area within the preset time period.

[0081] S106, determine the movement distance of the target object within the preset time period based on the pane distance information of the virtual pane occupied by the target object.

[0082] The virtual panes are typically square panes (also known as grids). The pane distance information of the occupied virtual panes can be the pane distance information of each occupied virtual pane. The pane distance information of each virtual pane can include the pane width of the virtual pane (i.e., the side length of the grid), or it can also include the diagonal length of the virtual pane, etc.

[0083] In addition, information such as the pane width and diagonal length of each virtual pane can be obtained when the court is gridded, that is, when the court area is gridded according to the set information such as the pane width and diagonal length of the virtual panes.

[0084] Specifically, after obtaining the virtual panes occupied by the target object in the court area within the preset time period and the pane distance information corresponding to each of these virtual panes, the pane distance information corresponding to each occupied virtual pane can be summed to obtain the total distance the target object moves within the preset time period, which is recorded as the movement distance. This summation can be a direct summation or a weighted summation.

[0085] In this embodiment, at least two frames of images of the court area within a preset time period are acquired, and target detection is performed on the objects in each frame to determine the target detection results for each frame. Then, based on the first positions of the target objects in the court area in each detection result, and considering that the court area is pre-divided into at least two virtual panes within the preset time period, the virtual pane occupied by the target object in the court area within the preset time period is determined. The movement distance of the target object within the preset time period is then determined based on the distance information of the occupied pane. This method automatically acquires and analyzes images of the court area and obtains the movement distance through the distance information of the pane occupied by the target object, thus avoiding manual statistics and recording of movement distances, saving labor and time costs. Furthermore, since the object movement analysis process does not require manual statistics and recording of movement distances, the influence of subjective factors on the object movement analysis process and results can be avoided, thereby ensuring the accuracy and objectivity of the object movement analysis process and results.

[0086] The following example illustrates the process of determining the virtual pane occupied by the target object within a preset time period by using the first position of the target object in each frame image.

[0087] In one exemplary embodiment, S104 above may include the following steps:

[0088] Step A1: Determine the target number corresponding to each first position based on the preset correspondence between the position and number of the virtual pane; each first position corresponds to the position of a virtual pane.

[0089] Step A2: Determine the virtual pane corresponding to each target number as the virtual pane occupied by the target object.

[0090] In this step, as mentioned above, the court area can be pre-gridized within a preset time period, dividing the court area into at least two virtual panes, where the position of each virtual pane corresponds to the actual position of the court area.

[0091] See Figure 3 The diagram illustrates the gridding of the sports field area. The left image shows the entire field area gridded, while the right image shows the marked monitoring areas of each image acquisition device within the gridded area. After gridding the field area, the virtual panes within the area can be numbered sequentially from left to right and from top to bottom. See also... Figure 4 The diagram illustrates how to number the virtual panes within the court area. For example, the first row of virtual panes is numbered from left to right, then the second row is numbered sequentially from left to right, and so on, until all virtual panes are numbered. Each virtual pane in the court area will then have a unique number. It's understood that each virtual pane's numbering is different; for instance, the numbering could start from 1 and proceed sequentially. Assuming there are 100 virtual panes, the numbers could be 1, 2, 3...99, 100.

[0092] After numbering each virtual pane in the stadium area as described above, each number can be bound to the corresponding virtual pane position, establishing a correspondence between the virtual pane position and the number. Then, after obtaining the first position of the target object in each frame image, each first position can be mapped to a virtual pane position, and the number at the corresponding virtual pane position can be obtained through this correspondence, all recorded as the target number. The virtual pane corresponding to each target number can then be used as the virtual pane occupied by the target object during a preset time period.

[0093] In this embodiment, the target number corresponding to each first position is determined according to the correspondence between the position and number of the virtual pane, and the virtual pane occupied by the target object is determined accordingly. This can accurately determine the virtual pane occupied by the target object within a preset time period, thereby improving the accuracy of calculating the target object's movement distance through the occupied virtual pane.

[0094] The following examples illustrate the process of determining the movement distance of a target object within a preset time period using distance information of the virtual grid occupied by the target object.

[0095] In one exemplary embodiment, S106 above may include the following steps:

[0096] Step B1: Determine the pane distance information corresponding to the virtual pane based on the historical motion data corresponding to the court area; the aforementioned historical motion data includes the historical step distance of each historical object in each team at different times of each game and under different sports types.

[0097] This involves pre-collecting historical movement data for each historical object (i.e., player) in different teams during different time periods of each game and under different sports types, such as historical stride (i.e., foot span), player number, player's movement status, player's total movement distance, and player's total movement duration. Then, by statistically analyzing the historical stride of each historical object in each team, the size of the virtual pane when dividing the field area into grids can be obtained, i.e., the pane distance information of the virtual pane.

[0098] Based on the aforementioned historical motion data, electronic devices, central control devices, or central control systems can automatically recommend the pane width (i.e., pane size) of the virtual pane to the user, such as the "Panpane Size Suggestions" in Tables 2 and 3 below. Because the geographical locations of historical players (i.e., players) within each team differ, their physical attributes will also vary. These differences will lead to variations in the historical stride distance of historical players within each team, consequently resulting in different pane widths for the virtual pane corresponding to each team. Generally, a more accurate virtual pane width needs to be determined using the historical motion data of each team. Therefore, in this embodiment, the pane width of the corresponding virtual pane can be determined for each different team.

[0099] The historical athletic data of each team collected in advance are shown in Table 2 below:

[0100] Table 2

[0101]

[0102] The historical sports data of each team can also be obtained from the historical sports data of the above-mentioned teams, as shown in Table 3 below:

[0103] Table 3

[0104]

[0105] The following example uses a football team to illustrate how to determine the width of the corresponding virtual pane.

[0106] Taking a grid-like virtual pane as an example, the pane distance information of the virtual pane can include the pane width of the virtual pane. As an optional embodiment, the pane width of each virtual pane can be determined by the following steps:

[0107] For each team, statistical analysis was conducted on the team's historical stride distance at different times and under different sports types in each game to determine the team's average stride distance at different times and under different sports types in each game;

[0108] Within a preset time period for each game, the duration of each team's movement under different sports types is obtained, and the target step distance corresponding to the preset time period is determined based on the proportion of each movement duration within the preset time period and the average step distance of each movement. The preset time period can be any time period among different time periods.

[0109] The pane width corresponding to the preset time period is determined based on the target step distance for the preset time period; among them, different time periods of each game correspond to different pane widths.

[0110] Specifically, based on the historical step distances of each team in Tables 2 and 3 above, a discrete data plot of the historical step distance for each team can be drawn. See [link to relevant documentation]. Figure 5 The diagram shows discrete data plots of a team's historical stride distance under different sports (or different sports states). Three plots represent the discrete data of historical individuals / players within the team under three sports: walking, running, and acceleration. The horizontal axis represents stride distance, and the vertical axis represents a time period. The green boxes in each plot indicate the data focus points—locations with a high concentration of data points. By analyzing the discrete data plots of each team's historical stride distance under different sports, the data focus points for each sports type can be identified, and the point with the highest density among these focus points can be used as the team's average stride distance for that corresponding sports type. For example... Figure 5 The average stride lengths corresponding to the three figures are as follows: 0.4 meters for slow walking, 1.15 meters for running, and 1.35 meters for sprinting.

[0111] Since a standard team match lasts 90 minutes with 30 minutes of overtime, players' athletic performance and physical condition change as the match progresses. Therefore, it's possible to statistically analyze the changing trends of average stride length for the same sport at different time points. Continuing with the above... Figure 5 Taking three different types of exercise as examples, the trend graphs of the average stride distance changing over time under the three different types of exercise can be found in [reference needed]. Figure 6 As shown in the figure. The horizontal axis of the three graphs can represent time periods, and the vertical axis can represent step size / average step size.

[0112] Based on the above Figure 5 and Figure 6 Analysis shows that the average stride distance of each team varies at different times of the game, so it can be based on... Figure 5 and Figure 6The analysis divides the regular match time into several different time slots, such as three slots: 0-45 minutes, 45-75 minutes, and 75-90 minutes. Then, the historical stride distances of each team during these three different time slots and across various sports types can be statistically analyzed to obtain data dispersion plots for different time slots. For details, please refer to [link / reference needed]. Figure 7 The charts shown depict historical step distance data at different time points. Specifically, the first row contains three charts showing historical step distance data for various exercise types within 0-45 minutes; the second row contains three charts showing historical step distance data for various exercise types within 45-75 minutes; and the third row contains three charts showing historical step distance data for various exercise types within 75-90 minutes. The meanings of the horizontal and vertical axes in these charts are the same as described above. Figure 5 The horizontal and vertical axes in the diagram have the same meaning.

[0113] The average stride distance for each team at different times and under different sports types can then be obtained by analyzing the above discrete graph. For example, taking Manchester United as an example, the average stride distance at different times and under different sports types can be seen in Table 4 below:

[0114] Table 4

[0115]

[0116] Next, the duration of each player's exercise in different sports types can be statistically analyzed at different times, such as walking for X minutes, running for Y minutes, and sprinting for Z minutes. Then, the percentage of exercise time in these three different sports types in the total duration of the time period can be calculated. For example, in the 0-45 minute time period, the percentages of the three different types can be X / 45, Y / 45, and Z / 45. Then, the largest percentage can be selected from these three percentages, and the average stride length corresponding to the largest percentage can be used as the target stride length for the team in that time period.

[0117] After determining the target step distance for each team at different times, this target step distance can be directly used as the pane width for grid division within that time period. It's understandable that since each team has a different target step distance at different times, the pane width will also differ when dividing the field area into grids. In other words, the size of the virtual panes used for grid division will be adjusted or switched according to the time period. For example, the pane width (or grid spacing) is 0.3 during the 0-45 minute period, 0.25 during the 45-75 minute period, and 0.22 during the 75-90 minute period.

[0118] By statistically analyzing the historical stride distances of different teams at different times and under different sports types, we can obtain the corresponding window widths for different time periods. This allows us to adjust the window width of the court area according to the time of the game, making it more consistent with the actual movement of objects in the game, thereby improving the accuracy of the movement distance of the determined target objects.

[0119] Step B2: Count the number of different target numbers among each target number to obtain the number of virtual panes occupied by the target object within the preset time period.

[0120] In this step, the above embodiment illustrates that the virtual panes occupied by the target object in the court area can be determined by the target number of the virtual panes. Furthermore, since the virtual panes are numbered sequentially, the target number changes each time the target object moves a virtual pane. That is, after determining the target numbers of each virtual pane occupied by the target object, duplicate or identical target numbers can be removed, leaving only one. Thus, all remaining target numbers are different. The number of these different target numbers can then be counted, and the total number is the total number of virtual panes occupied by the target object, denoted as the number of occupied virtual panes.

[0121] For example, see Figure 8 The diagram shows the virtual panes occupied by the target object over a period of time. It can be seen that the virtual panes occupied by the target object during this period of time are the green panes in the diagram, and the number of virtual panes occupied is 5.

[0122] Step B3: Determine the movement distance of the target object within a preset time period based on the number of panes and the distance between panes of the target object.

[0123] In this step, after obtaining the number of virtual panes occupied by the target object within a preset time period and the pane distance information of each virtual pane (specifically, the pane width), the two can be multiplied to obtain the movement distance of the target object within the preset time period.

[0124] For example, taking pane distance information including pane width as an example, assuming the movement distance of the target object is S, the distance information of each unit / virtual pane within the preset time period is denoted as F, and the number of virtual panes occupied by the target object within the preset time period is n, then S = n * F.

[0125] In this embodiment, the movement distance of the target object is determined by the number of virtual panes occupied by the target object within a preset time period and the distance information of a single pane. In this way, the movement distance of the target object can be calculated accurately and quickly by using the number of panes and the distance information between panes, thereby improving the accuracy and efficiency of the movement distance calculation.

[0126] In the actual analysis of images of the stadium area, in order to avoid the problem of excessive data volume caused by full image analysis, which would consume resources and affect the real-time performance of the analysis, this embodiment proposes to acquire and analyze images of the stadium area at certain time intervals. The following embodiment describes this process.

[0127] In an exemplary embodiment, obtaining at least two frames of images of the court area within a preset time period in S102 above may include the following steps:

[0128] Step C1: For each team, determine the team's movement frequency within a preset time period based on the historical movement data corresponding to the court area; the aforementioned movement frequency is used to represent the average time during which the team engages in cross-pane movement within the preset time period.

[0129] Step C2: Determine the time interval for motion analysis of objects within the team based on the team's movement frequency.

[0130] Step C3: Acquire at least two frames of images of the court area within a preset time period according to the time interval.

[0131] In this step, for each team, within a preset time period of each game, the distance a player travels per unit time is determined based on the total distance and duration of movement for each player on that team. Then, based on the average stride distance obtained from the statistics for each team, the corresponding pane width is determined. Next, the distance a player travels per unit time is divided by the pane width to obtain the time it takes for a player to cross one virtual pane within the preset time period. Then, the average time taken by all players on each team to cross one virtual pane is calculated to obtain the team's average time to cross one virtual pane within the preset time period. This average time can be used as the team's movement frequency within the preset time period. For example, if a player's total distance traveled is 10 meters and the total duration of movement is 10 seconds within the preset time period of 75-90 minutes, then the player travels 1 meter per second. Assuming the pane width corresponding to this preset time period is 0.25 meters, then the player will cross 4 virtual panes per second (1 meter divided by 0.25 meters = 4), that is, cross one virtual pane every 0.25 seconds (1 second divided by 4 = 0.25).

[0132] After obtaining the movement frequency of each team, the frequency or time interval for analyzing the acquired images can be adjusted according to that team's movement frequency. For example, the team's movement frequency can be directly used as the time interval for image analysis, or the team's movement frequency can be slightly lowered to obtain the time interval for image analysis, or other methods can be used to determine the time interval. It is understandable that within the team's movement frequency, players do not move across panes, so it is not necessary to acquire and analyze images of the field area frame by frame.

[0133] Then, each frame of the stadium area can be acquired according to the time intervals determined above, and each frame of the acquired frame can be analyzed separately.

[0134] For example, assuming a player moves across a window approximately every 2 seconds, the analysis of the field footage does not need to be done frame by frame per second. The analysis interval can be adjusted to 2 seconds or other intervals, thereby reducing the frequency of analysis and the amount of data.

[0135] Furthermore, an image of the stadium area can be obtained at the aforementioned time intervals, for example... Figure 9 The image shows a schematic diagram of the court area and its objects. Motion analysis is then performed on the objects in this frame to obtain the results, which can then be saved or recorded. For example, teams A and B are used to represent the two teams respectively. Players from different teams are labeled; for instance, player number 1 of team A is labeled A1, and player number 2 is labeled A2. This way, each player from both teams will have a unique label.

[0136] To improve image analysis performance, assume the electronic device analyzes and identifies players and their positions within the current frame every 500ms, and records the player's position. Based on this analysis, the virtual pane position of the player at different times (i.e., the corresponding virtual pane number, such as M15) can be determined. The storage method for recording the motion analysis results can be referenced in Table 5 below, which stores the virtual pane number occupied by the player at each time point:

[0137] Table 5

[0138]

[0139] For example, see Figure 10The diagram shown illustrates the acquisition and analysis of images of a sports field area. The diagram represents an 8-second video segment. Traditional full-view analysis requires analyzing the entire 8-second video. However, according to the analysis method in this embodiment, based on the above analysis, assuming that a 3-second time interval can achieve both accurate and efficient analysis, the 8-second video can be analyzed by extracting frames at 3-second intervals. Frame images are acquired and analyzed at analysis points a, b, and c. This reduces the performance consumption of electronic devices to 3 / 8 of the original, significantly reducing the amount of data analyzed and the resource consumption.

[0140] In this embodiment, the time interval for image acquisition and analysis is determined by the team's movement frequency within a preset time period, and images are acquired and analyzed according to the time interval. This eliminates the need for full frame-by-frame analysis of the video of the stadium area, thereby reducing the amount of data and resource consumption for image acquisition and analysis, and improving the efficiency of analyzing stadium area images.

[0141] The above embodiments illustrate the process of acquiring and analyzing images of the court area at certain time intervals. In reality, there may be cases where the target object in two adjacent frames acquired at time intervals may not be moving continuously. In this case, in order to more accurately calculate the movement distance of the target object, a full analysis of all images contained in the two adjacent frames can be performed to obtain the virtual pane occupied by the target object. The following embodiments illustrate this process.

[0142] In an exemplary embodiment, the at least two frames mentioned above include the current frame image and the previous frame image of the current frame image; S104 may include the following steps:

[0143] Step D1: If the second position of the virtual pane corresponding to the first position of the target object in the current frame image is different from the third position of the virtual pane corresponding to the first position of the target object in the previous frame image, then detect whether the virtual panes corresponding to the second position and the third position are consecutive virtual panes.

[0144] In this step, two adjacent frames can be obtained at the aforementioned time intervals, denoted as the current frame and the frame preceding it, respectively. Then, the objects in both the current and previous frames can be analyzed to obtain the first position of the target object in each frame. As mentioned above, the first position obtained from analyzing the frame images can also correspond to the position of the virtual pane after the field area is meshed. The position of the virtual pane corresponding to the first position of the target object in the current frame is denoted as the second position, and the position of the virtual pane corresponding to the first position of the target object in the previous frame is denoted as the third position.

[0145] Then, it can be checked whether the virtual panes corresponding to the second and third positions are consecutive virtual panes. For example, it can be checked whether the numbers of the virtual panes corresponding to the second and third positions are the numbers of adjacent virtual panes. If these two numbers are the numbers of adjacent virtual panes, then the virtual panes corresponding to the second and third positions are consecutive virtual panes; otherwise, they are not consecutive virtual panes.

[0146] For example, see Figure 11 The diagram shows the numbering of the stored virtual panes. Assuming the second and third positions correspond to virtual panes numbered M25 and M26 respectively, then from... Figure 11 As can be seen, virtual panes M25 and M26 are adjacent virtual panes, therefore virtual panes M25 and M26 are consecutive virtual panes. Further assuming that the second and third positions correspond to virtual panes numbered M25 and M10 respectively, then from... Figure 11 It can be seen that virtual panes M25 and M10 are not adjacent virtual panes, therefore virtual panes M25 and M10 are not consecutive virtual panes.

[0147] In step D2, if the virtual panes are not continuous, all frames between the previous frame and the current frame are acquired and analyzed to determine the motion trajectory of the target object.

[0148] In this step, the above steps determine whether the virtual panes corresponding to the target object in the previous and current frames are consecutive. If they are not consecutive, it indicates that the target object has completed movement between at least two virtual panes within the given time interval. To more accurately calculate the target object's movement distance within this time interval, the analysis time of the previous frame can be obtained from the electronic device, and the analysis time of the current frame can be used as the analysis duration. Then, all frame images within this analysis duration can be obtained from the electronic device, i.e., all frame images from the previous and current frames within the given time interval can be obtained and analyzed. When analyzing all frame images, the analysis can be performed frame by frame. After analyzing each frame, the first position corresponding to the target object in the corresponding frame image will be obtained. Arranging these first positions according to the time order of the frames yields the target object's movement trajectory within the given time interval.

[0149] For example, see Figure 12 The second illustration shows the acquisition and analysis of images of the court area. Assuming the acquisition and analysis of images of the court area continues at 3-second intervals, when the analysis reaches analysis point b, it is found that the target object in this image appears in a non-contiguous pane from the pane containing the target object at analysis point a (see [reference]). Figure 11(Examples M10 and M25 in the example) indicate that the target object moves across panes, meaning that the target object experiences significant speed and positional movement within the corresponding time interval. In this case, a detailed analysis of the target object's trajectory is required. Assuming analysis point a occurs at time Ta and analysis point b occurs at time Tb, all frames within the Ta-Tb time interval can be acquired and analyzed frame by frame in the electronic device. For example, if there are three frames within the Ta-Tb time interval, there are four analysis points: a1, a2, a3, and a4. Image analysis can be performed at each of these four analysis points to obtain the specific trajectory of the target object within that time interval.

[0150] Step D3: Based on the movement trajectory of the target object, determine the virtual panes occupied by the target object in the court area within the preset time period.

[0151] In this step, after obtaining all the frames in the previous frame image and the current frame image and performing frame-by-frame analysis, the first position corresponding to the target object in each frame image can be obtained. Then, each first position can be mapped to the position of the virtual pane, and the virtual pane occupied by the target object can be obtained through the number on each virtual pane.

[0152] Understandably, after performing a full frame-by-frame analysis of all frames between the current frame and the previous frame, the process will continue to extract / acquire frames and perform analysis at time intervals until the target object reappears in the cross-pane movement within the time interval. Then, the process of performing a full frame-by-frame analysis of all frames between the current frame and the previous frame will be repeated until the game ends.

[0153] In this embodiment, during the process of acquiring and analyzing images at time intervals, if the target object moves across panes in two adjacent frames, all frames within that time interval can be acquired and analyzed frame by frame to determine the specific virtual pane occupied by the target object. This makes the determination of the virtual pane occupied by the target object more accurate, and the subsequent calculation of the target object's movement distance based on the occupied virtual pane will be more accurate.

[0154] When multiple image acquisition devices are set up in the court area, there may be overlapping areas in the images acquired by the multiple image acquisition devices. This will result in the final determined image of the court area being inaccurate. Based on this, an embodiment that can remove overlapping areas is proposed to solve this problem. The following embodiment will explain the process of removing overlapping areas.

[0155] In one exemplary embodiment, the above method may further include the following steps:

[0156] Step E1: Based on the regional images captured by each image acquisition device within the court area, determine at least two candidate regions constituting the court area, as well as the length and width of each candidate region.

[0157] Step E2: Remove overlapping areas from each candidate area, and determine the actual length and width of the court area based on the length and width of each candidate area after removing overlapping areas.

[0158] Step E3: Divide the court area into at least two virtual panes within a preset time period according to the actual length and width of the court area.

[0159] In this step, the image acquisition device can be a camera, video camera, etc. Taking a camera as an example, suppose the stadium has six cameras pre-set, namely P1, P2, P3, P4, P5, and P6. These six cameras can respectively acquire images of different areas of the stadium, obtaining their respective acquired images. For example, see [link to relevant documentation]. Figure 13 The diagram illustrates images captured by each image acquisition device of the court area. The images captured by each camera can then be analyzed to obtain the length and width of the court area captured by each camera, for example, see [link to example]. Figure 14 The diagram shows the length and width of each court area. The length and width of the court area captured by camera P1 are L1 and L12, respectively; the length and width of the court area captured by camera P2 are (L2+L3) and L12, respectively; the length and width of the court area captured by camera P3 are L4 and L5, respectively; the length and width of the court area captured by camera P4 are L7 and L6, respectively; the length and width of the court area captured by camera P5 are (L8+L9) and L6, respectively; and the length and width of the court area captured by camera P6 are L10 and L11, respectively.

[0160] The length of the court area captured by the P2 and P5 cameras is divided into two parts. This is because the P2 and P5 cameras capture images at the center of the court, and the images captured are of the areas on both sides of the center, hence the length is divided into two parts.

[0161] The above analysis yields the length and width of each court area. It can be seen that the length or width calculations of L1 and L2, L3 and L4, L12 and L6, L5 and L11, L7 and L8, and L9 and L10 overlap due to the overlapping of the camera's visualization area. Therefore, it is necessary to perform deduplication processing on the overlapping length or width.

[0162] When performing deduplication, see [link / reference]. Figure 15The diagram illustrating the overlapping area shows how to identify the overlapping areas of the court areas captured by two adjacent cameras. This involves identifying the overlapping areas within the court areas captured by the two adjacent cameras, obtaining the length and width of the overlapping area, and then subtracting the length and width of the overlapping area from either of the two captured court areas to obtain the new length and width of the new court area. This method can be used to identify and remove duplicates from the overlapping areas of two court areas that both have overlapping regions, obtaining the new length and width of the court areas captured by each camera.

[0163] Then, the new lengths and widths of each court area can be stitched together to obtain the actual length and width of the entire court area. After obtaining the actual length and width of the court area, the pane width corresponding to the preset time period can be obtained, and within the preset time period of the game, the entire court area can be divided into at least two virtual panes according to the actual length and width of the court area.

[0164] In this embodiment, by removing overlapping areas in the court area captured by each image acquisition device, the accuracy of the actual length and width of the entire court area can be improved, achieving full coverage of the entire court area without duplication. Therefore, when the entire court area is gridded, duplicate panes can be avoided, which would cause duplicate motion distances to be calculated, further improving the accuracy of the calculated motion distance of the target object.

[0165] As described in the above embodiments, the object motion analysis method of the present invention can track and record the movement trajectory and running distance of a target object in a match in real time and accurately. This not only greatly improves the efficiency and accuracy of target object motion analysis but also provides more scientific and objective data support for football training and matches.

[0166] The object motion analysis device provided in the embodiments of the present invention will be described below. The object motion analysis device described below can be referred to in correspondence with the object motion analysis method described above.

[0167] Figure 16 This is a schematic diagram of the object motion analysis device provided in an embodiment of the present invention. See also... Figure 16 As shown, the device may include:

[0168] The analysis module 210 is used to acquire at least two frames of images of the court area within a preset time period, and to perform target detection on the objects in each frame of images to determine the target detection result corresponding to each frame of images; each target detection result includes the first position of the target object in the court area;

[0169] The occupancy pane determination module 220 is used to determine the virtual panes occupied by the target object in the court area within a preset time period based on the first positions of the target object in the court area; wherein the court area is pre-divided into at least two virtual panes within the preset time period;

[0170] The motion distance determination module 230 is used to determine the motion distance of the target object within the preset time period based on the pane distance information of the virtual pane occupied by the target object.

[0171] In an exemplary embodiment, the aforementioned occupancy pane determination module 220 is specifically configured to determine the target number corresponding to each first position based on the preset correspondence between the position and number of the virtual pane; each first position corresponds to the position of a virtual pane; and determine the virtual pane corresponding to each target number as the virtual pane occupied by the target object.

[0172] In one exemplary embodiment, the motion distance determination module 230 described above includes:

[0173] The pane distance information determination unit is used to determine the pane distance information corresponding to the virtual pane based on the historical motion data corresponding to the field area; the aforementioned historical motion data includes the historical step distance of each historical object in each team at different times of each game and under different sports types.

[0174] The pane number determination unit is used to count the number of different target numbers in each target number and obtain the number of virtual panes occupied by the target object within a preset time period;

[0175] The motion distance determination unit is used to determine the motion distance of the target object within a preset time period based on the number of panes and the pane distance information of the target object.

[0176] In an exemplary embodiment, the aforementioned pane distance information includes pane width. The pane distance information determining unit is specifically configured to, for each team, statistically analyze the team's historical stride distance at different times and under different sports types in each game, and determine the team's average stride distance at different times and under different sports types in each game; within a preset time period of each game, obtain the team's sports duration under different sports types, and determine the target stride distance corresponding to the preset time period based on the proportion of each sports duration within the preset time period and the average stride distance; the aforementioned preset time period can be any time period among different time periods; and determine the pane width corresponding to the preset time period based on the target stride distance corresponding to the preset time period; wherein, different time periods of each game correspond to different pane widths.

[0177] In an exemplary embodiment, the analysis module 210 is specifically configured to determine the movement frequency of each team within a preset time period based on historical movement data corresponding to the court area; the movement frequency is used to represent the average time during which the team exhibits cross-pane movement within the preset time period; determine the time interval for motion analysis of objects within the team based on the team's movement frequency; and acquire at least two frames of images of the court area within the preset time period according to the time interval.

[0178] In an exemplary embodiment, the aforementioned at least two frames include the current frame image and the previous frame image; the aforementioned occupancy pane determination module 220 is further configured to: if the second position of the virtual pane corresponding to the first position of the target object in the current frame image is different from the third position of the virtual pane corresponding to the first position of the target object in the previous frame image, then detect whether the virtual panes corresponding to the second position and the third position are consecutive virtual panes; if they are not consecutive virtual panes, then acquire all frame images between the previous frame image and the current frame image, and analyze all frame images to determine the motion trajectory of the target object; and determine the virtual panes occupied by the target object in the court area within a preset time period based on the motion trajectory of the target object.

[0179] In one exemplary embodiment, the above-described apparatus further includes:

[0180] The first length and width determination module is used to determine at least two candidate regions constituting the court area, as well as the length and width of each candidate region, based on the regional images captured by each image acquisition device within the court area.

[0181] The second length and width determination module is used to remove overlapping areas in each candidate area, and after removing overlapping areas, determine the actual length and actual width of the court area based on the length and width of each candidate area.

[0182] The division module is used to divide the court area into at least two virtual panes within a preset time period according to the actual length and width of the court area.

[0183] Figure 17 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 17As shown, the electronic device may include a processor 310, a communication interface 320, a memory 330, and a communication bus 340, wherein the processor 310, the communication interface 320, and the memory 330 communicate with each other via the communication bus 340. The processor 310 can call logical instructions in the memory 330 to execute an object motion analysis method, which includes: acquiring at least two frames of images of a court area within a preset time period, performing target detection on objects in each frame of images, and determining the target detection result corresponding to each frame of images; each target detection result includes a first position of the target object in the court area; determining the virtual panes occupied by the target object in the court area within the preset time period based on the first positions of the target object in the court area; wherein the court area is pre-divided into at least two virtual panes within the preset time period; and determining the movement distance of the target object within the preset time period based on the pane distance information of the virtual panes occupied by the target object.

[0184] Furthermore, the logical instructions in the aforementioned memory 330 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention embodiment, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0185] On the other hand, embodiments of the present invention also provide an object motion analysis system, including: an image acquisition device and an electronic device interconnected, wherein the image acquisition device is used to acquire images of a court area and transmit them to the electronic device; the electronic device is used to execute the object motion analysis method provided by the above methods based on the images transmitted by the image acquisition device, the method including: acquiring at least two frames of images of the court area within a preset time period, and performing target detection on objects in each frame of images to determine the target detection result corresponding to each frame of images; each target detection result includes a first position of the target object in the court area; determining the virtual panes occupied by the target object in the court area within the preset time period based on each first position of the target object in the court area; wherein the court area is pre-divided into at least two virtual panes within the preset time period; and determining the movement distance of the target object within the preset time period based on the pane distance information of the virtual panes occupied by the target object.

[0186] On the other hand, embodiments of the present invention also provide a computer program product, the computer program product including a computer program that can be stored on a computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the object motion analysis method provided by the above methods. The method includes: acquiring at least two frames of images of a court area within a preset time period, and performing target detection on objects in each frame of images to determine the target detection result corresponding to each frame of images; each target detection result includes a first position of the target object in the court area; determining the virtual panes occupied by the target object in the court area within the preset time period based on the first positions of the target object in the court area; wherein the court area is pre-divided into at least two virtual panes within the preset time period; and determining the movement distance of the target object within the preset time period based on the pane distance information of the virtual panes occupied by the target object.

[0187] In another aspect, embodiments of the present invention also provide a computer-readable storage medium storing a computer program thereon. When executed by a processor, the computer program implements the object motion analysis method provided by the above methods. The method includes: acquiring at least two frames of images of a court area within a preset time period, performing target detection on objects in each frame of images, and determining the target detection result corresponding to each frame of images; each target detection result includes a first position of the target object in the court area; determining the virtual panes occupied by the target object in the court area within the preset time period based on the first positions of the target object in the court area; wherein the court area is pre-divided into at least two virtual panes within the preset time period; and determining the movement distance of the target object within the preset time period based on the pane distance information of the virtual panes occupied by the target object.

[0188] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0189] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0190] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for analyzing the motion of an object, characterized in that, include: Acquire at least two frames of images of the court area within a preset time period, and perform target detection on objects in each frame of images to determine the target detection result corresponding to each frame of images; each target detection result includes the first position of the target object in the court area; Based on the first positions of the target object in the court area, determine the virtual panes occupied by the target object in the court area during the preset time period; wherein, the court area is pre-divided into at least two virtual panes during the preset time period; each first position corresponds to the position of a virtual pane, and each position of a virtual pane corresponds to a target number; Based on the pane distance information of the virtual pane occupied by the target object, the movement distance of the target object within the preset time period is determined; Determining the movement distance of the target object within the preset time period based on the pane distance information of the virtual pane occupied by the target object includes: Based on the historical motion data corresponding to the court area, the pane distance information corresponding to the virtual pane is determined; the historical motion data includes the historical step distance of each historical object in each team at different times of each game and under different sports types; The number of different target numbers among all target numbers is counted to obtain the number of virtual panes occupied by the target object within the preset time period; Based on the number of panes and the distance between panes of the target object, the movement distance of the target object within the preset time period is determined.

2. The object motion analysis method according to claim 1, characterized in that, The step of determining the virtual pane occupied by the target object in the court area within the preset time period based on each of the first positions of the target object in the court area includes: Based on the pre-defined correspondence between the positions and numbers of the virtual panes, the target number corresponding to each of the first positions is determined; The virtual pane corresponding to each of the target numbers is determined as the virtual pane occupied by the target object.

3. The object motion analysis method according to claim 1, characterized in that, The pane distance information includes the pane width. Determining the pane distance information of the virtual pane based on the historical motion data corresponding to the court area includes: For each team, statistical analysis is performed on the team's historical stride distance at different times and under different sports types in each game to determine the team's average stride distance at different times and under different sports types in each game; Within a preset time period of each game, the duration of the team's movements under different sports types is obtained, and the target step distance corresponding to the preset time period is determined based on the proportion of each movement duration within the preset time period and the average step distance of each movement; the preset time period can be any of the different time periods. The pane width corresponding to the preset time period is determined based on the target step distance corresponding to the preset time period; wherein, different time periods of each game correspond to different pane widths.

4. The object motion analysis method according to any one of claims 1 to 3, characterized in that, The acquisition of at least two frames of images of the court area within a preset time period includes: For each team, the team's movement frequency within a preset time period is determined based on historical movement data corresponding to the court area; the movement frequency is used to represent the average time during which the team engages in cross-pane movement within the preset time period. The time interval for motion analysis of objects within the team is determined based on the team's movement frequency; At least two frames of images of the court area within a preset time period are acquired according to the time interval.

5. The object motion analysis method according to any one of claims 1 to 3, characterized in that, The at least two frames include the current frame and the frame preceding the current frame; determining the virtual pane occupied by the target object in the court area within the preset time period based on the first position of the target object in the court area includes: If the second position of the virtual pane corresponding to the first position of the target object in the current frame image is different from the third position of the virtual pane corresponding to the first position of the target object in the previous frame image, then it is detected whether the virtual panes corresponding to the second position and the third position are consecutive virtual panes. If the virtual panes are not consecutive, all frames between the previous frame and the current frame are acquired, and all frames are analyzed to determine the motion trajectory of the target object. Based on the movement trajectory of the target object, determine the virtual pane occupied by the target object in the court area during the preset time period.

6. The object motion analysis method according to any one of claims 1 to 3, characterized in that, The method further includes: Based on the regional images captured by each image acquisition device within the court area, at least two candidate regions constituting the court area are determined, along with the length and width of each candidate region. Overlapping areas in each of the candidate areas are removed, and the actual length and width of the court area are determined based on the length and width of each candidate area after removing the overlapping areas. Based on the actual length and width of the court area, the court area is divided into at least two virtual panes within the preset time period.

7. An object motion analysis device, characterized in that, include: The analysis module is used to acquire at least two frames of images of the court area within a preset time period, and to perform target detection on objects in each frame of images to determine the target detection result corresponding to each frame of images; each target detection result includes the first position of the target object in the court area; The occupancy pane determination module is used to determine, based on the first positions of the target object in the court area, the virtual panes occupied by the target object in the court area during the preset time period; wherein, the court area is pre-divided into at least two virtual panes during the preset time period; each first position corresponds to the position of a virtual pane, and each position of the virtual pane corresponds to a target number; The motion distance determination module is used to determine the motion distance of the target object within the preset time period based on the pane distance information of the virtual pane occupied by the target object; The motion distance determination module includes: A pane distance information determination unit is used to determine the pane distance information corresponding to the virtual pane based on the historical motion data corresponding to the court area; the historical motion data includes the historical step distance of each historical object in each team at different times of each game and under different sports types; A pane number determination unit is used to count the number of different target numbers among all target numbers to obtain the number of virtual panes occupied by the target object within the preset time period; The movement distance determination unit is used to determine the movement distance of the target object within the preset time period based on the number of panes and the pane distance information of the target object.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the object motion analysis method as described in any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the object motion analysis method as described in any one of claims 1 to 6.

Images