Archery event capture visualization system and method

Abstract

The invention provides an archery event capture visualization system and method. The method comprises the steps that an image collection device collects image data in the archery event process; the laser radar device collects archery point cloud data on an archery track in the archery competition process; the intelligent identification module carries out arrow identification processing on the image data and the archery point cloud data to obtain arrow three-dimensional trajectory data; the storage module stores arrow three-dimensional trajectory data; and the three-dimensional engine module calls the arrow three-dimensional trajectory data to establish a three-dimensional archery event visual model. According to the obtained accurate arrow movement track data, more comprehensive and more accurate match analysis is performed, the process and the result of the match are analyzed, establishment of an arrow movement visual model is provided, match rebroadcasting has more display schemes, and analysis data can provide more powerful support for training of athletes.

Description

Technical Field

[0001] The present invention relates to the technical field of artificial intelligence recognition and processing, and particularly to an archery event capture visualization system and method. Background Art

[0002] The patent application 202311415349.7 provides a device for recognizing archery and its running trajectory based on dual cameras. It includes a structural frame, a projection screen, and customized floor glue. The customized floor glue is laid inside the structural frame. One end of the customized floor glue is vertically provided with a projection screen. Infrared lasers are arranged at the top and bottom of the projection screen. A wall-mounted camera recognition device is installed on one side of the archery logo. An archery interaction recognition device is installed on the top of the structural frame. Speakers and iron plates are arranged on the side frame of the structural frame. The iron plate is connected to the structural frame through a chuck. A touch screen and an ultra-thin light box description board are installed on the iron plate. An archery prop is arranged on the structural frame on the opposite side of the iron plate. The touch screen, infrared laser, wall-mounted camera recognition device, and archery interaction recognition device are respectively connected to an external computer through wires. This device for recognizing archery and its running trajectory based on dual cameras displays a nearly real archery movement interaction effect through an effect software of the simulated archery movement system.

[0003] The above technology only supports indoor archery simulation scenarios and is not actually used in outdoor venues. Different from indoor venues, when actually used in outdoor events, it may be affected by factors such as audiences, flying birds, and the environment. Therefore, the detection of arrows in outdoor archery projects needs to overcome challenges brought by large scenes, light changes, various weather conditions, etc. Relying solely on camera vision imaging will pose a greater risk of interference, thus affecting the detection recall rate and accuracy. And the existing technology that only uses camera recognition is easily affected by light and weather factors, resulting in a decrease in recognition accuracy. Therefore, the existing solutions have limited application scenarios and cannot support outdoor events.

[0004] In addition, for archery competitions in the existing technology, only the shooting degree of archery is analyzed based on the camera. Facing archery events, multi-angle analysis cannot be carried out, and visualization is not performed either, resulting in poor viewing effects for audiences during event broadcasts.

[0005] From the perspective of application scenarios, currently, the combined solutions of vision and radar are mostly applied in fields such as autonomous driving and drone navigation, and few technologies are applied to the scenarios of sports events. And archery sports events are very different from scenarios such as autonomous driving. The main difficulty in the archery event solution lies in that the arrow is a small target object moving at high speed. The diameter of the arrow is about 9 mm and the speed is about 67 m / s. Applying the combination of vision and radar to archery events becomes a difficult point. Summary of the Invention

[0006] Based on the above problems, the present invention provides an archery event capture visualization system and method, which aims to solve the technical problems such as the inability of the prior art to meet the detection accuracy of outdoor archery.

[0007] The present invention provides an archery event capture visualization system, including:

[0008] Image acquisition devices, which are distributed in the areas near the head and the tail of the shooting lanes on the competition field, and are used for acquiring image data during the archery competition;

[0009] LiDAR devices, which are distributed on the outer sides of the shooting lanes, and are used for acquiring the archery point cloud data on the shooting lanes during the archery competition;

[0010] An intelligent recognition module, which is respectively connected to the image acquisition devices and the LiDAR devices, and is used for performing arrow recognition processing on the image data and the archery point cloud data to obtain arrow three-dimensional trajectory data;

[0011] A storage module, which is connected to the intelligent recognition module, and is used for storing the arrow three-dimensional trajectory data;

[0012] A three-dimensional engine module, which is connected to the storage module, and is used for calling the arrow three-dimensional trajectory data to establish a three-dimensional visualization model of the archery event.

[0013] Furthermore, it further includes:

[0014] A video conversion module, which is connected to the three-dimensional engine module, and is used for converting the visualization model of the archery event into a video signal to be sent to the broadcaster to present the visualization model of the archery event in the event broadcast screen.

[0015] Furthermore, the image acquisition device in the area near the head of the shooting lane is used for acquiring the first image data of the starting point area, and the image acquisition device in the area near the tail of the shooting lane is used for acquiring the second image data of the target area; the intelligent recognition module includes:

[0016] A first image detection unit, which is used for identifying the two-dimensional key points of the athlete and the arrow in the first image data based on the first detection model, and calculating the three-dimensional key point coordinates of the athlete and the arrow according to the identified two-dimensional key points of the athlete and the arrow to obtain the first three-dimensional coordinate data;

[0017] A second image detection unit, which is used for identifying the two-dimensional key points of the arrow in the second image data based on the second detection model, and calculating the three-dimensional key point coordinates of the arrow by combining the identified two-dimensional key points of the arrow and the three-dimensional coordinates calibrated by the position of the target to obtain the second three-dimensional coordinate data, and the second three-dimensional coordinate data includes the target hitting position coordinates of the arrow;

[0018] A point cloud detection unit, which is used for identifying the arrow point cloud data in the archery point cloud data based on the third detection model;

[0019] The comprehensive detection unit is respectively connected to the first image detection unit, the second image detection unit and the point cloud detection unit, and is used to fuse the first three-dimensional coordinate data, the second three-dimensional coordinate data and the arrow point cloud data to obtain the arrow three-dimensional trajectory data.

[0020] Furthermore, the three-dimensional engine module is used for: calling the arrow three-dimensional trajectory data of a single archery shot for data analysis to obtain the first analysis result of the single archery shot, and constructing an archery event visualization model based on the first analysis result and the called arrow three-dimensional trajectory data.

[0021] Furthermore, the three-dimensional engine module is used for: calling the arrow three-dimensional trajectory data of at least two consecutive archery shots of the same athlete for data comparison and analysis to obtain the second analysis result of the multiple archery shots of a single athlete, and constructing an archery event visualization model based on the second analysis result and the called arrow three-dimensional trajectory data.

[0022] Furthermore, the three-dimensional engine module is used for: calling the arrow three-dimensional trajectory data of each winning arrow for data comparison and analysis to obtain the third analysis result regarding the winning arrow, and constructing an archery event visualization model based on the third analysis result and the called arrow three-dimensional trajectory data of each winning arrow.

[0023] Furthermore, the image acquisition devices in the area near the head of the arrow lane are four industrial cameras, with two industrial cameras distributed on each side of the head of the arrow lane;

[0024] The image acquisition devices in the area near the tail of the arrow lane are four industrial cameras; two industrial cameras are distributed on each side of the tail of the arrow lane;

[0025] The lidar device includes at least two lidars, and the lidars are all on the same outer side of the arrow lane.

[0026] Furthermore, the intelligent recognition module is used for performing arrow recognition processing on the image data and the archery point cloud data after time synchronization to obtain the arrow three-dimensional trajectory data.

[0027] An archery event capture and visualization method, using the aforementioned archery event capture and visualization system, includes:

[0028] Step A1, the image acquisition device acquires the image data during the archery competition, and the lidar device acquires the archery point cloud data on the arrow lane during the archery competition;

[0029] Step A2, the intelligent recognition module performs arrow recognition processing on the image data and the archery point cloud data to obtain the arrow three-dimensional trajectory data;

[0030] Step A3, the storage module stores the arrow three-dimensional trajectory data;

[0031] Step A4: The 3D engine module calls the 3D trajectory data of the arrow from the storage module to establish a 3D visualization model of the archery event.

[0032] Further, after step A4, the following steps are also included:

[0033] Step A5: The video conversion module converts the visualization model of the archery event into a video signal and sends it to the broadcaster to present the visualization model of the archery event in the event broadcast screen.

[0034] The beneficial technical effects of the present invention are as follows: By combining the image data collected by the camera and the point cloud data collected by the lidar for fusion analysis to obtain the movement trajectory of the arrow, the lidar has a natural advantage in anti-interference in the outdoor environment, which can largely make up for the weaknesses of the pure camera solution. The method of combining vision and lidar calculation can effectively improve the accuracy of arrow position detection. Moreover, based on the obtained accurate arrow movement trajectory data, more comprehensive and accurate event analysis can be carried out, analyzing the process and results of the competition, providing the establishment of a visualization model of the arrow movement, enabling the event broadcast to have more display options, further improving the professionalism and viewing pleasure of the archery event broadcast, and the analysis data can provide more powerful support for the training of athletes. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1-2 It is a module schematic diagram of a visualization system for archery event capture according to the present invention;

[0036] Figure 3 It is a step flow chart of a method for visualizing archery event capture according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0037] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0038] It should be noted that, without conflict, the embodiments in the present invention and the features in the embodiments can be combined with each other.

[0039] Next, the present invention will be further described in conjunction with the accompanying drawings and specific embodiments, but it is not intended to limit the present invention.

[0040] See Figure 1 , the present invention provides a visualization system for archery event capture, including:

[0041] The image acquisition device (1) is distributed in the area near the head and the area near the tail of the shooting lane on the competition field, and is used to acquire image data during the archery competition;

[0042] The lidar device (2) is distributed on the outer side of the shooting lane and is used to acquire the archery point cloud data on the shooting lane during the archery competition;

[0043] The intelligent recognition module (3) is respectively connected to the image acquisition device (1) and the lidar device (2), and is used to perform arrow recognition processing on the image data and the archery point cloud data to obtain the three-dimensional trajectory data of the arrow;

[0044] The storage module (4) is connected to the intelligent recognition module (3) and is used to store the three-dimensional trajectory data of the arrow;

[0045] The 3D engine module (5) is connected to the storage module (4) and is used to call the three-dimensional trajectory data of the arrow to establish a three-dimensional visual model of the archery event.

[0046] Further, the image acquisition device (1) in the area near the head of the shooting lane is four industrial cameras, and two industrial cameras are distributed on each side of the head of the shooting lane;

[0047] The image acquisition device (1) in the area near the tail of the shooting lane is four industrial cameras; two industrial cameras are distributed on each side of the tail of the shooting lane;

[0048] The lidar device (2) includes at least two lidars, and the lidars are all on the same outer side of the shooting lane.

[0049] The acquisition ranges of the image acquisition device (1) and the lidar device (2) completely cover the archery competition area (for example, an area of 70m×20m), and can adapt to complex scenarios including indoor, outdoor, and various weather conditions. The 3D coordinate error of the arrow trajectory reaches the centimeter level, and the system delay ≤500ms.

[0050] Each industrial camera of the image acquisition device (1) has a number, and each lidar of the lidar device (2) has a number. The cameras in the area near the head of the shooting lane are used to observe the starting point of shooting, and the cameras in the area near the tail of the shooting lane are used to observe the arrow trajectory and the target hitting data.

[0051] Specifically, the camera specification parameters of the image acquisition device (1) of the present invention are as follows:

[0052] 8 surrounding cameras;

[0053] Model: JAI-GOX-8105C-5GE;

[0054] Pixels: 8.1 million;

[0055] Resolution: 2856×2848px;

[0056] Frame rate: 66fps;

[0057] Interface: 5Gbps GigE Vision;

[0058] Image sensor: 1XCMOS;

[0059] Image sensor name: IMX546 Pregius S;

[0060] Image sensor size: 2 / 3inch;

[0061] Pixel size: 2.74×2.74μm;

[0062] Color: Color;

[0063] Shutter mode: Global shutter.

[0064] Specifically, the specification parameters of the lidar device are as follows:

[0065] 2 lidars;

[0066] Model: LS-ST-ARC600;

[0067] Band: 1550nm eye-safe;

[0068] Detection range: 2m~60m(10%), maximum 150m;

[0069] Bus speed: 500 lines / second;

[0070] Frame rate: 20Hz;

[0071] Ranging accuracy: ±2cm;

[0072] IP rating: IP6K9K;

[0073] Operating temperature: -40℃~85℃;

[0074] Power consumption: ≤40W.

[0075] The present invention obtains the movement trajectory of the arrow through the fusion analysis of the image data collected by the camera and the point cloud data collected by the lidar. The lidar has a natural advantage in anti-interference in the outdoor environment and can largely make up for the weaknesses of the pure camera solution. The combination of vision and lidar can effectively improve the accuracy of arrow position detection. Moreover, more comprehensive and accurate event analysis can be carried out based on the obtained accurate arrow movement trajectory data, analyzing the process and results of the competition, providing the establishment of a visual model of arrow movement, enabling the sports event broadcast to have more display options, further improving the professionalism and viewing experience of the archery event broadcast, and the analyzed data can provide more powerful support for the training of athletes.

[0076] The present invention includes a preliminary preparation stage, an algorithm processing stage, and a data application stage.

[0077] First, in the preliminary preparation stage. Several industrial cameras are arranged in the areas near the head and tail of the shooting range, and at least two lidars are arranged on the same outer side of the shooting range. The industrial cameras are used to collect image data, and the lidars are used to collect point cloud data.

[0078] In the preparation stage, internal parameter calibration and external parameter calibration are also carried out on each industrial camera. The internal parameters refer to the internal parameter matrix and distortion parameters of the industrial camera, etc., to ensure the accuracy of subsequent image reconstruction and measurement results. The external parameters refer to the 3D coordinate parameters of each camera relative to the real world, so as to restore the three-dimensional movement trajectory data of the arrow in subsequent video processing.

[0079] In the preparation stage, for the lidar, through the translation and rotation of the matrix, the lidar XYZ coordinate system is transformed into the world coordinate system. Manually select the calibration points to obtain multiple calibration points with known world coordinates in the point cloud, and calculate the transformation matrix. Obtain the actual layout of the competition venue and the point cloud imaging, cover the track, configure the point cloud imaging limit to exclude interferences such as buildings, walls, and spectator stands, so as to only perform point cloud imaging on the track area in the actual archery competition, and not perform point cloud imaging on the interference areas such as buildings, walls, and spectator stands, reducing the computational amount of point cloud imaging.

[0080] In the preparation stage, the image acquisition device (1) and the lidar device (2) are also connected to the intelligent recognition module (3) and business rules such as video stream parameters are configured, so that the intelligent recognition module (3) can obtain image data and archery point cloud data in real time.

[0081] Specifically, the intelligent recognition module (3) is an AI analysis server.

[0082] See Figure 2, Further, the intelligent recognition module (3) is used to perform arrow recognition processing on the image data and the archery point cloud data after time synchronization to obtain the three-dimensional trajectory data of the arrow. Further, the image acquisition device (1) in the area near the head of the arrow lane is used to acquire the first image data of the starting point area, and the image acquisition device (1) in the area near the tail of the arrow lane is used to acquire the second image data of the target surface area;

[0083] The intelligent recognition module (3) includes:

[0084] The first image detection unit (31) is used to identify the two-dimensional key points of the athlete and the arrow in the first image data based on the first detection model, calculate the three-dimensional key point coordinates of the athlete and the arrow according to the identified two-dimensional key points of the athlete and the arrow, and obtain the first three-dimensional coordinate data;

[0085] The second image detection unit (32) is used to identify the two-dimensional key points of the arrow in the second image data based on the second detection model, and calculate the three-dimensional key point coordinates of the arrow by combining the identified two-dimensional key points of the arrow and the three-dimensional coordinates calibrated by the arrow target position to obtain the second three-dimensional coordinate data. The second three-dimensional coordinate data includes the target hitting position coordinates of the arrow;

[0086] The point cloud detection unit (33) is used to identify the arrow point cloud data in the archery point cloud data based on the third detection model;

[0087] The comprehensive detection unit (34) is respectively connected to the first image detection unit (31), the second image detection unit (32) and the point cloud detection unit (33), and is used to fuse the first three-dimensional coordinate data, the second three-dimensional coordinate data and the arrow point cloud data to obtain the three-dimensional trajectory data of the arrow.

[0088] In the algorithm processing stage, it includes time synchronization, data acquisition, arrow recognition, trajectory fusion, etc.

[0089] The intelligent recognition module (3) performs time synchronization on multiple industrial cameras in the image acquisition device (1) and multiple lidars in the lidar device (2) to ensure that there is only a time error of milliseconds. The time error will affect the final coordinate deviation, which is very important for the reconstruction of the position coordinates of the arrow moving at high speed in the archery competition.

[0090] After the athlete prepares to start archery, each industrial camera in the image acquisition device starts taking pictures at a speed of 60 frames per second, and the lidar device starts to collect the arrow point cloud and its depth information. Specifically, the image acquisition device and the lidar device will transmit the collected data to the intelligent recognition module in real time.

[0091] The recognition module (3) performs algorithm analysis on the images captured by each industrial camera and the arrow point cloud data generated by the lidar to identify and track the arrows, and obtains the three-dimensional motion trajectory data of the arrows.

[0092] Specifically, the first detection model is the Retina Net Algorithm model. By using the first detection model to detect the two-dimensional key points of the archer (athlete, competitor) and the arrow in the image data of the starting point area of the arrow path captured by multiple industrial cameras, and then, the two-dimensional key points of the archer and the arrow identified by multiple industrial cameras in the starting point area of the arrow path are reconstructed three-dimensionally to obtain the first three-dimensional coordinate data. Specifically, the three-dimensional key point coordinates are reconstructed through a 3D triangulation model to form the first three-dimensional coordinate data.

[0093] RetinaNet is a one-stage object detection network. Different from traditional two-stage detectors (such as the R-CNN series), RetinaNet achieves end-to-end object detection through a single network. Its core idea is to use Focal Loss to solve the problem of class imbalance during the training process. In terms of the backbone network, RetinaNet adopts classic convolutional neural network structures such as ResNet, and improves the feature extraction ability by continuously deepening the network layers. At the same time, in order to make full use of feature information at different scales, RetinaNet also introduces the FPN (Feature Pyramid Network) structure to fuse features at different levels, thereby improving the model's detection ability for multi-scale objects. RetinaNet uses a combination of anchor boxes and convolutional neural networks for object detection. Specifically, RetinaNet first generates a series of anchor boxes on different feature maps, and then classifies and regresses each anchor box through the convolutional neural network to obtain the category and location information of the object.

[0094] Specifically, the second detection model is the Transformer Model model. The second detection model is used to identify the two-dimensional key points of the arrow in the image data of the target area captured by multiple industrial cameras. Then, the second image detection unit (32) reconstructs the two-dimensional key points of the arrow identified by multiple industrial cameras in the area near the tail, combined with the three-dimensional coordinates calibrated by the arrow target position, to obtain the second three-dimensional coordinate data. Combining the three-dimensional coordinates calibrated by the arrow target position is mainly to output the three-dimensional key point coordinates of the arrow hitting the target. If the arrow does not hit the target, the hitting position coordinates are output as none. If the arrow hits the target, the hitting position coordinates are the three-dimensional key point coordinates of the hitting point specifically calculated by combining the three-dimensional coordinates calibrated by the arrow target position.

[0095] The Transformer is essentially an Encoder-Decoder architecture. The middle part of the Transformer can be divided into two parts: the encoding component and the decoding component, the Transformer model (Encoder-Decoder architecture mode). Different from traditional models based on RNN (Recurrent Neural Network) or CNN (Convolutional Neural Network), the Transformer relies entirely on the attention mechanism, abandoning the sequential processing method, thus achieving parallel training and significantly improving the training efficiency and model performance.

[0096] Specifically, the third detection model is the DBSCAN clustering algorithm. The DBSCAN clustering algorithm is used to calculate and analyze the archery point cloud data to identify the arrow point cloud data.

[0097] DBSCAN (Density-Based Spatial Clustering of Applications with Noise) is a relatively representative density-based clustering algorithm. Different from partitioning and hierarchical clustering methods, it defines clusters as the largest set of density-connected points, can divide regions with sufficient high density into clusters, and can discover clusters of any shape in a spatial database with noise.

[0098] Specifically, the comprehensive detection unit corrects and fits the first three-dimensional coordinate data, the second three-dimensional coordinate data, and the arrow point cloud data through curve fitting (Curve-fitting) to obtain the three-dimensional arrow trajectory data, in order to restore the arrow release point, 3D flight trajectory, and target landing point. Specifically, the comprehensive detection unit also smooths the fitted trajectory function, that is, the trajectory three-dimensional coordinates, to obtain the final three-dimensional arrow trajectory data.

[0099] That is, the three-dimensional arrow trajectory data includes the three-dimensional coordinates of the arrow release point, the air movement trajectory, and the target hit point.

[0100] Furthermore, it also includes:

[0101] A video conversion module (6), connected to the three-dimensional engine module (5), is used to convert the archery event visualization model into a video signal to be sent to the broadcaster to present the archery event visualization model in the event broadcast screen.

[0102] The three-dimensional engine module (5) has previously modeled the competition venue, including the track, target, and arrow. When constructing the archery event visualization model, based on the models of the track, target, arrow, etc. that have been constructed, the arrow model is rendered based on the called three-dimensional arrow trajectory data to construct the archery event visualization model.

[0103] In the data application stage, the extracted three-dimensional trajectory data of the arrow can be applied to different scenarios. On the one hand, it can be used for event broadcasting. The three-dimensional engine module (5) analyzes the three-dimensional trajectory data of the arrow to form a visualization model of the archery event, which is converted into a video signal for output. The video signal of this visualization model of the archery event is superimposed and presented on the event broadcasting screen to present accurate and comprehensive event data on the event broadcasting screen. For example, effects such as the trajectory of a single arrow and the comparison of multiple arrows hitting the target are produced, reducing the viewing threshold for the audience. Based on the collection of on-site event data (starting point of shooting, arrow trajectory, target landing point), three-dimensional modeling and visual effect superimposition are carried out, thus enriching the display methods and visual effects of archery event broadcasting and achieving positive benefits in terms of the specialization, refinement, and interest of the broadcasting. On the other hand, the event data can be sorted and summarized. By analyzing the performance of each athlete in the event, data guidance and improvement direction suggestions are provided for post-match training. Further, the three-dimensional engine module (5) is used to: call the three-dimensional trajectory data of a single arrow shooting for data analysis to obtain the first analysis result of the single arrow shooting, and construct a visualization model of the archery event based on the first analysis result and the called three-dimensional trajectory data of the arrow.

[0104] As one of the application scenarios in the data application stage, it is to conduct data analysis and visualization modeling on the single arrow shooting of an athlete, analyze the starting point coordinates of the single arrow shooting, mark the starting point and present it visually.

[0105] By analyzing the three-dimensional trajectory data of the arrow, the three-dimensional coordinate information of the arrow in the air is obtained to get the first analysis result. The first analysis result includes, but is not limited to, the distance between the starting point and the starting line, the deviation distance between the aiming point and the final starting point, the height of the real-time arrow position from the ground, the maximum height of the trajectory parabola from the ground, the average speed of the arrow flight, the target ring number, etc.

[0106] Since the actual arrow shooting generally forms a parabola, after aiming, the arrow will be lifted up by a certain angle (starting point) and then fired. Therefore, the deviation distance between the aiming point and the final starting point refers to the deviation between the arrow coordinates at the aiming point and the actual arrow launching point coordinates.

[0107] After the visualization model of the archery event is visualized, the movement trajectory is visually presented. The visualization not only includes the visualization of the real-time movement trajectory of the rendered arrow, but also presents the first analysis result, that is, presents information including, but not limited to, the distance between the starting point and the starting line, the maximum height of the trajectory parabola from the ground, the target ring number, etc. The analysis parameters and the parameters to be visualized can be set as needed, and the analysis result is added to the model and then visualized.

[0108] Further, the 3D engine module (5) is used to: call the 3D trajectory data of at least two consecutive archery shots of the same athlete for data comparison and analysis to obtain a second analysis result of multiple archery shots of a single athlete, and construct an archery event visualization model based on the second analysis result and the called 3D trajectory data of the arrows.

[0109] As one of the application scenarios in the data application stage, it is also possible to perform comparison and analysis on several consecutive archery shots of an athlete, such as three consecutive archery shots. The second analysis result includes the average trajectory of multiple archery shots, the real-time deviation comparison between a single archery shot and the average trajectory, the comparison of the maximum height from the ground of the trajectories of multiple archery shots, the comparison result of the average speed of multiple archery shots, and the comparison of the target hitting distance deviation of multiple archery shots. These comparison and analysis results can be visually presented. The visual presentation renders and constructs the trajectories of multiple archery shots together, and the average trajectory visualization can be added for the audience to intuitively feel the differences between different archery shots. When presenting the figure, text and symbol content are also presented, such as marking the actual values of the target hitting distance deviation and the actual values of the maximum height from the ground of the trajectory in the figure. Further, the 3D engine module (5) is used to: call the 3D trajectory data of each winning arrow for data comparison and analysis to obtain a third analysis result regarding the winning arrow, and construct an archery event visualization model based on the third analysis result and the called 3D trajectory data of each winning arrow.

[0110] The winning arrow refers to the arrow shot at a critical moment in an archery competition, which may determine the outcome.

[0111] The present invention can not only perform comparison and analysis on multiple archery shots of the same athlete and present them during the live broadcast, but also perform comparison and analysis on the archery shots between different athletes. The analysis results are all helpful for providing data guidance and improvement direction suggestions for the athletes' post-match training.

[0112] Specifically, the third analysis result includes the real-time flight trajectory and the real-time trajectory deviation distance between two arrows, the target hitting deviation distance and the deviation distance from the bull's-eye between the two arrows, the comparison of the maximum height from the ground of the trajectories of the two arrows, the comparison of the average flight speed of the two arrows, etc. The third analysis result is marked in the archery event visualization model to visually present the comparison results.

[0113] As an implementation manner of the present invention, it may not be the winning arrow, but other archery shots of both sides of the competition. The 3D engine module (5) is used to: call the 3D trajectory data of one archery shot of each different athlete for data comparison and analysis to obtain a fourth analysis result, and construct an archery event visualization model based on the fourth analysis result and the called 3D trajectory data of the arrows.

[0114] See Figure 3 , the present invention also provides an archery event capture and visualization method, using the aforementioned archery event capture and visualization system, including:

[0115] Step A1: The image acquisition device acquires the image data during the archery competition, and the lidar device acquires the archery point cloud data on the arrow path during the archery competition.

[0116] Step A2: The intelligent recognition module performs arrow recognition processing on the image data and the archery point cloud data to obtain the three-dimensional arrow trajectory data.

[0117] Step A3: The storage module stores the three-dimensional arrow trajectory data.

[0118] Step A4: The 3D engine module calls the three-dimensional arrow trajectory data from the storage module to establish a three-dimensional visualization model of the archery event.

[0119] Furthermore, after step A4, it further includes:

[0120] Step A5: The video conversion module converts the three-dimensional visualization model of the archery event into a video signal to be sent to the broadcaster to present the three-dimensional visualization model of the archery event in the event broadcast screen.

[0121] The above are only the preferred embodiments of the present invention, and do not limit the implementation manners and protection scope of the present invention. For those skilled in the art, it should be able to realize that all equivalent replacements and obvious changes made by using the description and illustrations of the present invention should be included in the protection scope of the present invention.

Claims

1. An archery event capture visualization system, characterized in that: include: Image acquisition devices are distributed in the area near the head and the area near the tail of the arrow path on the competition field, and are used to collect image data during the archery competition; The laser radar device is distributed on the outer side of the archery path and is used to collect the archery point cloud data on the archery path during the archery competition; An intelligent recognition module, connected to the image acquisition device and the laser radar device, respectively, for performing arrow recognition processing on the image data and the archery point cloud data to obtain three-dimensional arrow trajectory data; A storage module, connected to the intelligent recognition module, and used to store the three-dimensional trajectory data of the arrow; The three-dimensional engine module is connected to the storage module and is used to call the three-dimensional trajectory data of the arrow to establish a three-dimensional archery event visualization model.

2. The archery event capture visualization system according to claim 1, characterized in that: Also includes: The video conversion module is connected to the three-dimensional engine module and is used to convert the archery event visualization model into a video signal to be sent to the broadcaster to present the archery event visualization model in the event broadcast screen.

3. The archery event capture visualization system according to claim 1, characterized in that: The image acquisition device in the area near the head of the arrow path is used to acquire first image data of the starting point area, and the image acquisition device in the area near the tail of the arrow path is used to acquire second image data of the target surface area; The intelligent recognition module comprises: A first image detection unit is used to identify two-dimensional key points of the athlete and the arrow in the first image data based on a first detection model, and calculate three-dimensional key point coordinates of the athlete and the arrow according to the identified two-dimensional key points of the athlete and the arrow to obtain first three-dimensional coordinate data; A second image detection unit is used to identify two-dimensional key points of the arrow in the second image data based on a second detection model, calculate the three-dimensional key point coordinates of the arrow in combination with the identified two-dimensional key points of the arrow and the three-dimensional coordinates of the target position calibration, and obtain second three-dimensional coordinate data, wherein the second three-dimensional coordinate data includes the coordinates of the arrow's target position; a point cloud detection unit, configured to identify arrow point cloud data in the archery point cloud data based on a third detection model; A comprehensive detection unit is respectively connected to the first image detection unit, the second image detection unit and the point cloud detection unit, and is used to fuse the first three-dimensional coordinate data, the second three-dimensional coordinate data and the arrow point cloud data to obtain the arrow three-dimensional trajectory data.

4. The archery event capture visualization system according to claim 1, characterized in that: The three-dimensional engine module is used to: call the three-dimensional trajectory data of the arrow of a single archery shot to perform data analysis to obtain a first analysis result of the single archery shot, and build the archery event visualization model based on the first analysis result and the called three-dimensional trajectory data of the arrow.

5. The archery event capture visualization system according to claim 1, characterized in that: The three-dimensional engine module is used to: call the three-dimensional trajectory data of the arrow from at least two consecutive archery shots by the same athlete to perform data comparison and analysis to obtain a second analysis result of multiple archery shots by a single athlete, and construct the archery event visualization model based on the second analysis result and the called three-dimensional trajectory data of the arrow.

6. The archery event capture visualization system according to claim 1, characterized in that: The three-dimensional engine module is used to: call the three-dimensional trajectory data of each winning arrow to perform data comparison and analysis to obtain a third analysis result about the winning arrow, and build the archery event visualization model based on the third analysis result and the called three-dimensional trajectory data of each winning arrow.

7. The archery event capture visualization system according to claim 3, characterized in that: The image acquisition device in the area near the arrow path head is four industrial cameras, with two of the industrial cameras distributed on both sides of the arrow path head; The image acquisition device in the area near the tail of the arrow path is four industrial cameras; two of the industrial cameras are distributed on each side of the tail of the arrow path; The laser radar device includes at least two laser radars, and both laser radars are located on the same outer side of the arrow path.

8. The archery event capture visualization system according to claim 1, characterized in that: The intelligent recognition module is used to perform arrow recognition processing after time synchronization of the image data and the archery point cloud data to obtain three-dimensional arrow trajectory data.

9. A method for capturing and visualizing an archery event, characterized in that: A system for capturing and visualizing an archery event as described in any one of claims 1 to 8, comprising: Step A1, the image acquisition device acquires image data during the archery competition, and the laser radar device acquires archery point cloud data on the arrow path during the archery competition; Step A2, the intelligent recognition module performs arrow recognition processing on the image data and the archery point cloud data to obtain three-dimensional arrow trajectory data; Step A3, a storage module stores the arrow three-dimensional trajectory data; Step A4: The three-dimensional engine module calls the three-dimensional trajectory data of the arrow from the storage module to establish a three-dimensional archery event visualization model.

10. The archery event capture visualization method according to claim 9, characterized in that: After step A4, the following steps are also included: Step A5: the video conversion module converts the archery event visualization model into a video signal, and sends it to the broadcaster to present the archery event visualization model in the event broadcast screen.

Images