Intelligent simulation system and method for flying disc shooting training based on fusion of vision and inertial navigation

Abstract

This application relates to the field of assisted shooting training technology, and discloses an intelligent simulation system and method for skeet shooting training. The system includes a gun body end unit, which includes an attitude measurement module and a shooting module. The attitude measurement module is installed inside the gun body and collects the gun body's motion data in real time. The shooting module is installed above or to the side of the gun body and collects image data of the target range in real time. A host computer is communicatively connected to the attitude measurement module and the shooting module and synchronized with their clocks. The host computer receives the motion data collected by the attitude measurement module and the image data collected by the shooting module, timestamps the motion data and image data, extracts the centroid coordinates of the target disc from the image data, fuses the centroid coordinates with the motion data to calculate the target disc's trajectory relative to the gun body, and calculates the user's training parameters. This application can collect and analyze user training data in real time to assist user training.

Description

Technical Field

[0001] This application relates to the field of skeet shooting training auxiliary equipment technology, and in particular to an intelligent simulation system and method for skeet shooting training based on vision and inertial navigation fusion. Background Technology

[0002] Skeet shooting is a sport that demands extremely high levels of reaction time, physical stability, and consistency from athletes. In competition or training, athletes must complete a series of actions—from spotting the target (skeet) to moving the gun, aiming, and firing—within 0.4 to 0.6 seconds. This extremely short time window makes it difficult for coaches to accurately assess and guide athletes' technical movements through visual observation.

[0003] Currently, existing skeet shooting auxiliary training equipment is usually based on technologies such as stereo vision, diffraction measurement, and GPS measurement. These technical solutions have the following problems.

[0004] 1. Stereo vision technology: It requires the installation of multiple cameras in the training field, the system is complex to install, has poor real-time performance, requires a large amount of computing power, and the target is easily lost due to occlusion during the athlete's rapid body rotation.

[0005] 2. Diffraction measurement technology: It has high requirements for lighting conditions and poor stability in outdoor lighting environments with varying lighting conditions, making it difficult to meet the needs of all-weather training.

[0006] 3. GPS measurement technology: The sampling frequency is low, which cannot meet the real-time attitude measurement requirements of fast-moving targets, and the measurement accuracy is insufficient.

[0007] In addition, existing training methods mainly rely on live-fire shooting and experience-based judgment, which are costly and cannot obtain quantitative data for scientific analysis of athletes' key technical parameters such as "three-dimensional displacement" (body displacement in three-dimensional space), gun handling speed, and changes in body center of gravity.

[0008] Therefore, there is an urgent need for an intelligent auxiliary training system that can collect athletes' training data in real time and accurately, and is easy to install and highly adaptable. Summary of the Invention

[0009] The purpose of this application is to provide an intelligent simulation system and method for skeet shooting training based on vision and inertial navigation fusion. By integrating high-precision sensors into the training firearms, the system can achieve real-time capture and quantitative analysis of athletes' movements, highly simulate the live-fire training environment, reduce training costs, and provide data support for scientific training.

[0010] In a first aspect, this application provides an intelligent simulation system for skeet shooting training, comprising: The gun body end unit includes an attitude measurement module and an imaging module. The attitude measurement module is installed inside the gun body and collects the motion data of the gun body in real time. The imaging module is fixedly installed above or on the side of the gun body barrel and collects image data of the target range in real time. The host computer is communicatively connected to the attitude measurement module and the imaging module and synchronized with their clocks. The host computer receives motion data collected by the attitude measurement module and image data collected by the imaging module, timestamps the motion data and image data, extracts the centroid coordinates of the target disc from the image data, fuses the centroid coordinates with the motion data to calculate the motion trajectory of the target disc relative to the gun body, and calculates the user's training parameters.

[0011] In a preferred embodiment, the gun body end unit further includes a data acquisition module and a communication module. The data acquisition module is connected to the attitude measurement module and the shooting module and realizes clock synchronization between the attitude measurement module and the shooting module. The communication module realizes the communication connection between the data acquisition module and the host computer and transmits the motion data and the image data to the host computer.

[0012] In a preferred embodiment, the optical axis of the imaging module is parallel to the axis of the gun barrel.

[0013] In a preferred embodiment, the attitude measurement module includes a MEMS inertial measurement unit, which includes an accelerometer and a gyroscope. The motion data includes the three-axis acceleration and three-axis angular velocity of the gun body measured by the accelerometer and the gyroscope, respectively. The sampling frequency of the attitude measurement module is greater than or equal to 500Hz.

[0014] In a preferred embodiment, the process of the host computer extracting the centroid coordinates of the target disk from the image data and fusing the centroid coordinates with the motion data further includes: Distortion correction is performed on the image data, adaptive threshold segmentation is performed on the image data within the detection box and the target disk region is extracted, and the centroid coordinates of the target disk are calculated using the gray-scale centroid method. The transformation relationship between the image coordinate system of the image data and the gun body coordinate system is established through the intrinsic parameters of the shooting module; The motion data is subjected to zero bias compensation and filtering, the motion data is updated using the quaternion method, and the three-axis acceleration and three-axis angular velocity data measured by the attitude measurement module are fused using extended Kalman filtering to output the attitude angle and position information of the gun body in the inertial coordinate system. The coordinates of the target disc in the image coordinate system are converted into a pointing deviation angle relative to the gun body coordinate system, and the motion trajectory of the target disc in the inertial coordinate system is obtained based on the attitude angle and the pointing deviation angle.

[0015] In a preferred embodiment, the centroid coordinates are ( , ),in , , For pixel grayscale values, ( , ) represents the pixel coordinates; the pointing deviation angle is (Δα, Δβ), where , ,in( , () represents the coordinates of the principal point of the image. , These are the focus coordinates of the shooting module in the image coordinate system.

[0016] In a preferred embodiment, the training parameters include reaction time, which is the time from when the target disc first appears in the image data to when the gun body moves.

[0017] In a preferred embodiment, the training parameters include the gun-carrying speed, which is obtained by calculating the angular acceleration based on the derivative of the angular velocity data.

[0018] In a preferred embodiment, the training parameters include the aiming trajectory, and the aiming accuracy is calculated based on the Euclidean distance between the centroid coordinates of the target disc and the center of the image.

[0019] In a second aspect, this application provides an intelligent simulation method for skeet shooting training, employing the aforementioned intelligent simulation system for skeet shooting training, the method comprising: The user holds the gun and prepares to fire, then selects the appropriate user in the host computer. As the user begins firing, the gun body end unit collects the motion data of the gun body and the image data of the target range in real time. The host computer calculates and displays the motion trajectory of the gun and the user's training parameters based on the motion data and the image data.

[0020] Compared with the prior art, the present invention has the following beneficial effects: 1. Solved the obstruction problem: The shooting module (or camera) is mounted on the gun body, and the shooting module moves with the gun body at the same time, keeping the target within the field of view at all times, fundamentally solving the problem of target loss caused by the athlete's body obstructing the view in the fixed camera solution.

[0021] 2. Improved real-time performance and accuracy: By adopting a high sampling rate inertial measurement unit (≥500Hz) and a high frame rate camera (≥120fps), it can accurately capture motion details within 0.4-0.6 seconds, achieving millisecond-level temporal resolution and sub-arcsecond-level spatial resolution.

[0022] 3. Multimodal data fusion was achieved: Visual information (target position) and inertial navigation information (gun attitude) were spatiotemporally fused, which can accurately calculate the relative motion relationship between the target and the muzzle, a function that cannot be achieved by a single sensor.

[0023] 4. Reduced training costs: The system supports "dry gun" training mode, allowing athletes to obtain complete motion data feedback without firing live ammunition, significantly reducing training ammunition costs and venue fees.

[0024] 5. Provides quantitative analysis basis: The system can generate quantitative data such as rifle carrying speed curve, aiming trajectory heat map, and body stability index, providing objective basis for coaches to scientifically guide training.

[0025] 6. High adaptability: The system has low requirements for lighting conditions and can adapt to varying outdoor lighting environments; it is easy to install, requires no modification to the training site, and is plug-and-play.

[0026] It should be understood that, within the scope of this invention, the above-described technical features of this invention and the technical features specifically described below (such as in the embodiments) can be combined with each other to form new or preferred technical solutions. Due to space limitations, they will not be described in detail here. Attached Figure Description

[0027] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. It should be understood that the accompanying drawings described below are merely some implementation examples of the present invention, and those skilled in the art can obtain other implementation examples based on these drawings without creative effort.

[0028] Figure 1 This is a schematic diagram of an intelligent simulation system for skeet shooting training according to one embodiment of this application.

[0029] Figure 2 This is a flowchart of data processing and fusion according to one embodiment of this application.

[0030] Figure 3 This is a display interface for training results according to one embodiment of this application.

[0031] Figure 4 This is a flowchart of an intelligent simulation method for skeet shooting training according to one embodiment of this application. Detailed Implementation

[0032] In the following description, many technical details are presented to help the reader better understand this application. However, those skilled in the art will understand that the technical solutions claimed in this application can be implemented even without these technical details and various variations and modifications based on the following embodiments.

[0033] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0034] The first embodiment of this application relates to an intelligent simulation system for skeet shooting training, the block diagram of which is shown below. Figure 1 As shown, the system includes a gun body end unit 10 and a host computer 20. The gun body end unit 10 includes an attitude measurement module and a shooting module (not shown, but installed, for example, at the location indicated by the dashed box in the figure). The attitude measurement module is installed inside the gun body and collects the gun body's motion data in real time. The shooting module is installed above or to the side of the gun barrel and collects image data of the firing range in real time. The shooting module can use a high frame rate camera, for example, with a frame rate greater than or equal to 120fps. The optical axis of the shooting module is preferably parallel to the axis of the gun barrel. The gun body end unit 10 also includes a data acquisition module and a communication module (not shown, but installed, for example, at the location indicated by the dashed box in the figure). The data acquisition module is connected to the attitude measurement module and the shooting module and synchronizes their clocks. The communication module establishes a communication connection between the data acquisition module and the host computer 20 and transmits motion data and image data to the host computer 20. The communication connection can be any wired or wireless connection method.

[0035] The host computer 20 communicates with the attitude measurement module and the imaging module and synchronizes with their clocks. The host computer 20 receives motion data collected by the attitude measurement module and image data collected by the imaging module, timestamps the motion data and image data, extracts the centroid coordinates of the target disc from the image data, fuses the centroid coordinates with the motion data to calculate the motion trajectory of the target disc relative to the gun body, and calculates the user's training parameters.

[0036] In one embodiment, the attitude measurement module can be a MEMS inertial measurement unit (IMU) (or inertial navigation system). The MEMS inertial measurement unit includes an accelerometer and a gyroscope. The motion data includes the three-axis acceleration and three-axis angular velocity of the gun body measured by the accelerometer and gyroscope respectively, or collectively referred to as inertial navigation data. The sampling frequency of the attitude measurement module is greater than or equal to 500Hz, for example, 1000Hz.

[0037] The host computer 20 performs fusion calculations on the image data and motion data to obtain the target disc's trajectory and the user's (or athlete's) training data. For example... Figure 2 As shown, the specific steps include: First, distortion correction is performed on the image data, and zero-bias compensation and filtering are performed on the motion data; adaptive threshold segmentation is performed on the image data within the detection box to extract the target disc region, and the centroid coordinates of the target disc are calculated using the gray-scale centroid method; the attitude is updated using the quaternion method, and the three-axis acceleration and three-axis angular velocity data measured by the attitude measurement module are fused using extended Kalman filtering to output the attitude angle and position information of the gun body in the inertial coordinate system; the transformation relationship between the image coordinate system and the inertial coordinate system of the image data is established; the coordinates of the target disc in the image coordinate system are converted into the pointing deviation angle relative to the inertial coordinate system, thereby obtaining the motion trajectory of the target disc relative to the gun body end unit 10, and further calculating the user's training parameters. The training parameters may include gun carrying speed, aiming trajectory, aiming accuracy, body stability, reaction time, firing timing, etc.

[0038] This invention integrates high-precision images and inertial navigation sensors into training firearms to achieve real-time capture and quantitative analysis of athletes' movements, highly simulating a live-fire training environment, reducing training costs, and providing data support for scientific training.

[0039] To better understand the technical solution of this application, specific examples are provided below. The details listed in these examples are for ease of understanding and are not intended to limit the scope of protection of this application.

[0040] This invention provides an intelligent simulation system for UFO shooting based on the fusion of vision and inertial navigation, including a gun-end unit and a host computer.

[0041] The gun's end unit includes the gun body, attitude measurement module, imaging module, data acquisition module, and communication module. The gun body is a standard skeet training weapon, such as the domestically produced TL-12 skeet training gun. Modifications are made to the standard skeet training weapon, allowing for pre-installation space inside the stock. Without altering the weapon's weight distribution, center of gravity, or the athlete's grip, the attitude measurement module can be miniaturized and integrated into the stock or handguard.

[0042] The attitude measurement module employs an industrial-grade or tactical-grade MEMS inertial measurement unit to acquire real-time three-axis angular velocity (including pitch, roll, and yaw) and three-axis acceleration data of the gun body. In this embodiment, an Analog Devices ADIS16470 tactical-grade MEMS-IMU is used, with a sampling rate set to 1kHz and a measurement range of ±2000° / s for angular velocity and ±40g for acceleration.

[0043] The imaging module includes a global shutter CMOS or CCD industrial camera and an optical lens, which is fixed above or to the side of the gun barrel using a special clamp. The camera's optical axis is parallel to the gun barrel axis, and it is used to capture range images during the athlete's aiming process. In this embodiment, the imaging module uses a Basler acA1300-200μm industrial camera with a resolution of 640×480, a frame rate of 240fps, and is equipped with an 8mm wide-angle lens.

[0044] The host computer comprises a high-performance portable workstation and a custom-developed data processing platform for receiving and processing data uploaded by the gun-mounted unit. The host computer utilizes a Dell Precision 7750 mobile workstation, equipped with an Intel i9 processor, 32GB of RAM, and an NVIDIA RTX 5000 graphics card. It is based on a hybrid Python and C++ programming language, using PyQt for the user interface, OpenCV for image processing, and TensorFlow for deploying deep learning models. The host computer establishes a communication connection with the gun-mounted unit via wireless or wired means and performs clock synchronization. The data acquisition module uses an STM32H7 series microcontroller, responsible for IMU data acquisition and camera trigger synchronization. The communication module employs a 5.8GHz Wi-Fi module to achieve wireless data transmission with the host computer.

[0045] The data processing platform includes a spatiotemporal synchronization module, an image processing module, an inertial navigation data processing module, a data fusion module, and a parameter generation module. The spatiotemporal synchronization module performs high-precision timestamp alignment between the image sequences acquired by the imaging module and the inertial navigation data acquired by the attitude measurement module. The image processing module includes a background estimation algorithm, a deep learning target detection model, and a centroid extraction algorithm, used to identify the target butterfly from a continuous image sequence and extract its centroid coordinates. The inertial navigation data processing module performs attitude calculation and Kalman filtering on the inertial navigation data to reconstruct the gun's trajectory in three-dimensional space. The data fusion module fuses the target butterfly's centroid coordinates with the gun's attitude data to calculate the target butterfly's trajectory relative to the muzzle. The parameter generation module calculates key technical parameters such as carrying speed, aiming accuracy, body stability, firing timing, and reaction time based on the fused data.

[0046] The following is a detailed explanation of the system workflow.

[0047] (1) System startup phase 1) Power on the host computer, and power on the imaging module and attitude measurement module; 2) Establish a communication connection, and the host computer interconnects and synchronizes the clock with the shooting module and attitude measurement module; 3) The shooting module acquires the current target range background image, and the attitude measurement module acquires the initial attitude information of the gun body and uploads it to the host computer for background modeling and initial calibration.

[0048] (2) Training data collection phase 1) The athlete aims the gun at the shooting range, and the host computer continuously receives video streams and inertial navigation data streams; 2) When the target butterfly appears, the system begins recording through automatic detection or manual triggering; 3) Record the entire process from the appearance of the target butterfly to the athlete pulling the trigger, including image sequences and inertial navigation data. All data are accompanied by high-precision timestamps.

[0049] (3) Data processing and fusion The host computer processes the collected data in real time, including: a) Identify target butterflies and extract their centroids using background estimation algorithms and deep learning models; b) Perform attitude calculation on the inertial navigation data to reconstruct the gun's motion trajectory; c) The centroid coordinates of the target butterfly and the attitude of the gun body are spatiotemporally fused to calculate the relative motion trajectory; d) Calculate parameters such as gun handling speed, aiming accuracy, body stability, firing timing, and reaction time.

[0050] The specific algorithms for data processing and fusion include the following steps 1 to 7.

[0051] Step 1: Data Preprocessing 1) Perform distortion correction on the image sequence acquired by the camera; the image sequence acquired by the camera belongs to the image coordinate system, which is a two-dimensional coordinate system with the origin at the center of the image and the unit is pixels (u, v). The image coordinate system is defined by the imaging module (camera) and is used to describe the position (centroid coordinates) of the target disk in the image.

[0052] 2) Perform zero-bias compensation and filtering on the IMU data; the IMU data belongs to the inertial coordinate system (or navigation coordinate system), which is a three-dimensional fixed coordinate system (such as the "north-south-sky" coordinate system). The origin is fixed at a certain point on the ground, the X-axis points east, the Y-axis points north, and the Z-axis points to the sky, and it does not move with the gun body. The inertial / navigation coordinate system is defined by the attitude measurement module (IMU) and is used to describe the absolute attitude (pitch angle θ, roll angle φ, yaw angle ψ) and position of the gun body in space.

[0053] Step 2: Target Butterfly Detection and Tracking 1) An improved model was used to detect target butterflies. The model was trained on a self-made UFO dataset. 2) An algorithm is used to perform multi-target tracking on the detected target discs, maintaining consistency in identity ID. During a training session, the athlete will perform multiple target disc shooting operations, tracking multiple target discs and maintaining consistency with the current athlete's identity.

[0054] Step 3: Extraction of the target butterfly's centroid 1) Adaptive threshold segmentation is used within the detection frame to extract the target butterfly region; 2) Calculate the sub-pixel centroid coordinates of the target butterfly using the gray-level centroid method: The gray-level centroid method is a feature localization method commonly used in image processing and computer vision. Essentially, it treats the gray-level distribution of the image as a mass distribution and determines the position of the target by calculating the centroid. The coordinates of the centroid are ( , ),in , , For pixel grayscale values, ( , ) represents the pixel coordinates, n represents the number of pixels in the camera in the algorithm, and i represents the pixel number.

[0055] Step 4: Attitude calculation of inertial navigation data 1) Use quaternions for pose update; 2) An extended Kalman filter is used to fuse accelerometer and gyroscope data to suppress integral drift. The Kalman filter is a recursive algorithm for dynamic system state estimation. It makes the optimal estimate of the system state (in the sense of minimum mean square error) in the presence of noise and uncertainty. 3) Output the attitude angles of the gun body in the inertial / navigation coordinate system (gun body coordinate system transformed to inertial / navigation coordinate system). And position information. Where θ is the pitch angle, φ is the roll angle, and ψ is the yaw angle.

[0056] Step 5: Spatiotemporal Synchronization and Coordinate Transformation 1) Establish the transformation relationship between the image coordinate system and the gun body coordinate system, and the camera intrinsic parameters ( , , , The target disc's centroid coordinates (u, v) are obtained through pre-calibration; the gun body coordinate system is a three-dimensional coordinate system with its origin at the camera's optical center (or a fixed reference point on the gun body), where the X-axis is to the right, the Y-axis is downward (or forward), and the Z-axis is forward (or right). The camera is fixedly connected to the gun body and moves with it. The gun body coordinate system is defined by the gun body end elements and is used to describe the direction of the target disc relative to the muzzle. Points in the image coordinate system can be transformed to their orientation angles (Δα, Δβ) in the gun body coordinate system using the camera's intrinsic parameters.

[0057] 2) Convert the target butterfly's coordinates (u, v) in the image coordinate system to its pointing deviation angles (Δα, Δβ) relative to the gun body coordinate system, where , ,in( , ) represents the coordinates of the principal point (i.e., the center point of the image). , These are the focus coordinates of the shooting module in the image coordinate system.

[0058] The inertial coordinate system is a globally fixed reference system, while the gun body coordinate system is a moving system fixed to the gun body. The gun body attitude angles measured by the IMU describe the rotational relationship of the gun body coordinate system relative to the inertial coordinate system. The image coordinate system and the gun body coordinate system have a fixed geometric relationship established through camera calibration (because the camera is fixed to the gun, their relative attitudes are known and unchanging).

[0059] The core algorithm is to add the position of the target disc in the image (pixel offset) to the absolute orientation of the gun body in inertial space (attitude angle) to obtain the absolute motion trajectory of the target disc in the inertial coordinate system.

[0060] Step 6: Relative trajectory fusion 1) The pointing angle of the target butterfly relative to the inertial coordinate system is a combination of the gun attitude angle and the image deviation angle; Combine the gun attitude angle with the image deviation angle: in αtarget The horizontal orientation of the target disc in the inertial coordinate system; βtarget Let be the pitch angle of the target disc in the inertial coordinate system.

[0061] 2) Obtain the trajectory of the target disc relative to the muzzle to analyze aiming errors. The aiming error angle of the target disc relative to the muzzle (barrel axis) is: The aiming error at the moment of firing is The relative motion trajectory of the target disc is in Indicates time, The time when the system started recording data. For the firing moment, It is a normalized direction vector that represents the unit pointing direction of the target disc in the inertial coordinate system.

[0062] Step 7: Calculation of key parameters 1) Gun movement speed: Differentiate the three-axis angular velocity data measured by the gyroscope to obtain the angular acceleration curve; extract the maximum angular velocity and average angular velocity from start to firing; Define the three-axis angular velocity vector measured by the gyroscope as follows: It should be understood that the roll angle φ, pitch angle θ, and yaw angle ψ are the three-axis angular velocities. It is obtained by integrating the components in the x, y, and z directions.

[0063] angular acceleration is Maximum angular velocity of gun transport is The average angular velocity of the gun is in The moment the gun begins to move. tfire The firing moment.

[0064] 2) Aiming accuracy: Calculate the Euclidean distance between the centroid of the target butterfly and the center of the image in each frame, and generate an aiming error curve; , t ∈[ ttarget , tfire ] ttarget This marks the moment the target disc first appeared.

[0065] 3) Body stability: High-frequency filtering is performed on the triaxial acceleration data measured by the accelerometer to extract the athlete's body sway component and calculate the sway amplitude and frequency; Let the triaxial acceleration vector measured by the accelerometer be... The extracted high-frequency shaking component is LPF(a(t)) is the result of low-pass filtering the high-frequency jitter component; The amplitude of the sway (i.e., the root mean square of the high-frequency sway component during the ΔT time interval before firing) is The smaller the sway amplitude, the more stable the athlete's body is before the shot. Coaches can use this to assess the athlete's core strength and mental stability, and can make horizontal comparisons between different athletes or track the training progress of the same athlete vertically. The main frequency of the shaking is 4) Reaction time: The reaction time is obtained by subtracting the time when the target butterfly appears from the moment when the angular velocity of the detection gun body first exceeds the threshold (e.g., 20° / s).

[0066] Define the angular velocity threshold as The moment the gun body begins to move is The reaction time is (4) Feedback phase The processing results are presented in a visual format, including slow-motion replays, motion trajectory curves, and key parameter charts, for coaches and athletes to analyze in real time.

[0067] The following provides a complete training scenario example.

[0068] The athlete used this system for 50 dry-fire training sessions. In one training session: 1. Preparation phase: The athlete stands at the shooting position with the gun. The coach selects "Athlete: Zhang San, Training Group: Group 1" in the host computer software. The system self-check is normal.

[0069] 2. Data Collection 1) 0ms: The target launcher launches the target disc; 2) 80ms: The host computer's vision algorithm detects the target butterfly and automatically starts caching data; 3) 100-580ms: Tracking the athlete's movement with a rifle. The inertial navigation IMU records the movement of the rifle at 1000Hz, and the camera records the footage at 240fps. 4) 600ms: The athlete pulls the trigger (dry gun), the gun vibrates and triggers the firing marker, the system stops recording.

[0070] 3. Data Analysis The system completes data processing within 2 seconds, displays the following results, and... Figure 3 The interface displays: 1) Maximum angular velocity of the rifle: 285° / s (standard value 280-300° / s, good); 2) Aiming deviation at the moment of firing: 0.35° (leaning to the upper right, requires adjustment); 3) Body sway amplitude: 0.12g (with obvious fluctuations 0.1s before firing); 4) Reaction time: 0.19s (excellent); 5) Generate a heatmap of the aiming trajectory to show the deviation of the aiming point from the target butterfly trajectory at the moment of firing.

[0071] 4. Coaching guidance: Based on the data, the coach points out the athlete's body stability issues at the moment of firing and arranges targeted core strength training.

[0072] After two weeks of systematic training, the athletes' firing accuracy improved by 15%, while training ammunition consumption decreased by 60%.

[0073] Compared with existing technologies, this application does not simply superimpose general image processing algorithms with general inertial navigation calculation methods. This application is a tightly coupled fusion system designed for the specific field of skeet shooting training, specifically for the problem of calculating the relative motion between a high-speed instantaneous target (skeet) and the gun muzzle, where the time window is extremely short (0.4–0.6 s). It addresses two specific technical challenges: target loss due to rapid athlete rotation and the lack of relative motion data between the gun and the target. Traditional methods cannot work stably under conditions without long-term feature tracking.

[0074] First, this application fixes the shooting module (camera) to the gun body, and the shooting module moves with the gun, which fundamentally eliminates the problem of the athlete's body blocking the target disc, which is unavoidable in the fixed camera solution.

[0075] Secondly, in response to the drastic changes in the image background caused by the camera's movement with the gun, this application employs a high frame rate (≥120fps) global shutter camera and a dedicated target detection and sub-pixel centroid extraction algorithm to ensure stable extraction of target coordinates even under high dynamic conditions.

[0076] Furthermore, for the 0.4-0.6 second action window of UFO shooting, this application adopts a high sampling rate (≥500Hz) MEMS-IMU and realizes hardware-level microsecond-level clock synchronization between the camera and the IMU to ensure that complete action details can be captured.

[0077] Finally, the core inventive point of this application lies in the fact that, under the premise of strict timestamp alignment of the two sensors, a coordinate transformation model is established to tightly couple and fuse the pointing deviation of the target disc relative to the gun body obtained based on vision with the absolute attitude of the gun body obtained based on inertial navigation. This allows for the accurate calculation of the target disc's trajectory relative to the gun body, and based on this, a series of parameters specifically for skeet shooting training and evaluation (such as reaction time, gun handling speed curve, aiming deviation at the moment of firing, etc.) are calculated. This fusion method and parameter system cannot be achieved by a single vision system or a single inertial navigation system, and is also different from the general vision-inertial navigation fusion SLAM method. Together, these systems constitute a complete, closed-loop feedback intelligent simulation training platform.

[0078] One embodiment of this application also relates to an intelligent simulation method for skeet shooting training, the process of which is as follows: Figure 4As shown, the method includes the following steps: the user holds the gun and prepares to fire, and selects the corresponding user in the host computer; the user starts firing, and the gun end unit collects the motion data of the gun and the image data of the target range in real time; the host computer calculates and displays the motion trajectory of the gun and the user's training parameters based on the motion data and image data.

[0079] Accordingly, embodiments of this application also provide an intelligent simulation system for skeet shooting training, including a memory for storing computer-executable instructions and a processor; the processor is used to implement the steps in the above-described method embodiments when executing the computer-executable instructions in the memory. The processor may be a Central Processing Unit (CPU), Graphics Processing Unit (GPU), Digital Signal Processor (DSP), Microcontroller Unit (MCU), Neural Processing Unit (NPU), Application Specific Integrated Circuit (ASIC), Field Programmable Gate Array (FPGA), or other programmable logic devices. The aforementioned memory may be read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or solid-state drive, etc. The steps of the methods disclosed in the embodiments of this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules in the processor.

[0080] Furthermore, embodiments of this application also provide a computer program product, including computer-executable instructions that, when executed by a processor, implement the steps in the above-described method embodiments.

[0081] The various technical features disclosed in the above-described invention, the various technical features disclosed in the following embodiments and examples, and the various technical features disclosed in the accompanying drawings can be freely combined to form various new technical solutions (all of which should be considered as having been recorded in this specification), unless such a combination of technical features is technically infeasible. For example, in one example, feature A+B+C is disclosed, and in another example, feature A+B+D+E is disclosed. Features C and D are equivalent technical means that serve the same function, and technically only one needs to be used; it is impossible to use both simultaneously. Feature E can be technically combined with feature C. Therefore, the solution A+B+C+D should not be considered as having been recorded because it is technically infeasible, while the solution A+B+C+E should be considered as having been recorded.

[0082] All references to this specification are considered to be incorporated integrally into the disclosure of this application so that they can serve as the basis for modifications if necessary. Furthermore, it should be understood that the above descriptions are merely preferred embodiments of this specification and are not intended to limit the scope of protection of this specification. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of one or more embodiments of this specification should be included within the scope of protection of one or more embodiments of this specification.

[0083] In some cases, the actions or steps described in the claims can be performed in a different order than that shown in the embodiments and still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. An intelligent simulation system for skeet shooting training, characterized in that, include: The gun body end unit includes an attitude measurement module and an imaging module. The attitude measurement module is installed inside the gun body and collects the motion data of the gun body in real time. The imaging module is fixedly installed above or on the side of the gun body barrel and collects image data of the target range in real time. The host computer is communicatively connected to the attitude measurement module and the imaging module and synchronized with their clocks. The host computer receives motion data collected by the attitude measurement module and image data collected by the imaging module, timestamps the motion data and image data, extracts the centroid coordinates of the target disc from the image data, fuses the centroid coordinates with the motion data to calculate the motion trajectory of the target disc relative to the gun body, and calculates the user's training parameters.

2. The system as described in claim 1, characterized in that, The gun body end unit also includes a data acquisition module and a communication module. The data acquisition module is connected to the attitude measurement module and the shooting module and realizes clock synchronization between the attitude measurement module and the shooting module. The communication module realizes the communication connection between the data acquisition module and the host computer and transmits the motion data and the image data to the host computer.

3. The system of claim 1, wherein, The optical axis of the imaging module is parallel to the axis of the gun barrel.

4. The system of claim 1, wherein, The attitude measurement module includes a MEMS inertial measurement unit, which includes an accelerometer and a gyroscope. The motion data includes the three-axis acceleration and three-axis angular velocity of the gun body measured by the accelerometer and gyroscope, respectively. The sampling frequency of the attitude measurement module is greater than or equal to 500Hz.

5. The system of claim 4, wherein, The host computer extracts the centroid coordinates of the target disc from the image data and performs a fusion calculation with the motion data, further including: Distortion correction is performed on the image data, adaptive threshold segmentation is performed on the image data within the detection box and the target disk region is extracted, and the centroid coordinates of the target disk are calculated using the gray-scale centroid method. The conversion relationship between the image coordinate system of the image data and the gun body coordinate system is established through the intrinsic parameters of the shooting module; The motion data is subjected to zero bias compensation and filtering, the motion data is updated using the quaternion method, and the three-axis acceleration and three-axis angular velocity data measured by the attitude measurement module are fused using extended Kalman filtering to output the attitude angle and position information of the gun body in the inertial coordinate system. The coordinates of the target disc in the image coordinate system are converted into a pointing deviation angle relative to the gun body coordinate system, and the motion trajectory of the target disc in the inertial coordinate system is obtained based on the attitude angle and the pointing deviation angle.

6. The system of claim 5, wherein, The centroid coordinates are ( , ),in , , For pixel grayscale values, ( , ) represents the pixel coordinates; the pointing deviation angle is (Δα, Δβ), where , ,in( , () represents the coordinates of the principal point of the image. , These are the focus coordinates of the shooting module in the image coordinate system.

7. The system of claim 5, wherein, The training parameters include reaction time, which is the time from when the target disc first appears in the image data to when the gun body moves.

8. The system as described in claim 5, characterized in that, The training parameters include the gun-carrying speed, which is obtained by calculating the angular acceleration based on the differential of the angular velocity data.

9. The system of claim 5, wherein, The training parameters include the aiming trajectory, and the aiming accuracy is calculated based on the Euclidean distance between the centroid coordinates of the target disc and the center of the image.

10. An intelligent simulation method for skeet shooting training, characterized in that, The method, employing the system as described in any one of claims 1 to 9, comprises: The user holds the gun and prepares to fire, then selects the appropriate user in the host computer. As the user begins firing, the gun body end unit collects the motion data of the gun body and the image data of the target range in real time. The host computer calculates and displays the motion trajectory of the gun and the user's training parameters based on the motion data and the image data.

Images