Intelligent shooting examination training system and method

Abstract

The invention relates to the technical field of intelligent training, in particular to an intelligent shooting examination training system and method.The intelligent shooting examination training system comprises a posture recognition module, a track monitoring module, a hit analysis module, a rhythm comparison module and a label output module. According to the method, on the basis of action images in the process from gun holding to firing of a trainee, synchronous angle tracks of shoulders and elbows are recognized in real time, comparison is conducted in combination with a standard action combination, the matching difference in each time period is accurately screened and quantified, and the action control stability is dynamically evaluated by analyzing laser track coordinate points and direction changes in the aiming period; structured scores are formed by using a hit point sequence and combining path ductility and hit area density, action unit starting and ending time in a training process is compared with rhythm arrangement one by one, advanced and delayed phenomena are identified, performance scores are flexibly adjusted according to multi-dimensional index weights, and accurate matching between training performance evaluation and actual ability requirements is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent training, and particularly to an intelligent shooting assessment and training system and method. Background Art

[0002] The technical field of intelligent training includes training support systems constructed by means of artificial intelligence, sensor fusion, virtual simulation, etc., aiming to achieve accurate capture of the training process, real-time feedback of data processing, and quantitative presentation of training evaluation. The core content is to digitally express the operation behaviors, physical states, and task execution processes of trainees through multi-source information perception means, and to evaluate and optimize the training effects with the help of analysis algorithms, covering the coordinated operation of sensor layout, behavior capture devices, evaluation terminals, and control systems, supporting process management, dynamic supervision, and task assessment of training projects, forming a closed-loop system from training implementation to feedback guidance, and improving the scientificity and efficiency of training.

[0003] Among them, an intelligent shooting assessment and training system aims to apply a variety of data collection and processing means to accurately track and evaluate the entire process of shooting training. The system records and analyzes matters such as the standardization of shooting actions, the stability of aiming trajectories, the accuracy of bullet hitting positions, and reaction time nodes. Specifically, an image acquisition device is used to obtain changes in shooting postures, a laser positioning device is used to capture the movement trajectory of the firearm, an intelligent target device is used to detect the hitting position of the firing, and a timer device is used to record the reaction time from the appearance of the target to the completion of the firing. The training data is uniformly collected by the data collection terminal and handed over to the processing unit for feature extraction, numerical classification, and rule comparison, forming a training record and outputting an assessment basis for subsequent training performance evaluation and comparison reference.

[0004] Traditional shooting assessment and training technologies independently process each link such as action data, aiming trajectories, hitting performances, and reaction times in the operation process, resulting in insufficient data correlation and a lack of linkage analysis of the entire process of actions and results. In the face of complex training scenarios, it is difficult to accurately track key action features and rhythm offsets, resulting in the omission of hidden offsets. The performance scores cannot fully reflect the actual requirements of task types, and in practical applications, it will cause the simplification and lack of pertinence of training performance evaluation, with weak adaptability to different ability structures, affecting the improvement of combat personnel's abilities and the pertinence of training. Summary of the Invention

[0005] In order to solve the technical problems existing in the prior art, an embodiment of the present invention provides an intelligent shooting assessment and training system and method. The technical solution is as follows: On the one hand, an intelligent shooting assessment and training system is provided, and the system includes: The posture recognition module analyzes real-time action images, recognizes the synchronous angular change trajectory of the shoulder and elbow according to the action images of the trainer during the process from holding the gun to firing, compares the corresponding angle combinations in the standard action, filters out the deviated segments and extracts the action deviation amplitude, and generates shoulder-elbow posture information; The trajectory monitoring module calls the shoulder-elbow posture information, obtains the continuous coordinate points of the laser pointer on the target surface during aiming, screens out the fluctuating trajectory segments by calculating the spatial distance and the orientation change angle between adjacent coordinate points, and evaluates the stability of the aiming action, and generates fluctuation frequency information; The hit analysis module calls the fluctuation frequency information, sorts the hit points on the target surface according to the firing time, connects adjacent hit points and analyzes the direction change trend, identifies the deviation path, and calculates the hit score by combining the extension continuity of the path and the coverage range of the hit area, and generates the landing point analysis result; The rhythm comparison module calls the landing point analysis result, analyzes the start time corresponding to each action unit in the training task, and by comparing with the preset rhythm arrangement, identifies the early and late trends of the action units, evaluates the fluctuation degree of the execution rhythm, and generates rhythm evaluation information.

[0006] As a further solution of the present invention, the shoulder-elbow posture information includes a ratio trajectory, a deviation section, and a joint angle sequence, the fluctuation frequency information includes a trajectory segment sequence, a direction jump frequency, and a stability mark, the landing point analysis result is specifically a hit concentration range, a deviation direction trend, and a path extension structure, and the rhythm evaluation information specifically refers to a time distribution structure, a rhythm deviation section, and an operation sequence difference.

[0007] As a further solution of the present invention, the posture recognition module includes: The action image processing sub-module analyzes real-time action images, extracts the contour coordinate point information corresponding to the shoulder and elbow in the action image sequence of the trainer, constructs the action time sequence of the shoulder and elbow, locates and corresponds the key nodes in the continuous images, and generates a shoulder-elbow dynamic trajectory sequence; The joint angle extraction sub-module calls the shoulder-elbow dynamic trajectory sequence, calculates the direction angle value of the line connecting the two points according to the corresponding coordinate positions of the shoulder and elbow in each frame of image, and obtains a synchronous angle sequence composed of multiple angle points according to the time axis order, and generates a joint angle change path; The synchronous deviation judgment sub-module calls the joint angle change path, compares the change direction difference between the two-joint angle combination in the actual action and the set combination in the standard action, filters out the synchronous deviation period, and identifies the amplitude of the action deviation, and generates shoulder-elbow posture information.

[0008] As a further solution of the present invention, the trajectory monitoring module includes: The coordinate data acquisition sub-module calls the shoulder-elbow posture information to obtain the coordinate data of the laser point positions within a continuous time period projected by the trainer onto the target surface using the laser pointer during aiming, records the coordinate points corresponding to each time node, arranges them in chronological order to form a coordinate sequence, establishes a spatial movement path of consecutive points, and generates laser movement trajectory data; The trajectory fluctuation identification sub-module calls the laser movement trajectory data, filters out the fluctuating trajectory segments by calculating the spatial distance and the change angle of the orientation between adjacent points in the trajectory, and records the corresponding start time to generate a set of trajectory fluctuation segments; The action stability evaluation sub-module calls the set of trajectory fluctuation segments, identifies the intervals of each fluctuating trajectory segment in the time series, and evaluates the stability of the trainer's aiming action based on the duration and frequency of the fluctuating trajectory segments in multiple time periods to generate fluctuation frequency information.

[0009] As a further solution of the present invention, the specific formula for evaluating the stability of the trainer's aiming action is: ; Calculate the aiming action stability score; Wherein, represents the aiming action stability score, is the score adjustment coefficient, represents the normalization value of the duration of the th fluctuating trajectory segment, represents the normalization value of the spatial trajectory fluctuation distance of the th fluctuating trajectory segment, represents the fluctuation frequency correction factor, represents the total number of fluctuating trajectory segments, represents the number index of the trajectory fluctuation segment in the time series.

[0010] As a further solution of the present invention, the hit analysis module includes: The landing point data sorting sub-module calls the fluctuation frequency information, collects the coordinate information of the hit points actually formed by firing on the target surface and obtains the corresponding firing time sequence, sorts and connects the hit points according to the time sequence to form a hit path structure, and generates a hit connection path sequence; The offset trend extraction sub-module calls the hit connection path sequence, extracts the angle of the connection direction of each path segment, and identifies the offset trend path of the connection direction between the hit points by comparing the direction change trajectories of the hit points on the target surface during multiple shootings to generate an offset path recognition result; The direction aggregation evaluation sub-module calls the offset path recognition result, analyzes the extension continuity of the offset path, combines the spatial coverage range of the hit points on the target surface, evaluates the stability and consistency degree of the aggregation direction of the hit points, and combines the path offset direction to generate a landing point analysis result; The specific formula for combining the spatial coverage range of the hit points on the target surface is: ; Calculate the aggregation consistency index; Among them, represents the coordinate value of the th hit point in the horizontal direction of the target surface, represents the coordinate value of the th hit point in the vertical direction of the target surface, represents the average value of the horizontal coordinates of all hit points, represents the average value of the vertical coordinates of all hit points, is a fixed set value of the reference radius of the shooting hit area, is the normalized value of the hit point density in the sector area where the th hit point is located, is the normalized value with the largest density value in all sector areas during this training process, is the number of all hit points within the current task stage, is the aggregation consistency index, represents the jth hit point calculated currently.

[0011] As a further solution of the present invention, the rhythm comparison module includes: The task period recognition sub-module calls the landing point analysis result, collects the start time and end time records of each action unit in the training task, calculates the duration length of each action unit, and generates an action time structure sequence; The rhythm trend comparison sub-module calls the action time structure sequence, compares the start period of each action unit with the preset rhythm arrangement, identifies the early and late trends of the action unit, and generates rhythm offset trend data; The fluctuation degree calculation sub-module calls the rhythm offset trend data, analyzes the number of action paragraphs with continuous rhythm offsets and the duration range of each paragraph, calculates the proportion of the rhythm offset segments during the training process, evaluates the fluctuation degree of the personnel execution rhythm in the training task, and generates rhythm evaluation information.

[0012] As a further solution of the present invention, the system further includes: The label output module calls the rhythm evaluation information, adjusts the weight distribution of each evaluation index according to the ability priority set by the training task, analyzes the attribution correspondence relationship between each index and the label level interval corresponding to the task type, calculates the performance score of the trainer, and generates a training performance evaluation level; The training performance evaluation level specifically includes the label interval matching result, the composition of the ability score, and the level classification number.

[0013] As a further solution of the present invention, the label output module includes: The index weight configuration sub-module calls the rhythm evaluation information, extracts the corresponding ability priority parameters according to the training task type, and adjusts the weight of each evaluation index, including the action offset amplitude, the stability of the aiming action, the hit score, and the fluctuation degree of the execution rhythm, to generate an ability weight distribution coefficient; The level attribution matching sub-module calls the ability weight distribution coefficient, classifies and judges the label level interval corresponding to each evaluation result, identifies the attribution position of each index value within the level interval, and generates a level interval matching relationship group; The result label generation sub-module calls the level interval matching relationship group, combines the weight of each evaluation index, calculates the performance score of the trainer, identifies the performance level of each trainer, and generates a training performance evaluation level.

[0014] On the other hand, an intelligent shooting assessment and training method is provided. This method is applied to an intelligent shooting assessment and training system, and the method includes: S1: Analyze the real-time action image, extract the continuous image sequence during the process of the trainer holding the gun to firing, perform inter-frame node extraction and angle path construction on the shoulder and elbow actions, conduct a difference comparison on the actual angle ratio change based on the standard action trajectory, screen out the deviated paragraphs and extract the action offset amplitude, and generate shoulder-elbow posture information; S2: Based on the shoulder-elbow posture information, record the point coordinates of the laser pointer on the target surface during aiming, construct a coordinate time series path, measure the distance and direction change amplitude between adjacent coordinates, screen out the fluctuation trajectory segments, and evaluate the stability of the aiming action to generate fluctuation frequency information; S3: Based on the fluctuation frequency information, sort the multiple hit points formed by firing in chronological order to establish a hit point path structure, identify the offset path and analyze the continuity of the path extension, and combine the coverage range of the hit area to calculate the hit score and generate a landing point analysis result; S4: Based on the landing point analysis result, collect the start and end times of each action unit in the training task, combine the standard rhythm arrangement set in the training plan, identify the early and late trends of the action unit, evaluate the fluctuation degree of the execution rhythm, and generate rhythm evaluation information; S5: Based on the rhythm assessment information, adjust the weights of each evaluation index according to the set priority of capabilities configured for each training task type, identify the grade labels corresponding to each index, calculate the performance scores of the trainees, and establish a training performance evaluation grade.

[0015] The beneficial effects brought by the technical solutions provided in the embodiments of the present invention at least include: Based on the action images of the trainees from holding the gun to firing, the synchronous angle trajectories of the shoulders and elbows are recognized in real time, compared with the standard action combinations, the ratio differences in each time period are accurately screened and quantified, the dynamic evaluation of the action control stability is carried out by analyzing the laser trajectory coordinate points and direction changes during aiming, the structured scores are formed by using the order of the hitting points and combining the path extensibility and the density of the hitting area, the start and end times of the action units in the training process are compared one by one with the rhythm arrangement, the phenomena of advance and delay are identified, and the performance scores are flexibly adjusted according to the weights of multi-dimensional indexes, realizing the accurate matching between the training performance evaluation and the actual ability requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained according to these drawings.

[0017] Figure 1 is the system flow chart of the present invention; Figure 2 is the schematic diagram of the system framework of the present invention; Figure 3 is the schematic diagram of the method steps of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] The following describes the technical solutions in the present invention in conjunction with the drawings.

[0019] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design solution described as "example" in the present invention should not be construed as being more preferred or having more advantages than other embodiments or design solutions. Exactly speaking, the use of the word "example" is intended to present concepts in a specific way. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one of the two.

[0020] In the embodiments of the present invention, "image" and "picture" can sometimes be used interchangeably. It should be noted that when the difference is not emphasized, their intended meanings are the same. "Of", "corresponding", and "corresponding" can sometimes be used interchangeably. It should be noted that when the difference is not emphasized, their intended meanings are the same.

[0021] In the embodiments of the present invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, their intended meanings are the same.

[0022] To make the technical problems, technical solutions, and advantages to be solved by the present invention clearer, the following will be described in detail with reference to the accompanying drawings and specific embodiments.

[0023] The embodiments of the present invention provide an intelligent shooting assessment and training system. Please refer to Figures 1 to 2 , the present invention provides a technical solution. An intelligent shooting assessment and training system includes: The posture recognition module analyzes real-time action images, identifies the synchronous angular change trajectory of the shoulder and elbow based on the action images of the trainer during the process from holding the gun to firing, compares the corresponding angle combinations in the standard action, filters out the deviated segments, extracts the action deviation amplitude, and generates shoulder-elbow posture information; The trajectory monitoring module calls the shoulder-elbow posture information, obtains the continuous coordinate points of the laser pointer on the target surface during aiming, filters out the fluctuating trajectory segments by calculating the spatial distance and the change angle of the orientation between adjacent coordinate points, and evaluates the stability of the aiming action to generate fluctuation frequency information; The hit analysis module calls the fluctuation frequency information, sorts the hit points on the target surface according to the firing time, connects adjacent hit points and analyzes the direction change trend, identifies the deviation path, and calculates the hit score by combining the extension continuity of the path and the coverage range of the hit area to generate the landing point analysis result; The rhythm comparison module calls the landing point analysis result, analyzes the start time corresponding to each action unit in the training task, compares it with the preset rhythm arrangement, identifies the early and late trends of the action units, evaluates the fluctuation degree of the execution rhythm, and generates rhythm evaluation information; The label output module calls the rhythm evaluation information, adjusts the weight distribution of each evaluation index according to the ability priority set by the training task, analyzes the corresponding relationship between the attribution of each index and the label level interval corresponding to the task type, and calculates the performance score of the trainer to generate the training performance evaluation level.

[0024] Shoulder-elbow posture information includes the matching trajectory, offset section, and joint angle sequence. The fluctuation frequency information includes the trajectory segment sequence, direction jump frequency, and stability marker. The specific result of the landing point analysis is the hit concentration range, offset direction trend, and path extension structure. The rhythm evaluation information specifically refers to the time distribution structure, rhythm offset section, and operation sequence difference. The training performance evaluation level is specifically the label interval matching result, ability score composition, and grade classification number.

[0025] The posture recognition module includes: The action image processing sub-module analyzes the real-time action image, extracts the contour coordinate point information corresponding to the shoulder and elbow in the action image sequence of the training personnel, constructs the action time series of the shoulder and elbow, locates and corresponds the key nodes in the continuous images, and generates the shoulder-elbow dynamic trajectory sequence; During the analysis of the real-time action image, first, a high-speed camera device is used to collect the full-process images of the training personnel from holding the gun to firing at a rate of 60 frames per second. Taking a 3-second complete shooting action as an example, 180 consecutive images are generated, and the key feature points of the shoulder and elbow of the training personnel are identified one by one in each frame of the image. The two-dimensional coordinate positions of the upper vertex of the shoulder and the elbow joint point are extracted respectively, and the shoulder coordinates are denoted as point ., and the elbow coordinates are denoted as point . For example, in the 50th frame of the image, the shoulder coordinates =(210, 325) pixels, and the elbow coordinates =(240, 410) pixels. Subsequently, in the 51st frame of the image, the shoulder coordinates are updated to =(211, 327) pixels, and the elbow coordinates are updated to =(242, 412) pixels. In this way, the shoulder and elbow coordinate points of each frame of the image are determined in turn, and the shoulder coordinate sequence and the elbow coordinate sequence are constructed. Subsequently, by correlating the spatial coordinates and positioning the time sequence of the key nodes in each frame of the image, the spatial trajectory data of the shoulder and elbow changing with the action are finally obtained, and the shoulder-elbow dynamic trajectory sequence is generated.

[0026] The joint angle extraction sub-module calls the shoulder-elbow dynamic trajectory sequence, calculates the direction angle value of the line connecting the two points according to the corresponding coordinate positions of the shoulder and elbow in each frame of the image, and obtains a synchronous angle sequence composed of multiple angle points according to the time axis order, and generates the joint angle change path; For the shoulder-elbow dynamic trajectory sequence, the shoulder coordinates and the elbow coordinates are called in each frame of the image, and the direction angle is calculated using the coordinate connection line, that is, the azimuth angle calculation method in the rectangular coordinate system is adopted, and the calculation formula is: , where is the angle value in the direction of the shoulder-elbow connection line, , are the horizontal and vertical coordinates of the shoulder-elbow coordinate pair in the same frame. Taking the 50th frame image as an example, substituting the coordinates and into the above formula, we get: . Similarly, for the 51st frame, the direction angle obtained is . By repeating the above calculation steps for all 180 frame images, a continuous angle value sequence is obtained, and then sorted in chronological order to generate the joint angle change path.

[0027] The synchronous offset judgment sub-module calls the joint angle change path, compares the difference in the change direction between the combination of the two joint angles in the actual action and the set combination in the standard action, filters out the synchronous deviation period, and identifies the amplitude of the action offset to generate the shoulder-elbow posture information; Based on the joint angle change path, the reference value of the shoulder-elbow angle combination of the standard shooting action is called. This reference value is usually determined by statistical experiments of standard actions. Based on the measured data of the standard shooting posture, after 100 repeated action tests, the optimal range of the shoulder-elbow direction angle combination is determined to be (see Table 1).

[0028] Table 1 Standard Shoulder-Elbow Angle Combination Reference Table

[0029] The average value of the experimental data of the shoulder-elbow combined angle in Table 1 is 66.5°, and the standard deviation is about 0.8°. Then, the reasonable range of the standard shooting posture direction angle combination is set to 65° to 68°. Compare and judge each frame of the actual action angle with this range frame by frame: for example, if the measured shoulder-elbow direction angle of the 50th frame image is 70.56°, which is significantly beyond the above reasonable range, it is marked as a deviated frame; if the measured direction angle of the 52nd frame is 66.7°, which falls within the above reasonable range, it is marked as a normal frame. Perform such frame-by-frame comparison and judgment on all 180 frames in turn, mark and continuously count the consecutive frame segments that deviate from the normal range. Suppose the 50th to 55th frames and the 120th to 130th frames are marked as deviated paragraphs, and the total number of frames is 17. Then calculate the proportion of the deviated segments in the total number of frames as 17 / 180≈0.094, that is, the proportion of the deviated paragraphs is about 9.4%. Further, calculate the difference between each frame's actual angle in the marked frame segments and the boundary values of the standard range. For example, the difference of the 50th frame is 70.56° - 68° = 2.56°, and the difference of the 120th frame is similarly processed to be 2.03°. Obtain the difference data of each deviated frame in turn. After analyzing all the difference data, select the maximum offset amplitude of 2.56° as the maximum action offset value, and calculate the average offset amplitude as (2.56° + 2.03° + …) / 17. Suppose the result is 2.1°. Finally, comprehensively calculate to obtain the shoulder-elbow posture information.

[0030] The trajectory monitoring module includes: The coordinate data acquisition sub-module calls the shoulder-elbow posture information, acquires the coordinate data of the laser points on the target surface projected by the laser pointer used by the training personnel during the aiming period within a continuous time period, records the coordinate points corresponding to each time node, arranges them in chronological order to form a coordinate sequence, establishes a continuous point space movement path, and generates laser movement trajectory data; During the coordinate data acquisition process, while the training personnel are performing the shooting aiming action, the laser pointer continuously projects to a fixed position on the front target surface. The coordinate capture device is used to collect the two-dimensional coordinates of the laser points in real time. The acquisition device acquires the laser coordinates on the target surface at a frequency of 50 times per second. If it continuously captures for 2 seconds, a total of 100 coordinate points are obtained. For example, the coordinate position of the first acquisition point is , the second acquisition point is , and so on. Suppose the coordinate position of the 100th acquisition point is . All the acquired coordinate points form a coordinate sequence . Each coordinate point is paired with the corresponding acquisition time node according to the acquisition order. For example, the time corresponding to the first coordinate is , the second is , until the 100th is . Subsequently, connect them in turn according to the acquisition order of the coordinate points. For example, connect the first point with the second point , then connect the second point to the third point until the 99th point is connected to the 100th point, forming a continuous point position space movement path to obtain laser movement trajectory data.

[0031] The trajectory fluctuation recognition sub-module calls the laser movement trajectory data, screens the fluctuation trajectory segments by calculating the spatial distance and the orientation change angle between adjacent point positions in the trajectory, and records the corresponding start time to generate a set of trajectory fluctuation segments; During the trajectory fluctuation recognition process, the continuous point position coordinate sequence of the above-mentioned laser movement trajectory data is called, and the spatial distance and the orientation change angle of each pair of adjacent coordinate points are calculated respectively. Taking the first point and the second point as an example, the spatial distance between the two points Calculated using the Euclidean distance formula as: , and the direction change angle Calculated as the angle between the line connecting the two point coordinates and the horizontal axis, that is: , similarly calculate all continuous point pairs such as the second point and the third point, the third point and the fourth point, etc. If the distance between two adjacent point pairs exceeds the set distance reference value of 2.0 mm, or the direction change angle exceeds the set angle reference value of 30°, it is marked as a trajectory fluctuation point segment. For example, the 20th to 24th point pairs meet the above distance or angle exceeding the benchmark, which is recorded as a fluctuation trajectory segment, and the start time node is recorded as the time corresponding to the 20th point , and the end time is the time corresponding to the 24th point , assuming that two more fluctuation trajectory segments of the 45th to 50th point pairs and the 80th to 85th point pairs are obtained in a similar marking method, and after summarization, they are recorded as a set of trajectory fluctuation segments. The example data is shown in Table 2.

[0032] Table 2 Example data table of laser trajectory fluctuation segments

[0033] Table 2 gives the time information of the fluctuation trajectory segments collected in the embodiment. After the trajectory fluctuation recognition sub-module calls the data, a set of trajectory fluctuation segments is formed.

[0034] The action stability evaluation sub-module calls the set of trajectory fluctuation segments, identifies the intervals of each fluctuation trajectory segment in the time series, and evaluates the stability of the training personnel's aiming action according to the duration and frequency of the fluctuation trajectory segments in multiple time periods to generate fluctuation frequency information; The specific formula for evaluating the stability of the training personnel's aiming action is: ; Calculate the aiming action stability score; Among them, represents the aiming action stability score, is the score adjustment coefficient, Represents the normalized value of the duration of the th fluctuation trajectory segment, Represents the normalized value of the spatial trajectory fluctuation distance of the th fluctuation trajectory segment, Represents the fluctuation frequency correction factor, Represents the total number of fluctuation trajectory segments, Represents the serial number index of the trajectory fluctuation segment in the time series.

[0035] Formula: ; Detailed explanation of the formula and the derivation process of the formula calculation: The formula is used to calculate the action stability score of the trainer during the aiming phase. The score result is used to reflect the level of continuous smoothness of the action during the training process. The higher the value, the more stable the aiming; Meaning of parameters and set values: Is the normalized value of the duration of the th fluctuation trajectory segment. The original duration is obtained by counting the change time period of consecutive laser pointing coordinates in the time series. This value is normalized by dividing by the total training duration If the durations of the 1st, 2nd, and 3rd fluctuation trajectory segments are 0.6 seconds, 0.9 seconds, and 1.2 seconds respectively, and the total training duration is 10 seconds, then there are: , , ; Is the normalized value of the spatial fluctuation distance of the th fluctuation trajectory segment. This value is the approximate arc length of the curve formed by the continuous change coordinate points of the laser spot, and is obtained by dividing the total distance of each segment of the trajectory by the standard reference distance If the fluctuation distances of the 1st, 2nd, and 3rd segments are 4.5 cm, 5.1 cm, and 5.7 cm respectively, and the standard reference distance is 10 cm, then there are: , , ; Is the frequency correction factor, which represents the normalized result of the difference between the number of trajectory fluctuations per unit time and the set threshold. The reasonable range of the fluctuation frequency is set from 0 to 6 times per 10 seconds. If the actual acquisition frequency is 8 times, then the correction factor is: ; Is the score adjustment coefficient, which is determined according to the fitting analysis between the evaluation accuracy of multiple samples and the fluctuation characteristics, and is usually set to 15; Is the number of trajectory fluctuation segments, which is 3 segments in the current sample; Substitute the parameters into the formula for calculation: ; The result of 94.34 indicates that the aiming action has a relatively high stability score in this training sample. The value is close to the full score, indicating that the trajectory fluctuation is small, the duration is short, and the frequency does not far exceed the set reference range, so the action stability is strong. This value is used as a key indicator in the subsequent comprehensive evaluation results to participate in the calculation of the final training level output.

[0036] The hit analysis module includes: The landing point data sorting sub-module calls the fluctuation frequency information, collects the coordinate information of the hit points actually formed by firing on the target surface and obtains the corresponding firing time sequence, sorts and connects the hit points according to the time sequence to form a hit path structure, and generates a hit connection path sequence; During the landing point data sorting process, the fluctuation frequency information corresponding to the shooting stage generated by the previous module is called, and the spatial coordinates of the impact points formed by each firing on the target surface are recorded in real time by the intelligent target device. Specifically, two-dimensional coordinate positioning is carried out in millimeters (mm). Assuming that a trainee conducts a shooting training and fires 5 times in total, the intelligent target obtains 5 hit point coordinates and the corresponding firing time nodes, as shown in Table 3.

[0037] Table 3 Record Table of Hit Point Coordinates and Firing Time

[0038] Based on the data in Table 3, the firing time is called to sort the 5 hit points. The sorting result is: the first firing point (85, 102), the second (88, 100), the third (91, 97), the fourth (95, 94), the fifth (100, 90), and the hit points are successively connected in this order. Specifically, a straight-line connection method is adopted, that is, the first point is connected to the second point, the second point is connected to the third point, and so on, finally forming a continuous hit connection path sequence.

[0039] The offset trend extraction sub-module calls the hit connection path sequence, extracts the angle of the connection direction of each path segment, and identifies the offset trend path of the connection direction between the hit points by comparing the direction change trajectories of the hit points in multiple shootings on the target surface, and generates an offset path recognition result; During the offset trend extraction process, the hit connection path sequence is called, and the angle of the connection direction of adjacent connection paths is extracted, and the azimuth angle of the connection line is calculated through the path coordinate difference. Taking the path from the first point (85, 102) to the second point (88, 100) as an example, the direction angle calculation formula of the connection path is , and substituting the specific data gives: , and similarly calculate the direction angles of the subsequent paths, such as the direction angle of the connection path from the second point to the third point The direction angle from the 3rd point to the 4th point The path angle from the 4th point to the 5th point to obtain a complete sequence of direction angles: . Compare with the direction change angle of the standard hit trajectory (for example, with the horizontal right 0° as the reference, and the standard deviation range allowing a float within ±10°), and judge whether each angle in the angle sequence exceeds the standard deviation range one by one. If the angle of the 1st path -33.69° exceeds the standard range of ±10° (-10° to 10°), it is marked as an offset direction path, and it is marked one by one in this way. Finally, the offset trend path of the connection direction of the hit points is identified, and the offset path recognition result is generated.

[0040] The direction aggregation evaluation sub-module calls the offset path recognition result, analyzes the extension continuity of the offset path, combines the spatial coverage range of the hit points on the target surface, evaluates the stability and consistency degree of the aggregation direction of the hit points, and combines the path offset direction to generate a landing point analysis result; The specific formula for combining the spatial coverage range of the hit points on the target surface is: ; Calculate the aggregation consistency index; Among them, represents the coordinate value of the th hit point in the horizontal direction of the target surface, represents the coordinate value of the th hit point in the vertical direction of the target surface, is a fixed set value of the reference radius of the shooting hit area, is the normalized value of the hit point density in the sector area where the th hit point is located, is the maximum normalized value of the density values in all sector areas during this training process, is the number of all hit points within the current task stage, is the aggregation consistency index,

[0041] Formula: ; Detailed explanation of the formula and the derivation process of the formula calculation: The formula calculates the hit point aggregation consistency index, which is used to quantify the aggregation degree and stability of the hit points in shooting training, and evaluate the stability of the aiming action of the training personnel; Meaning and setting value of parameters: and : They are the horizontal and vertical coordinates of the th hit point on the target surface, with the unit of meter. There are 5 hit points set, and the horizontal coordinates are , , , , , and the vertical coordinates are , , , , ; and : They are respectively the average values of the horizontal and vertical coordinates of all hit points. The average value of the horizontal coordinates is , and the average value of the vertical coordinates is ; : It is the reference radius of the shooting hit area, with the unit of meter, and is used to standardize the offset of the hit points. Assume it is set to meters; and : They are respectively the density normalization value of the area where the th hit point is located and the density value of the densest area, with the unit of dimensionless value. Assume , , , , , and the densest area is ; : It is the total number of hit points during the training phase. Assume it is .

[0042] Substitute the parameters into the formula for calculation: ; ; ; ; ; ; ; ; This result indicates that the aggregation consistency of the hit points is low, which means that the hit points are more dispersed on the target surface and the shooting stability is poor.

[0043] The rhythm comparison module includes: The task period recognition sub-module calls the landing point analysis results, collects the start time and end time records of each action unit in the training task, calculates the duration length of each action unit, and generates an action time structure sequence; During the task period recognition process, the training task phase corresponding to the landing point analysis results is called, and the start and end moments of each action unit in the shooting training process are collected in real time through the task data recording device. Taking a complete shooting training task as an example, assuming that the task process includes 4 action units: preparation action, aiming action, firing action, and recovery action, the data recording device collects the start and end time points of each unit and records them in seconds as and , as shown in Table 4.

[0044] Table 4 Action Unit Time Data Table for Shooting Training

[0045] According to the data in Table 4, calculate the duration length of each action unit one by one. For example, the duration of the preparation action is calculated as , the duration of the aiming action is calculated as , the duration of the firing action is , and the duration of the recovery action is . Record the duration of each action unit in the actual execution order to generate an action time structure sequence.

[0046] The rhythm trend comparison sub-module calls the action time structure sequence, compares the start period of each action unit with the preset rhythm arrangement, identifies the early and late trends of the action unit, and generates rhythm offset trend data; Call the above action time structure sequence and compare it with the time benchmark of the standard shooting training action rhythm arrangement. The standard rhythm arrangement is obtained through a large number of experiments. Assume that the data of the standard action unit rhythm arrangement is shown in Table 5.

[0047] Table 5 Standard Action Rhythm Arrangement Data Table

[0048] Compare the start time of each actual action unit with the standard start time item by item. For example, the actual start time of the preparation action is 0.00s, the standard is 0.00s, the time difference is 0.00s, and there is no advance or delay; the actual start time of the aiming action is 1.00s, the standard is 1.20s, and the difference is -0.20s, indicating that the action is advanced; the actual start time of the firing action is 3.00s, the standard is 3.20s, and the difference is -0.20s, 0.20s in advance; the actual start time of the recovery action is 3.50s, the standard is 3.70s, and the difference is -0.20s, also showing an advancing trend. According to this process, identify the advance or delay trend of each action unit relative to the standard rhythm arrangement one by one, record the rhythm offset trend of the above-mentioned action units in sequence, and generate rhythm offset trend data.

[0049] The fluctuation degree calculation sub-module calls the rhythm offset trend data, analyzes the number of action paragraphs with continuous rhythm offsets and the duration range of each paragraph, calculates the proportion of the rhythm offset segments during the training process, evaluates the fluctuation degree of the personnel's execution rhythm in the training task, and generates rhythm evaluation information; Call the rhythm offset trend data and conduct statistical analysis on the identified continuous rhythm offset paragraphs. First, record the offset trend types of each action unit. Taking the data in Table 4 as an example, action units 2, 3, and 4 all show continuous advance offset trends, and action unit 1 has no offset trend. The number of continuously offset action units is 3. Subsequently, calculate the total duration of the continuous offset paragraphs respectively. For example, the sum of the durations of action units 2, 3, and 4 is , the total duration of the training task is 4.50s, and the offset paragraph proportion is calculated as , that is, the rhythm offset segment proportion is 77.8%. Further, compare the standard allowable rhythm offset proportion threshold of 30%. The actually calculated offset segment proportion of 77.8% significantly exceeds the standard threshold. Based on a hundred-point system, set the rhythm fluctuation evaluation coefficient. For every 10% exceeding the standard allowable offset proportion, 10 points will be deducted from the score. Calculate the specific deduction as points. Therefore, the rhythm stability score is points. This score indicates that there are significant fluctuations in the personnel's execution rhythm during the training task, and the final evaluation calculation result is used as the rhythm evaluation information.

[0050] The label output module includes: The index weight configuration sub-module calls the rhythm evaluation information, extracts the corresponding ability priority parameters according to the training task type, and adjusts the weight of each evaluation index, including the action offset amplitude, the stability of the aiming action, the hit score, and the fluctuation degree of the execution rhythm, to generate the ability weight distribution coefficient; Call the rhythm evaluation information, configure the ability priorities according to the actual training task type, and specifically classify the task types into three categories: precision - priority tasks, stability - priority tasks, and rhythm - control - priority tasks. Taking the precision - priority tasks as an example, determine the priorities of different ability indicators. For the four evaluation indicators of action deviation amplitude, aiming action stability, hit score, and execution rhythm fluctuation degree, obtain the priority parameters according to the actual scenario statistics. Among them, the hit score is determined as the first priority, the action deviation amplitude is the second priority, the aiming stability is the third priority, and the execution rhythm fluctuation is the fourth priority. Set specific weight ratio parameters for different priorities. For example, the weight configured for the highest - priority hit score is 0.4, the secondary weight for the action deviation amplitude is 0.3, the aiming stability is 0.2, and the execution rhythm fluctuation is 0.1. Judge the training task type. When the trainer executes the precision - priority task, directly call the above parameters to adjust the weights of each evaluation indicator. Specifically, use the weight - adjustment formula to calculate the comprehensive weight distribution coefficient of the indicators: In the actual evaluation, the hit score is 0.4, the action deviation amplitude is 0.3, the aiming stability is 0.2, and the rhythm fluctuation is 0.1 as the ability weight distribution coefficient.

[0051] The level - attribution matching sub - module calls the ability weight distribution coefficient, classifies and judges the label level intervals corresponding to each evaluation result, identifies the attribution position of each index value within the level interval, and generates a level - interval matching relationship group; Call the above-mentioned ability weight distribution coefficients, and classify and judge the values obtained for each specific evaluation index one by one into grade intervals. Taking the hit score, action deviation amplitude, aiming stability, and execution rhythm fluctuation as specific indicators, the specific division criteria for each grade interval are obtained based on the comprehensive performance evaluation experiment of the shooting task. For example, according to the experimental setting, a hit score of 85 or above is judged as "excellent", 75 to 85 is "good", 60 to 75 is "medium", and below 60 is "poor". If the actual score is 92 points, it is directly classified as "excellent"; if the action deviation amplitude is judged by the angular error, the deviation amplitude in the range of 0 to 1.5 degrees is defined as "excellent", 1.5 to 3 degrees is "good", 3 to 5 degrees is "medium", and above 5 degrees is "poor". If the actually measured deviation amplitude is 2.1 degrees, it is judged as "good"; aiming stability is judged by the fluctuation score, a score of 85 or above is "excellent", 70 to 85 is "good", 55 to 70 is "medium", and below 55 is "poor". If the actual score is 69.5 points, the judged grade is "medium"; the degree of execution rhythm fluctuation is judged by the deviation section ratio, 0 to 20% is "excellent", 20 to 40% is "good", 40 to 60% is "medium", and above 60% is "poor". If the actual fluctuation ratio is 77.8%, the judged grade is "poor". Through the above grade classification operation, each index obtains the corresponding grade label position, that is, the hit score is "excellent", the action deviation amplitude is "good", the aiming stability is "medium", and the execution rhythm fluctuation is "poor", which is integrated into a grade interval matching relationship group.

[0052] The result label generation sub-module calls the grade interval matching relationship group, combines the weights of each evaluation index, calculates the performance scores of the training personnel, identifies the performance levels of each training personnel, and generates the training performance evaluation levels; Call the grade interval matching relationship group, take the ability weights of each evaluation index as parameters, and calculate the overall performance scores of the training personnel. The basic scores corresponding to the grades in the calculation process are: 95 points for excellent, 85 points for good, 70 points for medium, and 55 points for poor. Substitute the aforementioned grade matching results and ability weight distribution coefficients to calculate the total score item by item: the basic score of 95 points for the hit score of "excellent" is multiplied by the corresponding weight of 0.4 to calculate 38 points, the basic score of 85 points for the action deviation amplitude of "good" is multiplied by the weight of 0.3 to calculate 25.5 points, the basic score of 70 points for the aiming stability of "medium" is multiplied by the weight of 0.2 to calculate 14 points, and the basic score of 55 points for the execution rhythm fluctuation of "poor" is multiplied by the weight of 0.1 to calculate 5.5 points. The cumulative score is 38 + 25.5 + 14 + 5.5 = 83 points. Further match the score with the performance level standard, and define the overall score of 85 points or above as "excellent" in overall performance, 75 to 85 as "good", 60 to 75 as "general", and below 60 as "poor". The score obtained in this calculation is 83 points, and the overall performance level attribution is judged as "good", which is used as the training performance evaluation level.

[0053] Please refer to Figure 3 , on the other hand, a smart shooting assessment and training method is provided. This method is applied to a smart shooting assessment and training system, and the method includes: S1: Analyze the real-time action image, extract the continuous image sequence during the process of the trainer holding the gun to firing, perform inter-frame node extraction and angle path construction on the shoulder and elbow movements, conduct differential comparison on the actual angle ratio change according to the standard action trajectory, screen out the deviated paragraphs and extract the action deviation amplitude, and generate shoulder-elbow posture information; S2: Based on the shoulder-elbow posture information, record the point coordinates of the laser pointer on the target surface during aiming, construct a coordinate time series path, measure the distance and direction change amplitude between adjacent coordinates, screen out the fluctuation trajectory segments, and evaluate the stability of the aiming action, and generate fluctuation frequency information; S3: Based on the fluctuation frequency information, sort the multiple hit points formed by firing in chronological order to establish a hit point path structure, identify the deviation path and analyze the continuity of the path extension, and calculate the hit score in combination with the coverage range of the hit area, and generate a landing point analysis result; S4: Based on the landing point analysis result, collect the start and end times of each action unit in the training task, and in combination with the standard rhythm arrangement set in the training plan, identify the early and late trends of the action units, evaluate the fluctuation degree of the execution rhythm, and generate rhythm evaluation information; S5: Based on the rhythm evaluation information, adjust the weight of each evaluation index according to the ability priority setting configured for each training task type, identify the grade label corresponding to each index, and calculate the performance score of the trainer to establish a training performance evaluation grade.

[0054] The above embodiments can be implemented in whole or in part by software, hardware (such as circuits), firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present invention are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wired (such as infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that contains one or more collections of available media. The available media can be magnetic media (such as floppy disks, hard disks, magnetic tapes), optical media (such as DVDs), or semiconductor media. The semiconductor media can be a solid-state drive.

[0055] It should be understood that the term "and / or" in this document is merely a description of the association relationship between associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. Here, A and B can be singular or plural. In addition, the character " / " in this document generally represents an "or" relationship between the associated objects before and after, but it may also represent an "and / or" relationship, which can be specifically understood by referring to the context before and after.

[0056] In the present invention, "at least one" means one or more, and "a plurality" means two or more. "At least one of the following" or its similar expressions refer to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b, or c can represent: a, b, c, a - b, a - c, b - c, or a - b - c, where a, b, and c can be single or multiple.

[0057] It should be understood that in various embodiments of the present invention, the magnitudes of the sequence numbers of the above processes do not mean the order of execution. The order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0058] Those of ordinary skill in the art can realize that the units and algorithm steps of the examples described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. A professional technician can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0059] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the devices, apparatuses, and units described above can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated herein.

[0060] In several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0061] The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they can be located in one place, or can be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0062] In addition, the functional units in various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit.

[0063] When the above-mentioned function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this 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 for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0064] As described above, the above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. An intelligent shooting assessment and training system, characterized in that, The system includes: The posture recognition module analyzes real-time action images, recognizes the synchronous angular change trajectories of the shoulder and elbow according to the action images of the trainer during the process from holding the gun to firing, compares the corresponding angle combinations in the standard action, filters out the deviated paragraphs and extracts the action deviation amplitude, and generates shoulder-elbow posture information; The trajectory monitoring module calls the shoulder-elbow posture information, obtains the continuous coordinate points of the laser pointer on the target surface during aiming, filters out the fluctuating trajectory segments by calculating the spatial distance and the changing angle of the orientation between adjacent coordinate points, and evaluates the stability of the aiming action, and generates fluctuation frequency information; The hit analysis module calls the fluctuation frequency information, sorts the hit points on the target surface according to the firing time, connects adjacent hit points and analyzes the direction change trend, identifies the deviation path, combines the extension continuity of the path and the coverage range of the hit area, calculates the hit score, and generates the landing point analysis result; The rhythm comparison module calls the landing point analysis result, analyzes the start time corresponding to each action unit in the training task, and by comparing with the preset rhythm arrangement, identifies the early and late trends of the action units, evaluates the fluctuation degree of the execution rhythm, and generates rhythm evaluation information.

2. The intelligent shooting assessment and training system according to claim 1, wherein The shoulder-elbow posture information includes the ratio trajectory, the offset section, and the joint angle sequence. The fluctuation frequency information includes the trajectory segment sequence, the direction jump frequency, and the stability mark. The landing point analysis result is specifically the hit concentration range, the offset direction trend, and the path extension structure. The rhythm evaluation information specifically refers to the time distribution structure, the rhythm offset paragraph, and the operation sequence difference.

3. The intelligent shooting assessment and training system according to claim 1, characterized in that The posture recognition module includes: The action image processing sub-module analyzes real-time action images, extracts the contour coordinate point information corresponding to the shoulder and elbow in the action image sequence of the trainer, constructs the action time sequence of the shoulder and elbow, locates and corresponds the key nodes in the continuous images, and generates the shoulder-elbow dynamic trajectory sequence; The joint angle extraction sub-module calls the shoulder-elbow dynamic trajectory sequence, calculates the direction angle value of the line connecting the two points according to the corresponding coordinate positions of the shoulder and elbow in each frame of image, and obtains a synchronous angle sequence composed of multiple angle points according to the time axis order, and generates the joint angle change path; The synchronous offset judgment sub-module calls the joint angle change path, compares the change direction difference between the two-joint angle combination in the actual action and the set combination in the standard action, filters out the synchronous deviation period, and identifies the amplitude of the action deviation, and generates shoulder-elbow posture information.

4. The intelligent shooting assessment and training system according to claim 3, characterized in that The trajectory monitoring module includes: The coordinate data acquisition sub-module calls the shoulder-elbow posture information, obtains the coordinate data of the laser point positions during the continuous time period when the trainer projects the laser pointer onto the target surface during aiming, records the coordinate points corresponding to each time node, arranges them in time order to form a coordinate sequence, establishes a spatial movement path of continuous points, and generates laser movement trajectory data; The trajectory fluctuation identification sub-module calls the laser movement trajectory data, filters out the fluctuating trajectory segments by calculating the spatial distance and the changing angle of the orientation between adjacent points in the trajectory, and records the corresponding start time, and generates a set of trajectory fluctuation segments; The action stability evaluation sub-module calls the set of trajectory fluctuation segments, identifies the intervals of each fluctuating trajectory segment in the time series, evaluates the stability of the trainer's aiming action based on the duration and frequency of the fluctuating trajectory segments in multiple time periods, and generates fluctuation frequency information.

5. The intelligent shooting assessment and training system according to claim 4, wherein The specific formula for evaluating the stability of the trainer's aiming action is: ； Calculate the aiming action stability score; Among them, represents the aiming motion stability score, is the score adjustment coefficient, represents the normalization value of the duration of the th fluctuation trajectory segment, represents the normalization value of the spatial trajectory fluctuation distance of the th fluctuation trajectory segment, represents the fluctuation frequency correction factor, represents the total number of fluctuation trajectory segments, represents the serial number index of the trajectory fluctuation segment in the time series.

6. The intelligent shooting assessment and training system according to claim 4, characterized in that The hit analysis module includes: The landing point data sorting sub-module calls the fluctuation frequency information, collects the coordinate information of the hit points actually formed by the firing on the target surface and obtains the corresponding firing time sequence, sorts and connects the hit points according to the time sequence to form a hit path structure, and generates a hit connection path sequence; The offset trend extraction sub-module calls the hit connection path sequence, extracts the angle of the connection direction of each path segment, and identifies the offset trend path of the connection direction between the hit points by comparing the direction change trajectories of the hit points in multiple shootings on the target surface, and generates an offset path recognition result; The direction aggregation evaluation sub-module calls the offset path recognition result, analyzes the extension continuity of the offset path, combines the spatial coverage range of the hit points on the target surface, evaluates the stability and consistency degree of the aggregation direction of the hit points, and combines the path offset direction to generate a landing point analysis result; The specific formula for combining the spatial coverage range of the hit points on the target surface is: ； Calculate the aggregation consistency index; Among them, represents the coordinate value of the th hit point in the horizontal direction of the target surface, represents the coordinate value of the th hit point in the vertical direction of the target surface, represents the average value of the horizontal coordinates of all hit points, represents the average value of the vertical coordinates of all hit points, is a fixed set value of the reference radius of the shooting hit area, is the normalized value of the hit point density in the sector area where the th hit point is located, is the maximum normalized value of the density in all sector areas during this training process, is the number of all hit points within the current task phase, is the clustering consistency index, represents the jth hit point calculated currently.

7. The intelligent shooting assessment and training system according to claim 6, characterized in that, The rhythm comparison module includes: The task period recognition sub-module calls the landing point analysis result, collects the start time and end time records of each action unit in the training task, calculates the duration length of each action unit, and generates an action time structure sequence; The rhythm trend comparison sub-module calls the action time structure sequence, compares the start period of each action unit with the preset rhythm arrangement, identifies the early and late trends of the action unit, and generates rhythm offset trend data; The fluctuation degree calculation sub-module calls the rhythm offset trend data, analyzes the number of action paragraphs with continuous rhythm offsets and the duration range of each paragraph, calculates the proportion of the rhythm offset segments in the training process, evaluates the fluctuation degree of the execution rhythm of the personnel in the training task, and generates rhythm evaluation information.

8. The intelligent shooting assessment and training system according to claim 1, wherein, The system further includes: The label output module calls the rhythm evaluation information, adjusts the weight distribution of each evaluation index according to the ability priority set by the training task, analyzes the belonging correspondence relationship between each index and the label level interval corresponding to the task type, and calculates the performance score of the trainer to generate a training performance evaluation level; The training performance evaluation level is specifically the label interval matching result, the composition of the ability score, and the level classification number.

9. The intelligent shooting assessment and training system according to claim 8, wherein The label output module includes: The index weight configuration sub-module calls the rhythm evaluation information, extracts the corresponding ability priority parameters according to the training task type, and adjusts the weight of each evaluation index, including the action offset amplitude, the stability of the aiming action, the hit score, and the fluctuation degree of the execution rhythm, to generate an ability weight distribution coefficient; The level attribution matching sub-module calls the ability weight distribution coefficient to classify and judge the label level intervals corresponding to each evaluation result, identify the attribution positions of each index value within the level intervals, and generate a group of level interval matching relationships; The result label generation sub-module calls the group of level interval matching relationships, combines the weights of each evaluation index, calculates the performance scores of the training personnel, identifies the performance levels of each training personnel, and generates training performance evaluation levels.

10. An intelligent shooting assessment and training method, characterized in that, The method is used to implement the intelligent shooting assessment and training system described in any one of claims 1-9, and the method includes: S1: Analyze the real-time action image, extract the continuous image sequence during the process of the training personnel holding the gun to firing, extract the frame-by-frame nodes and construct the angle path for the shoulder and elbow movements, compare the difference in the actual angle ratio change with the standard action trajectory, screen the deviated paragraphs and extract the action deviation amplitude to generate shoulder-elbow posture information; S2: Based on the shoulder-elbow posture information, record the point coordinates of the laser pointer on the target surface during aiming, construct a coordinate time series path, measure the distance and direction change amplitude between adjacent coordinates, screen the fluctuating trajectory segments, and evaluate the stability of the aiming action to generate fluctuation frequency information; S3: Based on the fluctuation frequency information, sort the multiple hit points formed by firing in chronological order to establish a hit point path structure, identify the deviation path and analyze the extension continuity of the path, and combine the coverage range of the hit area to calculate the hit score and generate a landing point analysis result; S4: Based on the landing point analysis result, collect the start and end times of each action unit in the training task, combine the standard rhythm arrangement set in the training plan, identify the early and late trends of the action units, and evaluate the fluctuation degree of the execution rhythm to generate rhythm evaluation information; S5: Based on the rhythm evaluation information, adjust the weights of each evaluation index according to the ability priority settings configured for each training task type, identify the level labels corresponding to each index, calculate the performance scores of the training personnel, and establish training performance evaluation levels.

Images