Intelligent detection method for quick shooting action

Abstract

The application discloses a kind of intelligent detection methods of quick shooting action, comprising: using video camera real-time acquisition training personnel video image;Using human posture estimation model extracts the joint of human region, obtains joint information;Extract arm and spine information, obtain its angle information;Identify shooting preparation action and shooting aiming action and record time;Using shooting firing detection module identifies live ammunition shooting firing signal and records time;Calculate the time length used in each stage in the process of shooter shooting.The application can realize the quick intelligent detection of the shooting action of non-contact training personnel, and the recognition using the joint angle information of each human body can realize the detection of training personnel of different gender, age and body state. When there are many people in the video, only the joints of the nearest human body to the camera are selected for recognition, so as to avoid the influence on recognition in the case of many people, improve the detection speed and ensure the accuracy of recognition.

Description

Technical Field

[0001] This invention relates to an intelligent detection method for rapid shooting actions. Background Technology

[0002] In current shooting training, rapid shooting drills need to be practiced and assessed. The time from shooting preparation to aiming and firing is an important indicator for measuring the shooting level of trainees. Currently, this is achieved manually, using a stopwatch to calculate the duration of each stage of the shooting process. This method is cumbersome, time-consuming, and inaccurate. Therefore, there is an urgent need for a rapid and automatic method to detect shooting actions to meet training requirements. Summary of the Invention

[0003] Purpose of the invention: To address the shortcomings of existing technologies, this invention provides an intelligent detection method for rapid shooting actions, which can identify shooting actions and calculate shooting time in real time, providing strong support for shooting training and evaluation.

[0004] Beneficial effects: This invention enables rapid and intelligent detection of shooting actions of trainees without contact. By using the angle information of various joints of the human body for identification, it can detect trainees of different genders, ages, and body types. When there are multiple people in the video, only the joints of the human body closest to the camera are selected for identification, thereby avoiding the influence of multiple people on the identification, improving the detection speed, and ensuring the accuracy of the identification. Attached Figure Description

[0005] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments, and the advantages of the present invention as described above or otherwise will become clearer.

[0006] Figure 1 This is a diagram of the architecture of the present invention.

[0007] Figure 2 This is a flowchart of the present invention. Detailed Implementation

[0008] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0009] Step 1: Collect video image data of the trainee's shooting process in real time using a camera;

[0010] Step 2: Extract the joint point positions of the trainee in the video images of the trainee's shooting process and obtain joint point information. Specifically, this includes: reducing the frame rate of the video and processing one frame after another; using the OpenPose human pose estimation model algorithm library to extract the joint point positions of the trainee in the video images of the trainee's shooting process and obtain joint point information.

[0011] Step 3: Filter the key point information obtained in Step 2 and extract the key point information of the person closest to the camera; calculate the distance of the person from the perpendicular line of the shooting direction of the image based on the key point information of different people identified, and the person with the smallest distance is the person to be identified for training; the key point information includes key point coordinate information;

[0012] Step 4: Classify the key point information extracted in Step 3 to obtain a category sequence.

[0013] Step 5: Extract the joint information of the arm and spine, and calculate the angle information;

[0014] Step 6: Based on the information of the arm and spinal joints, identify the shooting preparation movements and record the time of occurrence of the movements;

[0015] Step 7: Based on the information of the arm and spinal joints, identify the shooting aiming action and record the time of the action;

[0016] Step 8: Obtain the firing signal from the trainee, obtain the firing time, and combine the time of the firing preparation action and the firing aiming action to calculate the time taken for the firing process, obtaining three times: the time taken from the firing preparation action to the firing action T1, the time taken from the firing aiming action to the firing action T2, and the time taken from the firing preparation action to the firing aiming action T3.

[0017] Step 3 includes: adjusting the camera view to be parallel to the shooting direction, then the perpendicular vector is v(0,W), where W is the image resolution width. The formula for calculating the average distance D between the personnel joint points and the perpendicular line is as follows:

[0018]

[0019] Where n is the total number of joints, x i Let D be the x-coordinate of the i-th joint; select the joint information of the trainee with the smallest D for processing.

[0020] In step 4, the joint information includes location, confidence level, and category p(x,y,c,f), where x is the horizontal coordinate, y is the vertical coordinate, c is the joint category, and f is the confidence level. The value of f ranges from 0 to 1, and the larger the value, the higher the confidence level. The condition f>0.5 is set to filter the joint information.

[0021] Step 5 includes:

[0022] The coordinates of the joints at both ends of the upper arm on the image are p(x1,y1), p(x2,y2), the coordinates of the joints at both ends of the forearm on the image are p(x2,y2), p(x3,y3), and the coordinates of the joints at both ends of the spine on the image are p(x5,y5), p(x6,y6). The angle formula is as follows:

[0023]

[0024]

[0025]

[0026] θ1 represents the angle between the upper arm and the forearm, θ2 represents the angle between the upper arm and the spine, and θ3 represents the angle between the forearm and the spine.

[0027] Step 6 includes: identifying firing preparation actions using the following formula:

[0028]

[0029] Where M1 represents the recognition condition of the shooting preparation action, θ1 represents the angle between the upper arm and the forearm, θ2 represents the angle between the upper arm and the spine, and α and δ are recognition adjustment parameters. When both conditions are met simultaneously -α≤θ1≤α, 0°≤α≤30° and -β≤θ2≤β, 0°≤β≤20°, the trainee's current action is determined to be the shooting preparation action.

[0030] Step 7 includes: identifying the shooting aiming action using the following formula:

[0031]

[0032] Where M2 represents the recognition condition of shooting aiming action, θ3 represents the angle between the forearm and the spine, and γ and δ are recognition adjustment parameters. When both conditions are met simultaneously, 90°-γ≤θ2≤90°+γ, 0°≤γ≤30° and 90°-δ≤θ3≤90°+δ, 0°≤δ≤20°, the trainee's current action is determined to be shooting aiming action.

[0033] Step 8 includes: calculating the time T1 from aiming to firing using the following formula:

[0034] T1 = T4 - T5

[0035] Where T4 represents the time when the shot is fired, and T5 represents the time when the shot is ready to be fired;

[0036] The time T2 from aiming to firing is calculated using the following formula:

[0037] T2 = T4 - T6

[0038] T6 represents the time when aiming the shot;

[0039] The time T3 from the shooting preparation action to the shooting aiming action is calculated using the following formula:

[0040] T3 = T5 - T6.

[0041] In its specific implementation, this application provides a computer storage medium and a corresponding data processing unit. The computer storage medium is capable of storing a computer program, which, when executed by the data processing unit, can run the invention's content regarding an intelligent detection method for rapid firing actions, as well as some or all of the steps in various embodiments. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0042] Those skilled in the art will clearly understand that the technical solutions in the embodiments of the present invention can be implemented using computer programs and their corresponding general-purpose hardware platforms. Based on this understanding, the technical solutions in the embodiments of the present invention, or the parts that contribute to the prior art, can be embodied in the form of computer programs, i.e., software products. These computer program software products can be stored in a storage medium and include several instructions to cause a device containing a data processing unit (which may be a personal computer, server, microcontroller, MUU, or network device, etc.) to execute the methods described in various embodiments or certain parts of the embodiments of the present invention.

[0043] This invention provides an intelligent detection method for rapid firing actions. Many methods and approaches exist for implementing this technical solution; the above description is merely a preferred embodiment of the invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of this invention, and these improvements and modifications should also be considered within the scope of protection of this invention. All components not explicitly stated in this embodiment can be implemented using existing technologies.

Claims

1. A method for intelligently detecting a quick shot action, the method comprising: The method comprises the following steps: Step 1: collecting video images of the shooting process of the trainer in real time through a camera; Step 2: extracting the joint position of the trainer in the video images of the shooting process of the trainer to obtain joint information; Step 3: screening the joint information obtained in Step 2 to extract the joint information of the person closest to the camera; Step 4: classifying the joint information extracted in Step 3 to obtain a category sequence; Step 5: extracting the joint information of the arm and the spine and calculating angle information; Step 6: identifying a shooting preparation action according to the joint information of the arm and the spine and recording the time of the action; Step 7: identifying a shooting aiming action according to the joint information of the arm and the spine and recording the time of the action; Step 8: obtaining a shooting trigger signal of the trainer, obtaining the time of the shooting trigger, combining the time of the identified shooting preparation action and the shooting aiming action to calculate the time used in the shooting process, and obtaining three times: the time T1 used from the shooting preparation action to the shooting trigger, the time T2 used from the shooting aiming action to the shooting trigger, and the time T3 used from the shooting preparation action to the shooting aiming action; Step 6 comprises: identifying the shooting preparation action by using the following formula: Wherein, M1 represents the identification condition of the shooting preparation action, θ1 represents the included angle between the upper arm and the lower arm, θ2 represents the included angle between the upper arm and the spine, and α and β are identification adjustment parameters. When the two conditions of -α≤θ1≤α, 0°≤α≤30° and -β≤θ2≤β, 0°≤β≤20° are simultaneously satisfied, it is determined that the current action of the trainer is the shooting preparation action. Step 7 comprises: identifying the shooting aiming action by using the following formula: Wherein, M2 represents the identification condition of the shooting aiming action, θ3 represents the included angle between the lower arm and the spine, and γ and δ are identification adjustment parameters. When the two conditions of 90°-γ≤θ2≤90°+γ, 0°≤γ≤30° and 90°-δ≤θ3≤90°+δ, 0°≤δ≤20° are simultaneously satisfied, it is determined that the current action of the trainer is the shooting aiming action.

2. The method of claim 1, wherein, Step 3 comprises: adjusting the camera screen to be parallel to the shooting direction, so that the perpendicular vector is v(0,W), W is the image resolution width, and the average distance D of the joint of the person from the perpendicular line is calculated according to the following formula: where n is the total number of joints, x i is the horizontal coordinate of the ith joint; the joint information of the training staff with the smallest D is selected for processing.

3. The method of claim 2, wherein, In Step 4, the joint information comprises position, confidence and category p(x,y,c,f), x is the horizontal coordinate, y is the vertical coordinate, c is the joint category, and f is the confidence, and f takes a value in the range of 0-1.

4. The method of claim 3, wherein, Step 5 comprises: The coordinates of the two end joints of the upper arm on the image are p(x1,y1) and p(x2,y2), the coordinates of the two end joints of the lower arm on the image are p(x2,y2) and p(x3,y3), and the coordinates of the two end joints of the spine on the image are p(x5,y5) and p(x6,y6). The angle formula is as follows: θ1 represents the included angle between the upper arm and the lower arm, θ2 represents the included angle between the upper arm and the spine, and θ3 represents the included angle between the lower arm and the spine.

5. The method of claim 4, wherein, Step 8 comprises: calculating the time T1 used from the shooting aiming action to the shooting trigger by using the following formula: T1=T4-T5 Wherein, T4 represents the time when the shooting trigger occurs, and T5 represents the time when the shooting preparation occurs. The time T2 used for the shooting aiming action to the shooting firing is calculated by the following formula: T2=T4-T6 Wherein, T6 represents the time when the shooting aiming is performed; The time T3 used for the shooting preparation action to the shooting aiming action is calculated by the following formula: T3=T5-T6.

Images