A method and system for reconstructing the outer profile function of a projectile head based on machine vision

Abstract

This invention provides a method and system for reconstructing the outer contour function of a projectile's head shape based on machine vision, belonging to the field of machine vision image processing and automatic parameter calculation technology. First, the invention performs semantic segmentation on the frontal view image of the projectile's head shape to obtain a binary mask of the projectile target, and extracts and crops the outer contour of the head shape segment. Second, through coordinate axis self-calibration and attitude alignment, the standard attitude coordinates of the projectile's head shape outer contour are obtained. Then, in the local coordinate system, geometric constraint feature points are extracted to obtain the endpoints, pixel areas, and single-point coordinates of the projectile's head shape outer contour. Third, a two-parameter head shape outer contour shape function satisfying endpoint constraints is constructed, and the shape parameters are solved using area constraints and single-point constraints to obtain the analytical expression of the head shape outer contour. Finally, the outer contour function and its derivative are substituted into the head shape factor integral formula to achieve automatic calculation of the head shape factor.

Description

Technical Field

[0001] This invention relates to the field of machine vision image processing and automatic parameter calculation technology, and in particular to a method and system for reconstructing the outer contour function of a projectile head based on machine vision. Background Technology

[0002] In high-speed penetration models, the projectile nose shape factor is one of the key parameters characterizing the sharpness of the projectile nose and its aerodynamic or penetration characteristics. It is often used to describe the influence of the projectile nose geometry on the penetration load response, velocity decay, and energy dissipation process, and to provide calibration basis for empirical or semi-empirical terms in the penetration model. With the continuous expansion of simulation and experimental data scale and the increasing demand for rapid evaluation in engineering applications, the efficient and accurate acquisition of the nose shape factor is of great significance. The calculation of the nose shape factor is highly dependent on the geometric information of the projectile nose's outer contour, especially on the functional expression of the outer contour curve, the curve derivative characteristics, and global quantities such as the area of ​​the region enclosed by the outer contour and the coordinate axes. Therefore, accurate extraction and stable modeling of the nose contour are the fundamental prerequisites for reliable calculation of the nose shape factor.

[0003] In current research and engineering practice, the acquisition of the projectile's head shape outline often employs manual measurement or semi-automatic image processing procedures. These include manually calibrating coordinate axes, manually selecting key points or contour points, and then obtaining the outline curve through polynomial fitting, spline fitting, or piecewise linear fitting, thereby calculating the head shape factor and related geometric parameters. However, this approach still has shortcomings in practical applications: Firstly, coordinate axis calibration and attitude alignment often rely on manual experience or simple rules, making them susceptible to interference from shooting angle, noise, lighting, and background, leading to unstable extraction of key geometric quantities such as head shape length and maximum radius. Secondly, high-order fitting based on dense contour points is sensitive to local burrs and gaps, easily causing fitting oscillations and unstable curve derivatives, thus amplifying errors in the integral calculation of the head shape factor, which includes derivative terms. Furthermore, many methods only utilize endpoints or a small number of geometric points for constraint, making it difficult to guarantee the uniqueness and repeatability of the curve shape when endpoints are consistent, resulting in significant differences in calculation results under different operators and different fitting strategies. Summary of the Invention

[0004] To address the problems in the prior art, this invention provides a method and system for reconstructing the outer contour function of a projectile's head shape based on machine vision. The invention first performs semantic segmentation on the frontal view image of the projectile's head shape to obtain a binary mask of the projectile target, and then extracts and crops the outer contour of the head shape segment. Second, through coordinate axis self-calibration and attitude alignment, the standard attitude coordinates of the projectile's head shape outer contour are obtained. Then, in the local coordinate system, geometric constraint feature points are extracted to obtain the endpoints, pixel areas, and single-point coordinates of the projectile's head shape outer contour. Third, a two-parameter head shape outer contour shape function satisfying endpoint constraints is constructed, and the shape parameters are solved using area constraints and single-point constraints to obtain the analytical expression of the head shape outer contour. Finally, the outer contour function and its derivative are substituted into the head shape factor integral formula to achieve automatic calculation of the head shape factor. To achieve the above objectives, the technical solution is as follows: On one hand, the present invention provides a method for reconstructing the outer contour function of a projectile head shape based on machine vision, the method comprising: S1. Use a camera to photograph the projectile during the aerodynamic or penetration process to obtain a frontal photograph of the projectile. S2. Based on the frontal view photo, a binarized mask is obtained through preprocessing and target segmentation; S3. Based on the binarized mask, the head shape outer contour point dataset is obtained by extracting the outer contour and cutting the head shape segment. S4. Based on the head shape outer contour point dataset, obtain the standard posture coordinates of the head shape outer contour through coordinate axis self-calibration and posture alignment; S5. Based on the standard posture coordinates of the outer contour of the head shape, geometric constraints are obtained through feature extraction. The geometric constraints include: endpoint data of the outer contour of the head shape, the point dataset of the outer contour of the head shape, the pixel area data enclosed by the coordinate axis and the endpoint data of the outer contour of the head shape, and single-point data of the curve. S6. Based on the geometric constraints, the outer contour function expression of the head shape is obtained by constructing a two-parameter projectile head shape outer contour function; S7. Based on the function expression of the head shape outer contour, a consistency check is performed. If the check is successful, a stable function expression of the head shape outer contour is obtained; otherwise, return to step S6. S8. Based on the stable function expression of the outer contour of the head shape, obtain the head shape factor data through the head shape factor calculation model.

[0005] Optionally, in S2, a binarized mask is obtained based on the frontal view photograph through preprocessing and target segmentation, including: S21. Based on the frontal photograph, a clear image is obtained through image preprocessing such as denoising and grayscale conversion; S22. Based on the clear image, obtain a pixel-level classification image through semantic segmentation; S23. Based on the pixel-level classification image, a preliminary binarization mask is obtained through binarization transformation; S24. Based on the preliminary binarized mask, a binarized mask is obtained through post-mask processing.

[0006] Optionally, in S3, based on the binarized mask, a head shape outer contour point dataset is obtained through outer contour extraction and head shape segment clipping, including: S31. Based on the binary mask, the edge features of the projectile are extracted using an edge detection algorithm to obtain the outer contour data; S32. Based on the outer contour data, obtain the head shape outer contour point dataset by extracting the head shape contour of the projectile.

[0007] Optionally, in S4, based on the head shape outer contour point dataset, the standard pose coordinates of the head shape outer contour are obtained through coordinate axis self-calibration and pose alignment, including: S41. Based on the dataset of head shape outer contour points, obtain the coordinates of the head shape outer contour by constructing a local coordinate system; S42. Based on the coordinates of the outer contour of the head shape, the standard posture coordinates of the outer contour of the head shape are obtained through posture alignment.

[0008] Optionally, in step S7, based on the head shape outer contour function expression, a consistency check is performed. If the check is successful, a stable function expression for the head shape outer contour is obtained; otherwise, the process returns to step S6, which includes: S71. Based on the function expression of the outer contour of the head shape, the function value of the single point of the curve is obtained by calculating the single point data of the curve. S72. Based on the function expression of the head shape's outer contour, the area of ​​the head shape is calculated using the pixel area data; S73. Based on the single-point function value of the curve and the area of ​​the head shape, a consistency threshold is compared to determine the following rules: Judgment Rule 1: If the threshold comparison is qualified, the stable function expression of the head shape outer contour is obtained; Judgment Rule 2: If the threshold comparison fails, return to step S6.

[0009] Optionally, in step S8, based on the stable function expression of the head shape's outer contour, head shape factor data is obtained through a head shape factor calculation model, including: S81. Based on the stable function expression of the outer contour of the head shape, the derivative expression is obtained by differentiation calculation; S82. Based on the derivative expression, the head shape factor data is obtained through the head shape factor physical formula (1); (1) In the formula, It is a head shape factor.R It is the coordinate axis x The coordinates of the endpoints of the axis, h It refers to the coordinate axes. y The coordinates of the endpoints of the axis p , q These are the two undetermined parameters of the two-parameter projectile head shape outer contour function. x It is the independent variable in the stable function expression of the head shape's outer contour. y It is the dependent variable in the stable functional expression of the head shape's outer contour. It is the derivative value of the stable function expression of the head shape's outer contour. yes x / h The normalized value.

[0010] On the other hand, the present invention provides a machine vision-based system for reconstructing the outer contour function of a projectile's nose shape. This system is applied to a machine vision-based method for reconstructing the outer contour function of a projectile's nose shape. The system includes: The image acquisition module is used to take pictures of the projectile during the aerodynamic or penetration process using a camera, and obtain a frontal photograph of the projectile. The binarization mask acquisition module is used to obtain a binarization mask based on the frontal view photo through preprocessing and target segmentation. The outer contour point acquisition module is used to obtain the head shape outer contour point dataset based on the binarized mask by extracting the outer contour and cropping the head shape segment; The standard head shape outer contour coordinate acquisition module is used to obtain the standard posture coordinates of the head shape outer contour based on the head shape outer contour point dataset, through coordinate axis self-calibration and posture alignment; The geometric constraint feature extraction module is used to extract geometric constraints based on the standard posture coordinates of the head's outer contour through feature element extraction. The module for obtaining the outer contour function expression of the head shape is used to obtain the outer contour function expression of the head shape by constructing a two-parameter projectile head shape outer contour shape function based on the geometric constraints. The expression verification module is used to obtain a stable function expression for the outer contour of the head shape by performing a consistency check based on the function expression of the head shape's outer contour. The head shape factor calculation module is used to obtain head shape factor data based on the stable function expression of the outer contour of the head shape through the head shape factor calculation model.

[0011] Compared with the prior art, the technical solution of the present invention has at least the following beneficial effects: The above-mentioned scheme has several advantages. First, by robustly fitting near-horizontal and near-vertical reference boundaries to the head shape outer contour point set, it achieves automatic determination of local coordinate axes and attitude alignment, reducing subjective errors caused by manual calibration and improving the stability and consistency of head shape geometry extraction under different imaging attitudes, noise, and background interference conditions. Second, it uses the pixel area of ​​the closed region formed by the outer contour and coordinate axes as a global constraint and introduces the curve intersection point at half the head shape length as a local constraint, so that the two-parameter head shape outer contour shape function can be uniquely solved by finite constraint quantities under endpoint conditions, avoiding the problems of shape uncertainty, non-repeatable results, and high sensitivity to contour quality caused by traditional methods that rely solely on endpoint or dense contour fitting. Third, the obtained outer contour function has analytical expression and differentiability, effectively suppressing the oscillation and derivative instability of polynomial and spline fitting, thereby improving the numerical stability and calculation accuracy of head shape factor integral calculation containing derivative terms, facilitating automated processing in engineering field or batch data conditions, and thus providing reliable support for rapid reconstruction of projectile head shape outer contour function and stable automatic extraction of head shape factors. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 This is a flowchart of an embodiment of the projectile head shape outer contour function reconstruction method based on machine vision of the present invention; Figure 2 This is a flowchart of the binarized mask acquisition process in an embodiment of the machine vision-based projectile head shape outer contour function reconstruction method of the present invention. Figure 3 This is a flowchart of the process for obtaining the outer contour points in an embodiment of the projectile head shape outer contour function reconstruction method based on machine vision of the present invention; Figure 4 This is a flowchart of the standard head shape outer contour coordinate acquisition process of an embodiment of the projectile head shape outer contour function reconstruction method based on machine vision of the present invention. Figure 5 Yes, this is a schematic diagram of the construction of the local coordinate axes of the projectile in an embodiment of the machine vision-based projectile head shape outer contour function reconstruction method of the present invention. Figure 6 This is a flowchart of the expression verification process for an embodiment of the projectile head shape outer contour function reconstruction method based on machine vision of the present invention. Figure 7This is a flowchart of the head shape factor calculation process in an embodiment of the projectile head shape outer contour function reconstruction method based on machine vision of the present invention. Figure 8 This is a system block diagram of an embodiment of the projectile head shape outer contour function reconstruction system based on machine vision of the present invention. Detailed Implementation

[0014] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0015] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0016] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0017] like Figure 1 The flowchart shown is an embodiment of the projectile nose shape outer contour function reconstruction method based on machine vision of the present invention. The present invention provides a projectile nose shape outer contour function reconstruction method based on machine vision, which is implemented by a projectile nose shape outer contour function reconstruction system based on machine vision. The method includes: S1. Use a camera to photograph the projectile during the aerodynamic or penetration process to obtain a frontal photograph of the projectile.

[0018] S2. Based on the frontal view photo, a binarized mask is obtained through preprocessing and target segmentation; Specifically, such as Figure 2 The flowchart shown is a representation of the machine vision-based projectile head shape contour function reconstruction method of the present invention, illustrating the binarized mask acquisition process. In step S2, based on the frontal view photograph, a binarized mask is obtained through preprocessing and target segmentation, including: S21. Based on the frontal photograph, a clear image is obtained through image preprocessing such as denoising and grayscale conversion; S22. Based on the clear image, obtain a pixel-level classification image through semantic segmentation; S23. Based on the pixel-level classification image, a preliminary binarization mask is obtained through binarization transformation; S24. Based on the preliminary binarized mask, a binarized mask is obtained through post-mask processing.

[0019] S3. Based on the binarized mask, the head shape outer contour point dataset is obtained by extracting the outer contour and cutting the head shape segment. Specifically, such as Figure 3 The flowchart shown is an embodiment of the projectile head shape outer contour function reconstruction method of the present invention. In step S3, based on the binarized mask, the outer contour extraction and head shape segment truncation are used to obtain the head shape outer contour point dataset, including: S31. Based on the binary mask, the edge features of the projectile are extracted using an edge detection algorithm to obtain the outer contour data; S32. Based on the outer contour data, obtain the head shape outer contour point dataset by extracting the head shape contour of the projectile.

[0020] Outer contour extraction involves connected component filtering and morphological processing (including isolated point removal, hole filling, and boundary smoothing) of the binarized mask to obtain the outer contour point set of the projectile. Then, based on the axial range or the location of abrupt curvature changes, the outer contour point set of the head-shaped segment is extracted, denoted as the head-shaped outer contour point dataset C. h .

[0021] S4. Based on the head shape outer contour point dataset, obtain the standard posture coordinates of the head shape outer contour through coordinate axis self-calibration and posture alignment; Specifically, such as Figure 4 The flowchart shown is a standard head shape outer contour coordinate acquisition flowchart of an embodiment of the projectile head shape outer contour function reconstruction method based on machine vision of the present invention. In step S4, based on the head shape outer contour point dataset, the standard posture coordinates of the head shape outer contour are obtained through coordinate axis self-calibration and posture alignment, including: S41. Based on the dataset of head shape outer contour points, obtain the coordinates of the head shape outer contour by constructing a local coordinate system; S42. Based on the coordinates of the outer contour of the head shape, obtain the standard posture coordinates of the outer contour of the head shape through posture alignment; Furthermore, such as Figure 5 The diagram shown illustrates the construction of the projectile's local coordinate axes in an embodiment of the machine vision-based projectile head shape outer contour function reconstruction method of the present invention. The local coordinate axes are constructed using the head shape outer contour point dataset C. h Two approximately orthogonal reference boundaries are identified, and robust line fitting is performed on each: 1) Let the approximate horizontal reference line obtained by fitting be denoted as L. x Its direction is used to define x Positive axis; 2) Let L be the approximate perpendicular line obtained by fitting. y Its direction is used to define y Positive axis; Among them, the straight line fitting can be robustly fitted using algorithms such as Hough transform, and an angle threshold is set to ensure that Lx and Ly are approximately orthogonal, with the intersection of the two lines as the coordinate origin O=L. x ∩L y This completes the self-calibration of the coordinate axes and the alignment of the orientation.

[0022] S5. Based on the standard posture coordinates of the outer contour of the head shape, geometric constraints are obtained through feature extraction. The geometric constraints include: endpoint data of the outer contour of the head shape, the point dataset of the outer contour of the head shape, the pixel area data enclosed by the coordinate axis and the endpoint data of the outer contour of the head shape, and single-point data of the curve. Specifically, in the local coordinate system, the outer contour of the head shape is represented as a single-valued function y=f(x), where x For axial coordinates, y Radial coordinates, Figure 5 Point M lies on the function. Extract the following geometric constraints: 1) Head shape length of outer contour endpoint data h : is the intercept corresponding to the intersection of the outer contour and the x-axis, satisfying f(h)=0; 2) Maximum radius of the outer contour endpoint data R : is the intercept corresponding to the intersection of the outer contour and the y-axis, satisfying f(0) = R ; 3) Single-point M constraint on the curve: x = h / 2 Draw a vertical line at point M, which intersects the outer contour to obtain the intersection point M. h / 2 , y 1 / 2 ),in, (1) 4) Pixel area constraint: determined by the outer contour curve and... x axis, y The area of ​​the closed region enclosed by the axes is denoted as S. The closed region is the area formed by the outer contour curve segment and the positive semi-axes of the two coordinate axes. The number of target pixels within the closed region is N, and the physical area corresponding to a single pixel is... The area of ​​the region is: (2) in, This can be obtained from image calibration, when the length-pixel conversion factors in the horizontal and vertical directions are respectively... , When, its value is: (3) S6. Based on the geometric constraints, the outer contour function expression of the head shape is obtained by constructing a two-parameter projectile head shape outer contour function.

[0023] Specifically, first, a two-parameter projectile head outer contour shape function is constructed: Dimensionless coordinates: Let (4) Selecting basis functions (in, p >0 is used to adjust the convergence distribution, and a power exponent is introduced. q >0 is used to adjust the overall "blunt / sharp" shape to meet the endpoint conditions. v(0) =1 and v(1) =0, thus obtaining the normalized shape function: (5) in, p and q The shape parameters of the two-parameter head shape function correspond to the physical scale head shape outer contour function expression: (6) The function satisfies f(0) = h , f(R)=0 And in p >0, q When >0, it can maintain non-negativity and convergence; its first derivative is... f'(x) It can be parsed as follows: (7) Then, the parameters are uniquely determined by the joint constraints of area and single point on the curve. p and q First, define the dimensionless constraint quantity. A and r for: (8) (9) By single-point constraint (in x=h / 2 place, that is u=1 / 2 )get: (10) At the same time, the curve and x axis, y The area enclosed by the axes satisfies: (11) Will f(x) Substitute and command u=x / h ,get: (12) Dividing both sides by Rh yields the normalized area constraint equation: (13) Finally, the one-dimensional root-finding method (preferably the bisection method) is used to obtain the solution. p , and by q(p) Seek q Thus, the outer contour function is uniquely determined. f(x) In the process of finding the root, the above integral can be calculated by numerical integration, and an error threshold ε can be set as the termination condition.

[0024] S7. Based on the function expression of the head shape outer contour, a consistency check is performed. If the check is successful, a stable function expression of the head shape outer contour is obtained; otherwise, return to step S6. Specifically, such as Figure 6 The flowchart shown is an example of the expression verification process for an embodiment of the projectile head shape outer contour function reconstruction method based on machine vision of the present invention. In step S7, the consistency verification is performed based on the head shape outer contour function expression. If the consistency verification is successful, a stable function expression for the head shape outer contour is obtained; otherwise, the process returns to step S6, including: S71. Based on the function expression of the outer contour of the head shape, the function value of the single point of the curve is obtained by calculating the single point data of the curve. Furthermore, according to equation (1), we obtain y 1 / 2 ,

[0025] S72. Based on the function expression of the head shape's outer contour, the area of ​​the head shape is calculated using the pixel area data; Furthermore, according to equation (12), the area S is obtained; S73. Based on the single-point function value of the curve and the area of ​​the head shape, a consistency threshold is compared to determine the following rules: Judgment Rule 1: If the threshold comparison is qualified, the stable function expression of the head shape outer contour is obtained; Furthermore, When the threshold comparison is qualified, the stable function expression of the head shape outer contour is obtained; Judgment Rule 2: If the threshold comparison fails, return to step S6; Furthermore, when If the conditions cannot be met simultaneously, the threshold comparison is deemed unqualified, and the process returns to step S6.

[0026] S8. Based on the stable function expression of the outer contour of the head shape, obtain the head shape factor data through the head shape factor calculation model; Specifically, such as Figure 7 The flowchart shown is a representation of the head shape factor calculation process of the machine vision-based projectile head shape contour function reconstruction method of the present invention. In step S8, head shape factor data is obtained through the head shape factor calculation model based on the stable function expression of the head shape contour, including: S81. Based on the stable function expression of the outer contour of the head shape, the derivative expression is obtained by differentiation calculation; S82. Based on the derivative expression, the head shape factor data is obtained through the head shape factor physical formula (14); Furthermore, head shape factor The dimensionless integral of the calculation formula is expressed as: (14) In physical coordinates Downtime: (15) like Figure 8 The diagram shown is a system block diagram of an embodiment of the projectile head shape outer contour function reconstruction system based on machine vision of the present invention. The present invention provides a projectile head shape outer contour function reconstruction system based on machine vision, which is applied to a projectile head shape outer contour function reconstruction method based on machine vision. The system includes: an image acquisition module, a binarization mask acquisition module, an outer contour point acquisition module, a standard head shape outer contour coordinate acquisition module, a geometric constraint feature extraction module, a head shape outer contour function expression acquisition module, an expression verification module, and a head shape factor calculation module. Specifically: The image acquisition module is used to take pictures of the projectile during the aerodynamic or penetration process using a camera, and obtain a frontal photograph of the projectile. The binarization mask acquisition module is used to obtain a binarization mask based on the frontal view photo through preprocessing and target segmentation. The outer contour point acquisition module is used to obtain the head shape outer contour point dataset based on the binarized mask by extracting the outer contour and cropping the head shape segment; The standard head shape outer contour coordinate acquisition module is used to obtain the standard posture coordinates of the head shape outer contour based on the head shape outer contour point dataset, through coordinate axis self-calibration and posture alignment; The geometric constraint feature extraction module is used to extract geometric constraints based on the standard posture coordinates of the head's outer contour through feature element extraction. The module for obtaining the outer contour function expression of the head shape is used to obtain the outer contour function expression of the head shape by constructing a two-parameter projectile head shape outer contour shape function based on the geometric constraints. The expression verification module is used to obtain a stable function expression for the outer contour of the head shape by performing a consistency check based on the function expression of the head shape's outer contour. The head shape factor calculation module is used to obtain head shape factor data based on the stable function expression of the outer contour of the head shape through the head shape factor calculation model.

[0027] This invention provides a method and system for reconstructing the outer contour function of a projectile's head shape based on machine vision. First, the invention performs semantic segmentation on the frontal view image of the projectile's head shape to obtain a binary mask of the projectile target, and extracts and crops the outer contour of the head shape segment. Second, through coordinate axis self-calibration and attitude alignment, the standard attitude coordinates of the projectile's head shape outer contour are obtained. Then, in the local coordinate system, geometric constraint feature points are extracted to obtain the endpoints, pixel areas, and single-point coordinates of the projectile's head shape outer contour. Third, a two-parameter head shape outer contour shape function satisfying endpoint constraints is constructed, and the shape parameters are solved using area constraints and single-point constraints to obtain the analytical expression of the head shape outer contour. Finally, the outer contour function and its derivative are substituted into the head shape factor integral formula to achieve automatic calculation of the head shape factor.

[0028] It is understood that the present invention has been described through the above embodiments and should not be construed as limiting the implementation and scope of the present invention. Those skilled in the art will recognize that various changes or equivalent substitutions can be made to these features and embodiments without departing from the spirit and scope of the present invention. Furthermore, under the teachings of the present invention, these features and embodiments can be modified to adapt to specific situations and materials without departing from the spirit and scope of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the protection scope of the present invention.

Claims

1. A method for reconstructing the outer contour function of a projectile's head shape based on machine vision, characterized in that, The method includes: S1. Use a camera to photograph the projectile during the aerodynamic or penetration process to obtain a frontal photograph of the projectile. S2. Based on the frontal view photograph, a binarized mask is obtained through preprocessing and target segmentation; S3. Based on the binarized mask, obtain the head shape outer contour point dataset by extracting the outer contour and cropping the head shape segment; S4. Based on the head shape outer contour point dataset, obtain the standard posture coordinates of the head shape outer contour through coordinate axis self-calibration and posture alignment; S5. Based on the standard posture coordinates of the head shape outer contour, geometric constraints are obtained through feature element extraction. The geometric constraints include: head shape outer contour endpoint data, head shape outer contour point dataset, pixel area data enclosed by the coordinate axes and the head shape outer contour endpoint data, and curve single point data. S6. Based on the geometric constraints, the outer contour function expression of the head shape is obtained by constructing a two-parameter projectile head shape outer contour function; S7. Based on the head shape outer contour function expression, a consistency check is performed. If the check is successful, a stable function expression for the head shape outer contour is obtained; otherwise, return to step S6. S8. Based on the stable function expression of the head shape outer contour, obtain the head shape factor data through the head shape factor calculation model.

2. The method for reconstructing the outer contour function of a projectile head shape based on machine vision according to claim 1, characterized in that, In step S2, a binarized mask is obtained based on the frontal view photograph through preprocessing and target segmentation, including: S21. Based on the frontal photograph, a clear image is obtained through image preprocessing such as denoising and grayscale conversion; S22. Based on the clear image, obtain a pixel-level classification image through semantic segmentation; S23. Based on the pixel-level classification image, a preliminary binarization mask is obtained through binarization transformation; S24. Based on the preliminary binarized mask, a binarized mask is obtained through post-mask processing.

3. The method for reconstructing the outer contour function of a projectile head shape based on machine vision according to claim 1, characterized in that, In step S3, based on the binarized mask, a head shape outer contour point dataset is obtained through outer contour extraction and head shape segment clipping, including: S31. Based on the binarized mask, the edge features of the projectile are extracted using an edge detection algorithm to obtain the outer contour data; S32. Based on the outer contour line data, obtain the head shape outer contour point dataset by extracting the head shape contour of the projectile.

4. The method for reconstructing the outer contour function of a projectile head shape based on machine vision according to claim 1, characterized in that, In step S4, based on the head shape outer contour point dataset, the standard posture coordinates of the head shape outer contour are obtained through coordinate axis self-calibration and posture alignment, including: S41. Based on the head shape outer contour point dataset, obtain the head shape outer contour coordinates by constructing a local coordinate system; S42. Based on the coordinates of the outer contour of the head shape, the standard posture coordinates of the outer contour of the head shape are obtained through posture alignment.

5. The method for reconstructing the outer contour function of a projectile head shape based on machine vision according to claim 1, characterized in that, In step S7, based on the head shape outer contour function expression, a consistency check is performed. If the check is successful, a stable function expression for the head shape outer contour is obtained; otherwise, the process returns to step S6, including: S71. Based on the head shape outer contour function expression, calculate the curve single-point function value using the curve single-point data; S72. Based on the head shape outer contour function expression, the head shape area is calculated using the pixel area data; S73. Based on the single-point function value of the curve and the head shape area, a consistency threshold is compared to determine the following rules: Judgment Rule 1: If the threshold comparison is qualified, the stable function expression of the head shape outer contour is obtained; Judgment Rule 2: If the threshold comparison fails, return to step S6.

6. The method for reconstructing the outer contour function of a projectile head shape based on machine vision according to claim 1, characterized in that, In step S8, based on the stable function expression of the head shape outer contour, head shape factor data is obtained through the head shape factor calculation model, including: S81. Based on the stable function expression of the head shape's outer contour, the derivative expression is obtained by differentiation calculation; S82. Based on the derivative expression, obtain the head shape factor data through the head shape factor physical formula (1); （1） In the formula, It is the head shape factor, and R is the coordinate axis. x The coordinates of the endpoints of the axis, h It refers to the coordinate axes. y endpoint coordinates of the axis ,p , q These are the two undetermined parameters of the two-parameter projectile head shape outer contour function. x It is the independent variable in the stable function expression of the head shape's outer contour. y It is the dependent variable in the stable functional expression of the head shape's outer contour. It is the derivative value of the stable function expression of the head shape's outer contour. yes x / h The normalized value.

7. A machine vision-based projectile head shape outer contour function reconstruction system, used to implement the machine vision-based projectile head shape outer contour function reconstruction method as described in any one of claims 1-6, characterized in that, The system includes: An image acquisition module is used to capture a frontal photograph of the projectile during aerodynamic or penetration processes using a camera. The binarization mask acquisition module is used to obtain a binarization mask based on the frontal view photo through preprocessing and target segmentation. The outer contour point acquisition module is used to obtain a head shape outer contour point dataset based on the binarized mask by extracting the outer contour and cropping the head shape segment; The standard head shape outer contour coordinate acquisition module is used to obtain the standard posture coordinates of the head shape outer contour based on the head shape outer contour point dataset by self-calibrating the coordinate axes and aligning with the posture. The geometric constraint feature extraction module is used to obtain geometric constraints by extracting feature elements based on the standard posture coordinates of the head shape's outer contour. The head shape outer contour function expression acquisition module is used to obtain the head shape outer contour function expression by constructing a two-parameter projectile head shape outer contour shape function according to the geometric constraints. The expression verification module is used to obtain a stable function expression for the head shape outer contour by performing a consistency check based on the head shape outer contour function expression. The head shape factor calculation module is used to obtain head shape factor data based on the stable function expression of the head shape outer contour and through the head shape factor calculation model.

Images