Building ceramic deformation degree automatic detection method based on vision

Abstract

The invention discloses an architectural ceramic deformation degree automatic detection method based on vision, and belongs to the technical field of deformation detection, and the method comprises the following steps: S1, obtaining a production digital image of architectural ceramic, and carrying out the preprocessing of the production digital image, and generating a standard digital image; s2, determining a plurality of grid models according to a plurality of angular points of the standard digital image, and calculating principal component parameters; and S3, generating a deformation detection result of the architectural ceramic according to the principal component parameters of the standard digital image and the edge strength value of each pixel point. The whole deformation detection process can be automated, manual intervention and errors are reduced, and tiny changes in the image can be captured, so that the deformation detection result has high sensitivity.

Description

Technical Field

[0001] The invention belongs to the technical field of deformation detection, and in particular relates to a vision-based automatic detection method for deformation of architectural ceramics. Background Art

[0002] As an important building material, the quality of building ceramics directly affects the safety and aesthetics of buildings. Traditional building ceramic deformation detection methods usually use manual measurement and observation, which is not only inefficient but also easily affected by human factors, resulting in inaccurate detection results. Therefore, it is of great significance to develop an efficient and accurate building ceramic deformation automatic detection method. Summary of the invention

[0003] In order to solve the above problems, the present invention proposes a method for automatically detecting deformation of architectural ceramics based on vision.

[0004] The technical solution of the present invention is: a method for automatically detecting deformation of architectural ceramics based on vision comprises the following steps: S1. Acquire the production digital image of building ceramics, and pre-process the production digital image to generate a standard digital image; S2. Determine a number of grid models according to a number of corner points of the standard digital image, and calculate the principal component parameters; S3. Generate deformation detection results of building ceramics according to the principal component parameters of the standard digital image and the edge intensity value of each pixel point.

[0005] Furthermore, S2 includes the following sub-steps: S21, extracting a number of corner points of the standard digital image; S22, calculating a stability weight of a standard digital image according to a plurality of corner points; S23, determining a plurality of grid models using a plurality of corner points of the standard digital image and a stability weight of the standard digital image; S24. Calculate the principal component parameters of the standard digital image based on a plurality of grid models.

[0006] The beneficial effects of the above further scheme are: in the present invention, corner points are important feature points in the image, and by extracting corner points, the key information in the image can be accurately captured. Moreover, as feature points in the image, the number of corner points is relatively small, so the amount of data for subsequent processing can be greatly reduced; the grid model extracted by corner points can flexibly express the pixel value of important areas in the image, and the size of the grid can be adaptively adjusted. By calculating the principal component parameters, the main features in the image can be extracted.

[0007] Furthermore, in S22, the calculation formula of the stability weight Q of the standard digital image is: ;k n represents the curvature of the nth corner point, k n+1 represents the curvature of the n+1th corner point, and N represents the number of corner points of a standard digital image.

[0008] The beneficial effect of the above further scheme is: in the present invention, by calculating the curvature of each corner point and deriving the stability weight Q accordingly, the local stability of the image in different areas can be reflected. The stability weight Q can comprehensively consider the curvature values ​​of these various corner points and provide a quantitative evaluation for the overall stability of the image.

[0009] Further, S23 includes the following sub-steps: S231, generating a corresponding mesh side length for each corner point; S232, constructing a grid for each corner point based on the grid side length corresponding to each corner point and with the corner point position center; S233, generating a corresponding grid model according to the grid of each corner point and the stable weight of the standard digital image.

[0010] The beneficial effect of the above further scheme is that in the present invention, different grid side lengths are allowed to be generated for each corner point, providing great flexibility, and the grid side length can be adjusted according to the local features and shape changes of the image, thereby optimizing the representation ability of the grid model. By generating accurate grid side lengths for each corner point, it can be ensured that the grid model can accurately capture the details and features in the image. By combining the grids of each corner point, a complete grid model can be generated, which helps to retain all the information and features in the image and provide comprehensive support for subsequent image processing and analysis tasks.

[0011] Furthermore, in S231, the calculation formula of the grid side length L corresponding to the corner point is: ; Wherein, l represents the length of the standard digital image, w represents the width of the standard digital image, M represents the number of corner points of the standard digital image, max(·) represents the maximum value operation, and T represents the Euclidean distance between the corner point and the centroid of the standard digital image.

[0012] Furthermore, in S233, the expression of the grid model Y is: ; In the formula, σ(·) represents the activation function, W represents the weight of the convolution kernel, b represents the bias of the convolution kernel, and z n represents the pixel value of the nth pixel of the grid, N represents the number of pixels in the grid, and Q represents the stability weight of the standard digital image.

[0013] The beneficial effect of the above further solution is that in the present invention, the activation function can map the pixels of the input grid to a nonlinear space, thereby capturing more image details and features.

[0014] Furthermore, in S24, the calculation formula of the main component parameter MAIN of the standard digital image is: Where, Y m represents the grid model value of the mth corner point, P m represents the feature matrix of the grid corresponding to the mth corner point, ||·||2 represents the F2 norm operation, M represents the number of corner points, S m Represents the translation parameter between the mth corner point and the corner point closest to its Euclidean distance.

[0015] The translation parameter is used to describe the translation distance of one grid relative to another grid in a plane or space. If necessary, the translation parameter can be dimensionless. The pixel values ​​of the pixels in each grid are arranged according to the pixel position to obtain the characteristic matrix of the grid.

[0016] Furthermore, S3 includes the following sub-steps: S31, using the Kirsch operator to extract the edge intensity value of each pixel in the standard digital image; S32, constructing deformation determination conditions using principal component parameters, and taking pixel points whose edge strength values ​​do not meet the deformation determination conditions as deformation areas.

[0017] The beneficial effect of the above further scheme is: in the present invention, the Kirsch operator is an edge detection algorithm, which can extract edge feature parameters of pixel points in an image; after the image is processed by the Kirsch operator, each pixel point will have a corresponding edge intensity value. The deformation judgment condition constructed by the principal component parameters can realize the detection of image deformation.

[0018] Further, in S32, the expression of the deformation determination condition is: ; In the formula, MAIN represents the principal component parameter of the standard digital image, V represents the edge intensity value of the pixel point, Indicates rounding up, rand(0,1) indicates a random number generated between 0 and 1.

[0019] The beneficial effects of the present invention are as follows: the present invention extracts more characteristic grid areas in a standard digital image through corner points, calculates principal component parameters using grid model values, thereby capturing more detailed information and features, and having a certain robustness to image changes; determines whether the edge strength value of the pixel point satisfies the deformation determination condition constructed by the principal component parameters, and generates a final deformation determination result. The entire deformation detection process of the present invention can be automated, reducing manual intervention and errors, and can capture slight changes in the image, so the deformation detection result has high sensitivity. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1This is a flow chart of the vision-based automatic detection method for deformation of building ceramics. DETAILED DESCRIPTION

[0021] The embodiments of the present invention will be further described below in conjunction with the accompanying drawings.

[0022] like Figure 1 As shown, the present invention provides a method for automatically detecting deformation of architectural ceramics based on vision, comprising the following steps: S1. Acquire the production digital image of building ceramics, and pre-process the production digital image to generate a standard digital image; S2. Determine a number of grid models according to a number of corner points of the standard digital image, and calculate the principal component parameters; S3. Generate deformation detection results of building ceramics according to the principal component parameters of the standard digital image and the edge intensity value of each pixel point.

[0023] In this embodiment of the present invention, S2 includes the following sub-steps: S21, extracting a number of corner points of the standard digital image; S22, calculating a stability weight of a standard digital image according to a plurality of corner points; S23, determining a plurality of grid models using a plurality of corner points of the standard digital image and a stability weight of the standard digital image; S24. Calculate the principal component parameters of the standard digital image based on a plurality of grid models.

[0024] In the present invention, corner points are important feature points in an image. By extracting corner points, key information in an image can be accurately captured. Corner points are relatively few in number as feature points in an image, so the amount of data for subsequent processing can be greatly reduced. The grid model extracted by corner points can flexibly express the pixel values ​​of important areas in an image, and the size of the grid can be adaptively adjusted. By calculating the principal component parameters, the main features in an image can be extracted.

[0025] In the embodiment of the present invention, in S22, the calculation formula of the stability weight Q of the standard digital image is: ;k n represents the curvature of the nth corner point, k n+1 represents the curvature of the n+1th corner point, and N represents the number of corner points of a standard digital image.

[0026] In the present invention, by calculating the curvature of each corner point and deriving the stability weight Q accordingly, the local stability of the image in different areas can be reflected. The stability weight Q can comprehensively consider the curvature values ​​of these various corner points and provide a quantitative evaluation for the overall stability of the image.

[0027] In this embodiment of the present invention, S23 includes the following sub-steps: S231, generating a corresponding mesh side length for each corner point; S232, constructing a grid for each corner point based on the grid side length corresponding to each corner point and with the corner point position center; S233, generating a corresponding grid model according to the grid of each corner point and the stable weight of the standard digital image.

[0028] In the present invention, different grid side lengths are allowed to be generated for each corner point, providing great flexibility. The grid side length can be adjusted according to the local features and shape changes of the image, thereby optimizing the representation capability of the grid model. By generating accurate grid side lengths for each corner point, it can be ensured that the grid model can accurately capture the details and features in the image. By combining the grids of each corner point, a complete grid model can be generated, which helps to retain all the information and features in the image and provide comprehensive support for subsequent image processing and analysis tasks.

[0029] In the embodiment of the present invention, in S231, the calculation formula of the grid side length L corresponding to the corner point is: ; Wherein, l represents the length of the standard digital image, w represents the width of the standard digital image, M represents the number of corner points of the standard digital image, max(·) represents the maximum value operation, and T represents the Euclidean distance between the corner point and the centroid of the standard digital image.

[0030] In the embodiment of the present invention, in S233, the expression of the grid model Y is: ; In the formula, σ(·) represents the activation function, W represents the weight of the convolution kernel, b represents the bias of the convolution kernel, and z n represents the pixel value of the nth pixel of the grid, N represents the number of pixels in the grid, and Q represents the stability weight of the standard digital image.

[0031] In the present invention, the activation function can map the pixels of the input grid to a nonlinear space, thereby capturing more image details and features.

[0032] In the embodiment of the present invention, in S24, the calculation formula of the main component parameter MAIN of the standard digital image is: Where, Y m represents the grid model value of the mth corner point, P m represents the feature matrix of the grid corresponding to the mth corner point, ||·||2 represents the F2 norm operation, M represents the number of corner points, S m Represents the translation parameter between the mth corner point and the corner point closest to its Euclidean distance.

[0033] In the present invention, the translation parameter is used to describe the translation distance of one grid relative to another grid in a plane or space. If necessary, the translation parameter can be dimensionless. The pixel values ​​of the pixels in each grid are arranged according to the pixel positions to obtain the characteristic matrix of the grid.

[0034] In this embodiment of the present invention, S3 includes the following sub-steps: S31, using the Kirsch operator to extract the edge intensity value of each pixel in the standard digital image; S32, constructing deformation determination conditions using principal component parameters, and taking pixel points whose edge strength values ​​do not meet the deformation determination conditions as deformation areas.

[0035] In the present invention, the Kirsch operator is an edge detection algorithm that can extract edge feature parameters of pixel points in an image; after the image is processed by the Kirsch operator, each pixel point will have a corresponding edge intensity value. The deformation judgment condition constructed using the principal component parameters can realize the detection of image deformation.

[0036] In the embodiment of the present invention, in S32, the expression of the deformation determination condition is: ; In the formula, MAIN represents the principal component parameter of the standard digital image, V represents the edge intensity value of the pixel point, Indicates rounding up, rand(0,1) indicates a random number generated between 0 and 1.

[0037] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the principles of the present invention, and should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific variations and combinations that do not deviate from the essence of the present invention based on the technical revelations disclosed by the present invention, and these variations and combinations are still within the protection scope of the present invention.

Claims

1. A vision-based automatic detection method for deformation of architectural ceramics, characterized in that: The following steps are involved: S1. Acquire the production digital image of building ceramics, and pre-process the production digital image to generate a standard digital image; S2. Determine a number of grid models according to a number of corner points of the standard digital image, and calculate the principal component parameters; S3. Generate deformation detection results of building ceramics according to the principal component parameters of the standard digital image and the edge intensity value of each pixel point.

2. The visual-based automatic detection method for deformation of architectural ceramics according to claim 1 is characterized in that: The S2 comprises the following sub-steps: S21, extracting a number of corner points of the standard digital image; S22, calculating a stability weight of a standard digital image according to a plurality of corner points; S23, determining a plurality of grid models using a plurality of corner points of the standard digital image and a stability weight of the standard digital image; S24. Calculate the principal component parameters of the standard digital image based on a plurality of grid models.

3. The visual-based automatic detection method for deformation of architectural ceramics according to claim 2 is characterized in that: In S22, the calculation formula of the stability weight Q of the standard digital image is: ;k n represents the curvature of the nth corner point, k n+1 represents the curvature of the n+1th corner point, and N represents the number of corner points of a standard digital image.

4. The visual-based automatic detection method for deformation of architectural ceramics according to claim 2 is characterized in that: The S23 comprises the following sub-steps: S231, generating a corresponding mesh side length for each corner point; S232, constructing a grid for each corner point based on the grid side length corresponding to each corner point and with the corner point position center; S233, generating a corresponding grid model according to the grid of each corner point and the stable weight of the standard digital image.

5. The visual-based automatic detection method for deformation of architectural ceramics according to claim 4 is characterized in that: In S231, the calculation formula of the grid side length L corresponding to the corner point is: ; Wherein, l represents the length of the standard digital image, w represents the width of the standard digital image, M represents the number of corner points of the standard digital image, max(·) represents the maximum value operation, and T represents the Euclidean distance between the corner point and the centroid of the standard digital image.

6. The visual-based automatic detection method for deformation of architectural ceramics according to claim 4 is characterized in that: In S233, the expression of the grid model Y is: ; In the formula, σ(·) represents the activation function, W represents the weight of the convolution kernel, b represents the bias of the convolution kernel, and z n represents the pixel value of the nth pixel of the grid, N represents the number of pixels in the grid, and Q represents the stability weight of the standard digital image.

7. The visual-based automatic detection method for deformation of architectural ceramics according to claim 2 is characterized in that: In S24, the calculation formula of the main component parameter MAIN of the standard digital image is: Where, Y m represents the grid model value of the mth corner point, P m represents the feature matrix of the grid corresponding to the mth corner point, ||·||2 represents the F2 norm operation, M represents the number of corner points, S m Represents the translation parameter between the mth corner point and the corner point closest to its Euclidean distance.

8. The visual-based automatic detection method for deformation of architectural ceramics according to claim 1 is characterized in that: The S3 comprises the following sub-steps: S31, using the Kirsch operator to extract the edge intensity value of each pixel in the standard digital image; S32, constructing deformation determination conditions using principal component parameters, and taking pixel points whose edge strength values ​​do not meet the deformation determination conditions as deformation areas.

9. The visual-based automatic detection method for deformation of architectural ceramics according to claim 8 is characterized in that: In S32, the expression of the deformation determination condition is: ; In the formula, MAIN represents the principal component parameter of the standard digital image, V represents the edge intensity value of the pixel point, Indicates rounding up, rand(0,1) indicates a random number generated between 0 and 1.

Images