Three-dimensional digital analogue simulation system based on industrial VR

Abstract

The invention relates to the technical field of analogue simulation, in particular to a three-dimensional digital analogue simulation system based on industrial VR. The system comprises a simulation judgment unit, a deviation defect unit, a shape comparison unit and a data screening unit. A texture defect module receives data in a pixel matrix module and a sequence label industrial hexagonal nut surface reconstructed in an angle deviation module, and extracts texture features according to extracted pixel values in the pixel matrix module, a calculated mean value of the pixel values and a calculated standard deviation of the pixel values. And mapping the extracted texture features to the reconstructed industrial hexagonal nut surface with sequential labels to form a reconstructed industrial hexagonal nut surface containing the texture features, judging whether the simulated industrial hexagonal nut surface has texture defects or not according to the extracted texture features, capturing surface fine texture changes through the texture features, and determining whether the texture defects exist in the industrial hexagonal nut surface or not. And tiny defects can be accurately identified, so that the detection precision is improved, and the surface quality of each nut is ensured to meet the standard.

Description

Technical Field

[0001] The present invention relates to the field of simulation technology, and in particular to a three-dimensional digital simulation system based on industrial VR. Background Art

[0002] With the development of technology and the deepening of its application, virtual reality (VR) technology will play an increasingly important role in the optimization of industrial production processes. For example, the three-dimensional digital simulation of industrial VR can achieve detailed simulation of the production process of industrial hexagonal nuts through virtual reality (VR) technology and simulation. When simulating the industrial hexagonal nuts produced, due to the size, angle and surface texture defects of the actual industrial hexagonal nuts produced, the size, angle and surface texture defects of the actual industrial hexagonal nuts can be found through general image comparison and image processing. It is impossible to accurately capture the subtle texture changes and tiny defects of the surface of the simulated industrial hexagonal nuts, resulting in reduced accuracy of image comparison and detection. Therefore, we provide a three-dimensional digital simulation system based on industrial VR. Summary of the Invention

[0003] The purpose of the present invention is to provide a three-dimensional digital simulation system based on industrial VR to solve the problems raised in the above background technology.

[0004] To achieve the above-mentioned purpose, the present invention provides a three-dimensional digital simulation system based on industrial VR, which includes a simulation judgment unit, a deviation defect unit, a shape comparison unit, and a data screening unit;

[0005] The simulation judgment unit uses a camera device to obtain real-time flow chart image data of the actual production of industrial hexagonal nuts for simulation, extracts the boundary lines of the simulated industrial hexagonal nuts based on the simulated industrial hexagonal nut images, and then uses the extracted boundary lines to compare and judge whether there is a size deviation in the simulated industrial hexagonal nuts;

[0006] The deviation defect unit is used to receive data from the simulation judgment unit, obtain the extracted boundary line from the simulation judgment unit to reconstruct the industrial hexagonal nut surface, use the reconstructed industrial hexagonal nut surface to compare and judge whether the simulated industrial hexagonal nut has an angular deviation, extract texture features based on the data in the simulation judgment unit, and then map them to the reconstructed sequentially labeled industrial hexagonal nut surface to form a reconstructed industrial hexagonal nut surface containing texture features, and at the same time judge whether the simulated industrial hexagonal nut surface has texture defects based on the extracted texture features;

[0007] The shape comparison unit is used to receive data from the deviation defect unit, obtain the reconstructed industrial hexagonal nut surface containing texture features from the deviation defect unit, and splice it into a complete industrial hexagonal nut shape, and then compare the complete industrial hexagonal nut shape with the standard industrial hexagonal nut shape;

[0008] The data screening unit is used to receive data from the shape comparison unit or the deviation defect unit or the simulation judgment unit to screen deviations and defects in the industrial hexagonal nuts actually produced.

[0009] As a further improvement of the present technical solution, the simulation judgment unit includes a simulation module, a boundary line extraction module and a size deviation module;

[0010] The simulation module uses a camera to obtain the flow chart image data of the actual production of industrial hexagonal nuts in real time, and obtains historical data, and then uses computer vision technology and image processing algorithms to generate a digital production three-dimensional model based on the obtained image data and historical data. When the digital production three-dimensional model receives the start of production of industrial hexagonal nuts, the digital production three-dimensional model is triggered, and the digital production three-dimensional model sends an instruction to the camera. The camera will obtain the image data of the industrial hexagonal nuts actually produced and the corresponding coordinate data are fed back to the digital production three-dimensional model in real time. The digital production three-dimensional model simulates the fed-back industrial hexagonal nut image data and simulates the industrial hexagonal nut image. The digital three-dimensional model can accurately simulate the actual appearance and size of the industrial hexagonal nuts, ensure high-precision detection, and accurately identify dimensional deviations, shape errors and other subtle manufacturing defects in the image;

[0011] Historical data includes standard boundary lines for sequentially numbered hexagonal nuts, standard faces for sequentially numbered hexagonal nuts, and standard industrial hexagonal nut shapes;

[0012] The boundary line extraction module is used to receive the simulated industrial hexagonal nut image and corresponding coordinate data in the simulation module, use the Sobel operator to calculate the horizontal and vertical gradients based on the simulated industrial hexagonal nut image, calculate the gradient direction based on the calculated horizontal and vertical gradients, and then use the Hough space method to extract the boundary line of the simulated industrial hexagonal nut based on the calculated gradient direction and the corresponding coordinate data. By calculating the gradient direction, subtle changes in the boundary line can be captured, thereby accurately identifying dimensional deviations, and then the extracted boundary lines are sequentially numbered (referring to the different boundary lines of the simulated industrial hexagonal nut being correspondingly numbered 1, 2, 3...);

[0013] The implementation principle of calculating horizontal and vertical gradients using the Sobel operator:

[0014]

[0015] The implementation principle of calculating the gradient direction:

[0016] Collect the calculated horizontal gradient G x and vertical gradient G y Data and calculate the gradient direction to get the calculated gradient direction θ. The specific algorithm formula is:

[0017]

[0018] This formula is used to calculate the gradient direction. By calculating the horizontal and vertical gradients, the edge position in the image can be accurately detected. Because the edge is usually where the gradient changes the most, the accuracy of edge detection is improved, and the outline and boundary of the object can be accurately identified.

[0019] The size deviation module is used to receive the historical data in the simulation module and the sequential number boundary line extracted in the boundary line extraction module, obtain the standard boundary line of the sequential number from the historical data, and compare the extracted sequential number boundary line with the standard boundary line of the sequential number. According to the comparison results, it is judged whether there is a size deviation in the simulated industrial hexagonal nut. By comparing the extracted boundary line with the standard boundary line, the deviation between the actual size of the hexagonal nut and the standard size can be accurately measured. The specific comparison judgment situation is:

[0020] Case 1: When the extracted sequence number boundary line is consistent with the standard sequence number boundary line, it means that the simulated industrial hexagonal nut has no size deviation, which reflects that the actual industrial hexagonal nut has no size deviation. The command of no size deviation is transmitted to the deviation defect unit;

[0021] Case ②: When the extracted sequence number boundary line is inconsistent with the standard sequence number boundary line, it indicates that the simulated industrial hexagonal nut has a size deviation, which reflects that the industrial hexagonal nut actually produced has a size deviation, and the size deviation command is passed to the data screening unit.

[0022] As a further improvement of this technical solution, the size deviation module uses the Hough space method to extract the boundary line of the simulated industrial hexagonal nut. The implementation principle is:

[0023] Get the calculated gradient direction θ and the corresponding coordinate data Z b =(x,y) and extract the simulated industrial hexagonal nut boundary line ρ. The specific algorithm formula is:

[0024] ρ=xcosθ+ysinθ;

[0025] This formula is used to extract the boundary lines of simulated industrial hexagonal nuts. By extracting the boundary lines, subtle changes in the boundary lines can be captured, thereby accurately identifying dimensional deviations. Small dimensional errors can be discovered to ensure that the size of the nut meets the specification requirements.

[0026] As a further improvement of the present technical solution, the deviation defect unit includes an angle deviation module, a pixel matrix module and a texture defect module;

[0027] The angle deviation module is used to receive the command that there is no size deviation in the size deviation module. The angle deviation module obtains the extracted sequentially labeled boundary lines and historical data in the simulation module from the size deviation module, and uses the Graham scanning algorithm to connect the extracted sequentially labeled boundary lines into closed contours. These contours represent the external boundaries of each side of the hexagonal nut. The contour reconstruction algorithm is used to reconstruct each face of the industrial hexagonal nut according to the closed contour, and the reconstructed industrial hexagonal nut face is sequentially labeled. The reconstructed sequentially labeled industrial hexagonal nut face is then compared with the sequentially labeled hexagonal nut standard face. The compared face results are used to determine whether there is an angle deviation in the simulated industrial hexagonal nut. By comparing the angle between the reconstructed hexagonal nut face and the standard face, the deviation between the actual angle of each face and the standard angle can be accurately measured. The specific comparison judgment situation is:

[0028] Case 1: When the reconstructed serial numbered industrial hexagonal nut surface is consistent with the serial numbered standard hexagonal nut surface, it indicates that the simulated industrial hexagonal nut has no angle deviation, which reflects that the actual industrial hexagonal nut has no angle deviation. The command indicating no angle deviation is transmitted to the pixel matrix module;

[0029] Case 2: When the reconstructed serial numbered industrial hexagonal nut surface is inconsistent with the serial numbered standard hexagonal nut surface, it indicates that the simulated industrial hexagonal nut has an angle deviation, which reflects that the actual industrial hexagonal nut has an angle deviation. The angle deviation command is transmitted to the data screening unit;

[0030] The pixel matrix module is used to receive the command that there is no angle deviation in the angle deviation module. The pixel matrix module obtains the image data of the industrial hexagonal nut actually produced from the simulation module, and then extracts the pixel values ​​from the image data of the industrial hexagonal nut actually produced. The grayscale co-occurrence matrix is ​​used to form a co-occurrence matrix according to the extracted pixel values, and the mean and standard deviation of the pixel values ​​are calculated according to the extracted pixel values. The co-occurrence matrix is ​​normalized according to the formed co-occurrence matrix and the extracted pixel values ​​by normalization so that the sum is 1, so as to be used for calculating texture features;

[0031] The principle of using normalization to realize the normalized co-occurrence matrix:

[0032] Collect the formed co-occurrence matrix p(i,j) and the extracted pixel value (i,j) and perform normalized co-occurrence matrix to obtain the normalized co-occurrence matrix p′(i,j). The specific algorithm formula is:

[0033]

[0034] The texture defect module is used to receive the normalized co-occurrence matrix, extracted pixel values, formed co-occurrence matrix, calculated pixel value mean and calculated pixel value standard deviation in the pixel matrix module, receive the reconstructed sequentially labeled industrial hexagonal nut surface in the angle deviation module, extract texture features based on the normalized co-occurrence matrix, extracted pixel values, formed co-occurrence matrix, calculated pixel value mean and calculated pixel value standard deviation, map the extracted texture features to the reconstructed sequentially labeled industrial hexagonal nut surface to form a reconstructed industrial hexagonal nut surface containing texture features, and use texture mapping technology to extract texture features including contrast, energy and correlation texture features, and judge whether the simulated industrial hexagonal nut surface has texture defects based on the extracted contrast, energy and correlation texture features. Contrast, energy and correlation can respectively reflect the contrast, uniformity and spatial relationship of the texture, helping to identify subtle defects.

[0035] As a further improvement of this technical solution, the steps for extracting texture features in the texture defect module are as follows:

[0036] Step ①: Collect the extracted pixel values ​​(i, j) and the normalized co-occurrence matrix p′(i, j) and calculate the texture contrast to obtain the calculated texture contrast Contrast. The specific algorithm formula is:

[0037] Contrast=∑ i,j (ij) 2 p′(i,j);

[0038] This formula is used to calculate texture contrast, which measures the brightness difference of the texture. Normal texture has consistent contrast, while defects (such as scratches, pits, and surface damage) will cause abnormal changes in contrast.

[0039] Step ②, calculate the texture correlation based on the extracted pixel value (i, j), the calculated pixel value mean μ, the calculated pixel value standard deviation σ and the normalized co-occurrence matrix p′(i, j) to obtain the calculated texture correlation Correlation. The specific algorithm formula is:

[0040]

[0041] Among them, μ i represents the average value of pixel value i in the image data of industrial hexagonal nuts actually produced, μj represents the average value of pixel value j in the image data of industrial hexagonal nuts actually produced, σ i represents the standard deviation of pixel value i, σ j Represents the standard deviation of pixel value j. This formula is used to calculate texture correlation, which measures the spatial relationship between texture elements. By calculating the correlation, inconsistencies in texture patterns, such as local defects or irregular distribution of textures, can be discovered.

[0042] Step 3: Calculate the texture energy based on the normalized co-occurrence matrix p′(i,j) to obtain the calculated texture energy Energy. The specific algorithm formula is:

[0043] Energy=∑ i,j [p′(i,j)] 2 ;

[0044] This formula is used to calculate texture energy, which measures the overall strength and contrast of the texture. By calculating texture energy, areas with abnormal texture intensity can be identified, which are often signs of defects such as scratches, dents or uneven surfaces.

[0045] As a further improvement of this technical solution, the specific judgment conditions in the texture defect module include:

[0046] Case 1: When it is determined that a texture defect exists in one of the extracted contrast, energy, and correlation texture features, it indicates that a texture defect exists on the surface of the simulated industrial hexagonal nut, and a texture defect existence command is transmitted to the data screening unit;

[0047] S1. If the texture contrast calculated from the extracted texture features is greater than the standard texture contrast, it indicates that the simulated industrial hexagonal nut surface has defects such as unevenness and scratches, which reflects that the actual industrial hexagonal nuts produced have defects such as unevenness and scratches;

[0048] If the texture contrast calculated in the extracted texture features is less than the standard texture contrast, it indicates that the simulated industrial hexagonal nut surface has wear defects, which reflects that the actual industrial hexagonal nuts produced have wear defects;

[0049] S2. If the texture correlation calculated from the extracted texture features is less than the standard texture correlation, it indicates that the simulated industrial hexagonal nut surface has a concave or convex defect, thereby reflecting that the actual industrial hexagonal nut production has a concave or convex defect;

[0050] If the texture correlation calculated from the extracted texture features is greater than the standard texture correlation, it indicates that the simulated industrial hexagonal nut surface has no defects, thus reflecting that the actually produced industrial hexagonal nuts have no defects.

[0051] S3. If the texture energy calculated from the extracted texture features is greater than the standard texture energy, it indicates that the simulated industrial hexagonal nut surface has no defects, thereby reflecting that the actually produced industrial hexagonal nuts have no defects;

[0052] If the texture energy calculated from the extracted texture features is less than the standard texture energy, it indicates that the simulated industrial hexagonal nut surface has uneven texture, which reflects that the actual industrial hexagonal nut production also has uneven texture.

[0053] Case 2: When it is determined that there are no texture defects in the extracted contrast, energy and correlation texture features, it means that there are no texture defects on the simulated industrial hexagonal nut surface, and a command that there is no texture defect is passed to the shape comparison unit.

[0054] As a further improvement of the present technical solution, the shape comparison unit is used to receive a command indicating that there is no texture defect in the texture defect module. The shape comparison unit obtains the reconstructed industrial hexagonal nut surface containing texture features and historical data in the simulation module from the angle deviation module, and uses grid splicing technology to splice the reconstructed industrial hexagonal nut surface containing texture features into a complete industrial hexagonal nut shape. The complete industrial hexagonal nut shape is then accurately compared with the standard industrial hexagonal nut shape. When the complete industrial hexagonal nut shape is consistent with the standard industrial hexagonal nut shape, the comparison result is visualized through VR. When the complete industrial hexagonal nut shape is inconsistent with the standard industrial hexagonal nut shape, the inconsistent comparison command is passed to the data screening unit.

[0055] As a further improvement of the present technical solution, the data screening unit is used to receive a command of inconsistent comparison in the shape comparison unit or a command of texture defect in the texture defect module or an angle deviation command in the angle deviation module or a size deviation command in the size deviation module. The data screening unit obtains the corresponding coordinate data from the simulation module, and performs deviation and defect screening on the industrial hexagonal nuts actually produced according to the corresponding coordinate data. Each important dimension and geometric feature of the hexagonal nut can be accurately measured, thereby effectively identifying deviations and defects in the production process.

[0056] Compared with the prior art, the present invention has the following beneficial effects:

[0057] 1. In the three-dimensional digital simulation system of the industrial VR, the texture defect module is used to receive the normalized co-occurrence matrix, the extracted pixel values, the formed co-occurrence matrix, the calculated pixel value mean and the calculated pixel value standard deviation in the pixel matrix module, and receive the reconstructed sequentially labeled industrial hexagonal nut surface in the angle deviation module. Texture features are extracted according to the normalized co-occurrence matrix, the extracted pixel values, the formed co-occurrence matrix, the calculated pixel value mean and the calculated pixel value standard deviation. The extracted texture features are mapped to the reconstructed sequentially labeled industrial hexagonal nut surface to form a reconstructed industrial hexagonal nut surface containing texture features. The extracted texture features are used to determine whether the simulated industrial hexagonal nut surface has texture defects. The texture features can capture subtle texture changes on the surface and accurately identify tiny defects, thereby improving the detection accuracy and ensuring that the surface quality of each nut meets the standards.

[0058] 2. In the three-dimensional digital simulation system of the industrial VR, the angle deviation module is used to receive the command that there is no size deviation in the size deviation module. The angle deviation module obtains the extracted sequential number boundary lines and historical data in the simulation module from the size deviation module, connects the extracted sequential number boundary lines into a closed contour, and reconstructs each face of the industrial hexagonal nut according to the closed contour. The reconstructed industrial hexagonal nut face is sequentially numbered, and then the reconstructed sequentially numbered industrial hexagonal nut face is compared with the sequentially numbered hexagonal nut standard face. The simulated industrial hexagonal nut is judged whether there is an angle deviation based on the compared face results. By comparing the actual reconstructed hexagonal nut face with the standard face, the angle of each nut can be accurately calibrated to ensure that the produced nuts meet the design and standard requirements and avoid the angle deviation affecting the product performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 It is an overall block diagram of the present invention;

[0060] Figure 2 This is a block diagram of the simulation judgment unit of the present invention;

[0061] Figure 3 This is a block diagram of the deviation defect unit of the present invention.

[0062] The meaning of each number in the figure is:

[0063] 1. Simulation judgment unit; 11. Simulation module; 12. Boundary line extraction module; 13. Dimension deviation module;

[0064] 2. Deviation defect unit; 21. Angle deviation module; 22. Pixel matrix module; 23. Texture defect module;

[0065] 3. Shape comparison unit; 4. Data screening unit. DETAILED DESCRIPTION

[0066] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts shall fall within the scope of protection of the present invention.

[0067] Example 1

[0068] The present invention provides a three-dimensional digital simulation system based on industrial VR, please refer to Figure 1-Figure 3 , including a simulation judgment unit 1, a deviation defect unit 2, a shape comparison unit 3, and a data screening unit 4;

[0069] The simulation judgment unit 1 uses a camera device to obtain real-time flow chart image data of the actual production of industrial hexagonal nuts for simulation, extracts the boundary lines of the simulated industrial hexagonal nuts based on the simulated industrial hexagonal nut images, and then uses the extracted boundary lines to compare and judge whether there is a size deviation in the simulated industrial hexagonal nuts;

[0070] The deviation defect unit 2 is used to receive the data in the simulation judgment unit 1. The deviation defect unit 2 obtains the extracted boundary line from the simulation judgment unit 1 to reconstruct the industrial hexagonal nut surface, uses the reconstructed industrial hexagonal nut surface to compare and judge whether the simulated industrial hexagonal nut has an angle deviation, extracts texture features based on the data in the simulation judgment unit 1, and then maps them to the reconstructed sequentially labeled industrial hexagonal nut surface to form a reconstructed industrial hexagonal nut surface containing texture features. At the same time, the extracted texture features are used to judge whether the simulated industrial hexagonal nut surface has texture defects;

[0071] The shape comparison unit 3 is used to receive data from the deviation defect unit 2. The shape comparison unit 3 obtains the reconstructed industrial hexagonal nut surface containing texture features from the deviation defect unit 2 and splices it into a complete industrial hexagonal nut shape. The complete industrial hexagonal nut shape is then compared with the standard industrial hexagonal nut shape.

[0072] The data screening unit 4 is used to receive the data in the shape comparison unit 3 or the data in the deviation and defect unit 2 or the data in the simulation judgment unit 1 to screen the deviations and defects of the industrial hexagonal nuts actually produced.

[0073] The following is a refinement of the above units, see Figure 1-Figure 3 ;

[0074] The simulation judgment unit 1 includes a simulation module 11, a boundary line extraction module 12 and a size deviation module 13;

[0075] The simulation module 11 uses a camera device to obtain real-time flow chart image data of the actual production of industrial hexagonal nuts and obtains historical data, and then uses computer vision technology and image processing algorithms to generate a digital production three-dimensional model based on the obtained image data and historical data. When the digital production three-dimensional model receives the start of production of industrial hexagonal nuts, the digital production three-dimensional model is triggered, and the digital production three-dimensional model sends an instruction to the camera device. The camera device obtains the real-time image data of the industrial hexagonal nuts produced and the corresponding coordinate data and feeds it back to the digital production three-dimensional model. The digital production three-dimensional model simulates the fed-back industrial hexagonal nut image data, and simulates the industrial hexagonal nut image. The digital three-dimensional model can accurately simulate the actual appearance and size of the industrial hexagonal nuts, ensure high-precision detection, and accurately identify dimensional deviations, shape errors and other subtle manufacturing defects in the image;

[0076] Historical data includes standard boundary lines for sequentially numbered hexagonal nuts, standard faces for sequentially numbered hexagonal nuts, and standard industrial hexagonal nut shapes;

[0077] The boundary line extraction module 12 is used to receive the industrial hexagonal nut image simulated in the simulation module 11 and the corresponding coordinate data, calculate the horizontal and vertical gradients based on the simulated industrial hexagonal nut image using the Sobel operator, calculate the gradient direction based on the calculated horizontal and vertical gradients, and then use the Hough space method to extract the boundary line of the simulated industrial hexagonal nut based on the calculated gradient direction and the corresponding coordinate data. By calculating the gradient direction, subtle changes in the boundary line can be captured, thereby accurately identifying dimensional deviations, and then sequentially numbering the extracted boundary lines (referring to the different boundary lines of the simulated industrial hexagonal nut being correspondingly numbered 1, 2, 3...);

[0078] The implementation principle of calculating horizontal and vertical gradients using the Sobel operator:

[0079]

[0080] The implementation principle of calculating the gradient direction:

[0081] Collect the calculated horizontal gradient G x and vertical gradient G y Data and calculate the gradient direction to get the calculated gradient direction θ. The specific algorithm formula is:

[0082]

[0083] This formula is used to calculate the gradient direction. By calculating the horizontal and vertical gradients, the edge position in the image can be accurately detected. Because the edge is usually where the gradient changes the most, the accuracy of edge detection is improved, and the outline and boundary of the object can be accurately identified.

[0084] The principle of using Hough space method to extract the boundary line of industrial hexagonal nuts for simulation:

[0085] Get the calculated gradient direction θ and the corresponding coordinate data Z b =(x,y) and extract the simulated industrial hexagonal nut boundary line ρ. The specific algorithm formula is:

[0086] ρ=xcosθ+ysinθ;

[0087] This formula is used to extract the boundary lines of simulated industrial hexagonal nuts. By extracting the boundary lines, subtle changes in the boundary lines can be captured, thereby accurately identifying dimensional deviations. It can detect small dimensional errors and ensure that the nut size meets the specification requirements.

[0088] The size deviation module 13 is used to receive the historical data in the simulation module 11 and the sequentially labeled boundary lines extracted by the boundary line extraction module 12, obtain the standard boundary lines of the sequential labels from the historical data, and compare the extracted sequentially labeled boundary lines with the standard boundary lines of the sequential labels. According to the comparison results, it is judged whether there is a size deviation in the simulated industrial hexagonal nut. By comparing the extracted boundary lines with the standard boundary lines, the deviation between the actual size of the hexagonal nut and the standard size can be accurately measured. The specific comparison judgment situation is:

[0089] Case 1: When the extracted sequence number boundary line is consistent with the standard sequence number boundary line, it indicates that the simulated industrial hexagonal nut has no size deviation, which reflects that the actual industrial hexagonal nut has no size deviation. The command indicating that there is no size deviation is transmitted to the deviation defect unit 2;

[0090] Case 2: When the extracted sequence number boundary line is inconsistent with the standard sequence number boundary line, it indicates that the simulated industrial hexagonal nut has a size deviation, which reflects that the actual industrial hexagonal nut has a size deviation, and the size deviation command is transmitted to the data screening unit 4;

[0091] The deviation defect unit 2 includes an angle deviation module 21, a pixel matrix module 22 and a texture defect module 23;

[0092] The angle deviation module 21 is used to receive a command from the size deviation module 13 that there is no size deviation. The angle deviation module 21 obtains the extracted sequentially labeled boundary lines and the historical data in the simulation module 11 from the size deviation module 13, and uses the Graham scanning algorithm to connect the extracted sequentially labeled boundary lines into closed contours. These contours represent the external boundaries of each side of the hexagonal nut. The contour reconstruction algorithm is used to reconstruct each face of the industrial hexagonal nut according to the closed contours, and the reconstructed industrial hexagonal nut face is sequentially labeled. The reconstructed sequentially labeled industrial hexagonal nut face is then compared with the sequentially labeled hexagonal nut standard face. The compared face results are used to determine whether the simulated industrial hexagonal nut has an angle deviation. By comparing the angles of the reconstructed hexagonal nut face and the standard face, the deviation between the actual angle of each face and the standard angle can be accurately measured. The specific comparison and judgment situation is:

[0093] Case 1: When the reconstructed serial numbered industrial hexagonal nut surface is consistent with the serial numbered standard hexagonal nut surface, it indicates that the simulated industrial hexagonal nut has no angle deviation, which reflects that the actual industrial hexagonal nut has no angle deviation. The command indicating that there is no angle deviation is transmitted to the pixel matrix module 22;

[0094] Case 2: When the reconstructed serial numbered industrial hexagonal nut surface is inconsistent with the serial numbered standard hexagonal nut surface, it indicates that the simulated industrial hexagonal nut has an angle deviation, which reflects that the actual industrial hexagonal nut has an angle deviation, and the angle deviation command is transmitted to the data screening unit 4;

[0095] The pixel matrix module 22 is used to receive a command from the angle deviation module 21 indicating that there is no angle deviation. The pixel matrix module 22 obtains the image data of the industrial hexagonal nut actually produced from the simulation module 11, extracts pixel values ​​from the image data of the industrial hexagonal nut actually produced, forms a co-occurrence matrix based on the extracted pixel values ​​using a grayscale co-occurrence matrix, calculates the mean and standard deviation of the pixel values ​​based on the extracted pixel values, and then uses normalization to normalize the co-occurrence matrix based on the formed co-occurrence matrix and the extracted pixel values ​​so that the sum is 1, so as to be used for calculating texture features;

[0096] The principle of using normalization to realize the normalized co-occurrence matrix:

[0097] Collect the formed co-occurrence matrix p(i, j) and the extracted pixel value (i, j) and perform normalized co-occurrence matrix to obtain the normalized co-occurrence matrix p′(i, j). The specific algorithm formula is:

[0098]

[0099] The texture defect module 23 is used to receive the normalized co-occurrence matrix, the extracted pixel values, the formed co-occurrence matrix, the calculated pixel value mean and the calculated pixel value standard deviation in the pixel matrix module 22, and receive the sequentially labeled industrial hexagonal nut surface reconstructed in the angle deviation module 21, extract texture features according to the normalized co-occurrence matrix, the extracted pixel values, the formed co-occurrence matrix, the calculated pixel value mean and the calculated pixel value standard deviation, map the extracted texture features to the reconstructed sequentially labeled industrial hexagonal nut surface to form a reconstructed industrial hexagonal nut surface containing texture features, and use texture mapping technology to extract texture features including contrast, energy and correlation texture features, and judge whether the simulated industrial hexagonal nut surface has texture defects based on the extracted contrast, energy and correlation texture features. Contrast, energy and correlation can respectively reflect the contrast, uniformity and spatial relationship of the texture, and help identify subtle defects;

[0100] Implementation steps for extracting texture features:

[0101] Step ①: Collect the extracted pixel values ​​(i, j) and the normalized co-occurrence matrix p′(i, j) and calculate the texture contrast to obtain the calculated texture contrast Contrast. The specific algorithm formula is:

[0102] Contrast=∑ i,j (ij) 2 p′(i,j);

[0103] This formula is used to calculate texture contrast, which measures the brightness difference of the texture. Normal texture has consistent contrast, while defects (such as scratches, pits, and surface damage) will cause abnormal changes in contrast.

[0104] Step ②, calculate the texture correlation based on the extracted pixel value (i, j), the calculated pixel value mean μ, the calculated pixel value standard deviation σ and the normalized co-occurrence matrix p′(i, j) to obtain the calculated texture correlation Correlation. The specific algorithm formula is:

[0105]

[0106] Among them, μ i represents the average value of pixel value i in the image data of industrial hexagonal nuts actually produced, μ j represents the average value of pixel value j in the image data of industrial hexagonal nuts actually produced, σ i represents the standard deviation of pixel value i, σ jRepresents the standard deviation of pixel value j. This formula is used to calculate texture correlation, which measures the spatial relationship between texture elements. By calculating the correlation, inconsistencies in texture patterns, such as local defects or irregular distribution of textures, can be discovered.

[0107] Step 3: Calculate the texture energy based on the normalized co-occurrence matrix p′(i,j) to obtain the calculated texture energy Energy. The specific algorithm formula is:

[0108] Energy=∑ i,j [p′(i,j)] 2 ;

[0109] This formula is used to calculate texture energy, which measures the overall strength and contrast of the texture. By calculating texture energy, areas with abnormal texture intensity can be identified, which are often signs of defects such as scratches, dents or uneven surfaces.

[0110] Specific judgment situations include:

[0111] Case 1: When it is determined that a texture defect exists in one of the extracted contrast, energy, and correlation texture features, it indicates that a texture defect exists on the surface of the simulated industrial hexagonal nut, and a texture defect existence command is transmitted to the data screening unit 4;

[0112] S1. If the texture contrast calculated from the extracted texture features is greater than the standard texture contrast, it indicates that the simulated industrial hexagonal nut surface has defects such as unevenness and scratches, which reflects that the actual industrial hexagonal nuts produced have defects such as unevenness and scratches;

[0113] If the texture contrast calculated in the extracted texture features is less than the standard texture contrast, it indicates that the simulated industrial hexagonal nut surface has wear defects, which reflects that the actual industrial hexagonal nuts produced have wear defects;

[0114] S2. If the texture correlation calculated from the extracted texture features is less than the standard texture correlation, it indicates that the simulated industrial hexagonal nut surface has a concave or convex defect, thereby reflecting that the actual industrial hexagonal nut production has a concave or convex defect;

[0115] If the texture correlation calculated from the extracted texture features is greater than the standard texture correlation, it indicates that the simulated industrial hexagonal nut surface has no defects, thus reflecting that the actually produced industrial hexagonal nuts have no defects.

[0116] S3. If the texture energy calculated from the extracted texture features is greater than the standard texture energy, it indicates that the simulated industrial hexagonal nut surface has no defects, thereby reflecting that the actually produced industrial hexagonal nuts have no defects;

[0117] If the texture energy calculated from the extracted texture features is less than the standard texture energy, it indicates that the simulated industrial hexagonal nut surface has uneven texture, which reflects that the actual industrial hexagonal nut production also has uneven texture.

[0118] Case 2: When it is determined that there are no texture defects in the extracted contrast, energy, and correlation texture features, it means that there are no texture defects on the simulated industrial hexagonal nut surface, and a texture defect-free command is transmitted to the shape comparison unit 3;

[0119] The shape comparison unit 3 is used to receive a command from the texture defect module 23 indicating that there is no texture defect. The shape comparison unit 3 obtains the reconstructed industrial hexagonal nut surface containing texture features and the historical data in the simulation module 11 from the angle deviation module 21, and uses the grid splicing technology to splice the reconstructed industrial hexagonal nut surface containing texture features into a complete industrial hexagonal nut shape. The complete industrial hexagonal nut shape is then accurately compared with the standard industrial hexagonal nut shape. When the complete industrial hexagonal nut shape is consistent with the standard industrial hexagonal nut shape, the comparison result is visualized through VR. When the complete industrial hexagonal nut shape is inconsistent with the standard industrial hexagonal nut shape, the inconsistent comparison command is passed to the data screening unit 4.

[0120] The data screening unit 4 is used to receive the command of inconsistent comparison from the shape comparison unit 3 or the command of texture defect existence from the texture defect module 23 or the command of angle deviation existence from the angle deviation module 21 or the command of size deviation existence from the size deviation module 13. The data screening unit 4 obtains the corresponding coordinate data from the simulation module 11, and performs deviation and defect screening on the industrial hexagonal nuts actually produced according to the corresponding coordinate data. It can accurately measure each important dimension and geometric feature of the hexagonal nuts, thereby effectively identifying deviations and defects in the production process.

[0121] Usage process:

[0122] The size deviation module 13 is used to receive the historical data in the simulation module 11 and the sequentially labeled boundary lines extracted in the boundary line extraction module 12, and use the extracted sequentially labeled boundary lines to compare and judge whether there is a size deviation in the simulated industrial hexagonal nut, and pass the command that there is no size deviation to the angle deviation module 21. The angle deviation module 21 is used to receive the command that there is no size deviation in the size deviation module 13. The angle deviation module 21 obtains the extracted sequentially labeled boundary lines and the historical data in the simulation module 11 from the size deviation module 13, uses the Graham scanning algorithm to connect the extracted sequentially labeled boundary lines into a closed contour, and then uses the contour reconstruction algorithm to reconstruct each face of the industrial hexagonal nut according to the closed contour, and then uses the reconstructed sequentially labeled industrial hexagonal nut face comparison to judge whether the simulated industrial hexagonal nut has a size deviation. Whether the nut has an angle deviation, the command that there is no angle deviation is passed to the pixel matrix module 22. The pixel matrix module 22 is used to receive the command that there is no angle deviation from the angle deviation module 21. The pixel matrix module 22 obtains the image data of the industrial hexagonal nut actually produced from the simulation module 11 and then extracts the pixel value, and passes the extracted pixel value to the texture defect module 23. The texture defect module 23 is used to receive the pixel value extracted from the pixel matrix module 22 and the sequentially labeled industrial hexagonal nut surface reconstructed in the angle deviation module 21, extract texture features according to the extracted pixel values, map the extracted texture features to the reconstructed sequentially labeled industrial hexagonal nut surface, and form a reconstructed industrial hexagonal nut surface containing texture features. Then, the extracted texture features are used to judge whether the simulated industrial hexagonal nut surface has texture defects.

[0123] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Various changes and improvements may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and improvements fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. The three-dimensional digital simulation system based on industrial VR is characterized by: It includes a simulation judgment unit (1), a deviation defect unit (2), a shape comparison unit (3), and a data screening unit (4); The simulation judgment unit (1) uses a camera device to obtain real-time flow chart image data of the actual production of industrial hexagonal nuts for simulation, extracts the boundary lines of the simulated industrial hexagonal nuts based on the simulated industrial hexagonal nut images, and then uses the extracted boundary lines to compare and judge whether the simulated industrial hexagonal nuts have size deviations; The deviation defect unit (2) is used to receive data from the simulation judgment unit (1). The deviation defect unit (2) obtains the extracted boundary line from the simulation judgment unit (1) to reconstruct the industrial hexagonal nut surface, uses the reconstructed industrial hexagonal nut surface to compare and judge whether the simulated industrial hexagonal nut has an angle deviation, extracts texture features based on the data in the simulation judgment unit (1), and then maps them to the reconstructed sequentially labeled industrial hexagonal nut surface to form a reconstructed industrial hexagonal nut surface containing texture features, and at the same time judges whether the simulated industrial hexagonal nut surface has texture defects based on the extracted texture features. The shape comparison unit (3) is used to receive data from the deviation defect unit (2). The shape comparison unit (3) obtains the reconstructed industrial hexagonal nut surface containing texture features from the deviation defect unit (2) and splices it into a complete industrial hexagonal nut shape, and then compares the complete industrial hexagonal nut shape with the standard industrial hexagonal nut shape. The data screening unit (4) is used to receive data from the shape comparison unit (3) or the deviation defect unit (2) or the simulation judgment unit (1) to screen deviations and defects in industrial hexagonal nuts actually produced.

2. The three-dimensional digital simulation system of industrial VR according to claim 1 is characterized in that: The simulation judgment unit (1) includes a simulation module (11), a boundary line extraction module (12) and a size deviation module (13); The simulation module (11) uses a camera device to obtain the flow chart image data of the actual production of industrial hexagonal nuts in real time to generate a digital production three-dimensional model. When the digital production three-dimensional model receives the start of production of industrial hexagonal nuts, the digital production three-dimensional model is triggered. The digital production three-dimensional model sends a size deviation module command to the camera device. The camera device obtains the image data of the industrial hexagonal nuts actually produced in real time and the corresponding coordinate data and feeds them back to the digital production three-dimensional model. The digital production three-dimensional model simulates the fed-back industrial hexagonal nut image data. The boundary line extraction module (12) is used to receive the industrial hexagonal nut image simulated in the simulation module (11) and the corresponding coordinate data, calculate the horizontal and vertical gradients based on the simulated industrial hexagonal nut image using the Sobel operator, calculate the gradient direction based on the calculated horizontal and vertical gradients, and then use the Hough space method to extract the boundary line of the simulated industrial hexagonal nut based on the calculated gradient direction and the corresponding coordinate data, and then sequentially label the extracted boundary lines; The size deviation module (13) is used to receive the historical data in the simulation module (11) and the sequentially labeled boundary lines extracted in the boundary line extraction module (12), and compare the extracted sequentially labeled boundary lines with the standard boundary lines of the sequential labels in the historical data to determine whether the simulated industrial hexagonal nut has a size deviation.

3. The three-dimensional digital simulation system of industrial VR according to claim 2 is characterized in that: The boundary line extraction module (12) uses the Hough space method to extract the boundary line of the simulated industrial hexagonal nut to achieve the following principle: Get the calculated gradient direction θ and the corresponding coordinate data Z b =(x,y) and extract the simulated industrial hexagonal nut boundary line ρ. The specific algorithm formula is: ρ=xcosθ+ysinθ.

4. The three-dimensional digital simulation system of industrial VR according to claim 2, characterized in that: The deviation defect unit (2) comprises an angle deviation module (21), a pixel matrix module (22) and a texture defect module (23); The angle deviation module (21) is used to receive a command from the size deviation module (13) indicating that there is no size deviation. The angle deviation module (21) obtains the extracted sequential number boundary line and the historical data from the simulation module (11) from the size deviation module (13), uses the Graham scanning algorithm to reconstruct each face of the industrial hexagonal nut with the extracted sequential number boundary line, sequentially labels the reconstructed industrial hexagonal nut face, and then compares the reconstructed sequential numbered industrial hexagonal nut face with the sequential numbered hexagonal nut standard face to determine whether the simulated industrial hexagonal nut has an angle deviation. The pixel matrix module (22) is used to receive a command indicating that there is no angle deviation in the angle deviation module (21). The pixel matrix module (22) obtains image data of industrial hexagonal nuts actually produced from the simulation module (11) and then extracts pixel values. A grayscale co-occurrence matrix is ​​used to form a co-occurrence matrix based on the extracted pixel values. A mean and a standard deviation of the pixel values ​​are calculated based on the extracted pixel values. The co-occurrence matrix is ​​then normalized based on the formed co-occurrence matrix and the extracted pixel values ​​using a normalization method. The texture defect module (23) is used to receive data in the pixel matrix module (22) and the reconstructed sequentially labeled industrial hexagonal nut surface in the angle deviation module (21), extract texture features based on the data in the pixel matrix module (22), map the extracted texture features onto the reconstructed sequentially labeled industrial hexagonal nut surface, and form a reconstructed industrial hexagonal nut surface containing texture features. The texture mapping technology is used, and the extracted texture features are used to determine whether the simulated industrial hexagonal nut surface has texture defects.

5. The three-dimensional digital simulation system of industrial VR according to claim 4 is characterized in that: The steps for extracting texture features in the texture defect module (23) are as follows: Step ①: Collect the extracted pixel values ​​(i, j) and the normalized co-occurrence matrix p′(i, j) and calculate the texture contrast to obtain the calculated texture contrast Contrast. The specific algorithm formula is: Contrast=∑ i,j (i-j) 2 p′(i,j); Step ②, calculate the texture correlation based on the extracted pixel value (i, j), the calculated pixel value mean μ, the calculated pixel value standard deviation σ and the normalized co-occurrence matrix p′(i, j) to obtain the calculated texture correlation Correlation. The specific algorithm formula is: Among them, μ i represents the average value of pixel value i in the image data of industrial hexagonal nuts actually produced, μ j represents the average value of pixel value j in the image data of industrial hexagonal nuts actually produced, σ i represents the standard deviation of pixel value i, σ j represents the standard deviation of pixel value j; Step 3: Calculate the texture energy based on the normalized co-occurrence matrix p′(i,j) to obtain the calculated texture energy Energy. The specific algorithm formula is: Energy=∑ i,j [p′(i,j)] 2 。 6. The three-dimensional digital simulation system of industrial VR according to claim 4 is characterized in that: The specific judgment situations in the texture defect module (23) include: Case 1: When it is determined that a texture defect exists in one of the extracted contrast, energy and correlation texture features, it indicates that a texture defect exists on the surface of the simulated industrial hexagonal nut, and a texture defect existence command is transmitted to the data screening unit (4); S1. If the texture contrast calculated from the extracted texture features is greater than the standard texture contrast, it indicates that the simulated industrial hexagonal nut surface has defects such as unevenness and scratches; If the texture contrast calculated from the extracted texture features is less than the standard texture contrast, it indicates that the simulated industrial hexagonal nut surface has wear defects; S2. If the texture correlation calculated from the extracted texture features is less than the standard texture correlation, it indicates that the simulated industrial hexagonal nut surface has a concave or convex defect; If the texture correlation calculated from the extracted texture features is greater than the standard texture correlation, it means that there are no defects on the surface of the simulated industrial hexagonal nut; S3. If the texture energy calculated from the extracted texture features is greater than the standard texture energy, it indicates that there is no defect on the surface of the simulated industrial hexagonal nut; If the texture energy calculated from the extracted texture features is less than the standard texture energy, it means that the simulated industrial hexagonal nut surface has uneven texture; Case 2: When it is determined that there are no texture defects in the extracted contrast, energy and correlation texture features, it means that there are no texture defects in the simulated industrial hexagonal nut surface, and a command indicating that there are no texture defects is passed to the shape comparison unit (3).

7. The three-dimensional digital simulation system of industrial VR according to claim 4 is characterized in that: The shape comparison unit (3) is used to receive a texture defect absence command from the texture defect module (23). The shape comparison unit (3) obtains the reconstructed industrial hexagonal nut surface containing texture features and historical data from the simulation module (11) from the angle deviation module (21), and uses a grid splicing technology to splice the reconstructed industrial hexagonal nut surface containing texture features into a complete industrial hexagonal nut shape. The complete industrial hexagonal nut shape is then accurately compared with the standard industrial hexagonal nut shape. When the complete industrial hexagonal nut shape is consistent with the standard industrial hexagonal nut shape, the comparison result is visualized through VR. When the complete industrial hexagonal nut shape is inconsistent with the standard industrial hexagonal nut shape, the inconsistent comparison command is transmitted to the data screening unit (4).

8. The three-dimensional digital simulation system of industrial VR according to claim 7, characterized in that: The data screening unit (4) is used to receive data from the shape comparison unit (3) or data from the texture defect module (23) or data from the angle deviation module (21) or data from the size deviation module (13). The data screening unit (4) obtains corresponding coordinate data from the simulation module (11) and performs deviation and defect screening on the industrial hexagonal nuts actually produced based on the corresponding coordinate data. Each important dimension and geometric feature of the hexagonal nuts can be accurately measured, thereby effectively identifying deviations and defects in the production process.

Images