Five-axis ball milling free-form surface overall texture morphology evaluation method

Abstract

The invention discloses a five-axis ball-end milling free-form surface overall texture morphology evaluation method, which comprises the following five steps: S1, designing an orthogonal experiment by taking a cutter attitude angle as a factor, establishing five-axis ball-end milling free-form surface texture morphology digital models H (i, j), projecting a height value matrix corresponding to each group of digital models to an XWOWYW surface, and establishing a five-axis ball-end milling free-form surface texture morphology model H (i, j); obtaining a Z value residual height matrix Hc (i, j); s2, extracting overall texture morphology feature indexes of the digital model of the curved surface texture morphology, and establishing a preliminary classification method of the curved surface texture morphology; s3, establishing a curved surface texture morphology subdivision method based on the Z value residual height matrix Hc (i, j), and obtaining a curved surface texture morphology subdivision result; and S4, establishing a method for determining the range of the tool attitude angle corresponding to each type of the curved surface texture morphology. According to the method, reference can be provided for limiting adjustment of the cutter attitude angle in the machining process, and the problem that the texture morphology is poor due to improper cutter attitude adjustment is solved.

Description

Technical Field

[0001] The invention relates to the technical field of numerical control milling processing, and in particular to a method for evaluating the overall texture morphology of a free-form surface during five-axis ball head milling. Background Art

[0002] Surface topography refers to the height and texture left on the surface of a workpiece after machining, and is an important indicator of the quality of the workpiece surface topography. The overall trend of the surface texture topography of complex curved high-precision parts has a significant impact on the accuracy and performance of the workpiece, and directly affects its friction, wear, contact stiffness, assembly performance, etc. How to obtain the ideal surface texture topography during milling is of great significance.

[0003] Most of the existing surface morphology characterization parameters are extracted based on partial areas of the surface morphology, and they focus on characterizing the surface morphology quality. It is difficult to reflect the overall texture morphology characteristics and trends of the surface, which are crucial to physical properties such as sealing and coordination. Most of the existing studies explore the influence of processing parameters on surface morphology to obtain the best surface morphology quality, but do not pay much attention to the overall texture morphology trend. Five-axis CNC machine tools are widely used in the processing of complex parts surfaces due to their advantages such as high degrees of freedom. However, the overall texture morphology quality control considering the influence of the tool posture of five-axis machine tools has rarely received attention. Precision maintenance strategies in five-axis milling, such as path optimization, singularity avoidance and error compensation, often improve the quality of machined surface morphology by adjusting the tool posture. However, unconstrained adjustment of the tool posture may affect the texture morphology quality of the free-form surface, especially the overall trend of the texture, resulting in irregular changes in the surface texture, affecting the overall texture morphology quality and integrity of the surface. Summary of the invention

[0004] In order to avoid the problem of poor texture morphology caused by improper tool posture adjustment, the present invention provides a method for evaluating the overall texture morphology of a free-form surface in five-axis ball head milling.

[0005] The method for evaluating the overall texture morphology of a free-form surface by five-axis ball head milling provided by the present invention comprises the following steps:

[0006] S1, tool inclination angle φ and rotation angle θ based on five-axis ball end milling, select tool inclination angle range (φ l ,φ r ) and tool rotation angle range (θ l ,θ r ) Design an orthogonal experiment, and establish a digital model H(i,j) of the surface texture morphology of the workpiece milled by a five-axis ball head based on the orthogonal experiment, where H is the height value matrix of the digital model, i is the row number of the workpiece surface grid point, and j is the column number of the workpiece surface grid point. Project the height value matrix corresponding to each group of digital models onto X W O WY W Surface, get the Z value residual height matrix Hc(i,j).

[0007] S2. Extract the overall texture morphology characteristic index of the digital model H(i,j) of the surface texture morphology of the five-axis ball head milling workpiece, and establish a preliminary classification method for the surface texture morphology. The specific method is as follows:

[0008] S21, scaling the element values ​​in the Z value residual height matrix Hc(i,j) to the grayscale interval [0,255], where 0 represents black and 255 represents white, to obtain the grayscale image matrix G(i,j) of the surface texture morphology;

[0009] S22, based on the gray image matrix G(i,j), calculate the gray co-occurrence matrix in four directions, respectively obtain the second-order statistical feature quantity of each co-occurrence matrix, and use the mean value of each feature quantity as the final evaluation index;

[0010] S23, calculating Tamura texture morphology features based on the grayscale image matrix G(i,j);

[0011] S24. Extract contour lines based on the Z-value residual height matrix Hc(i,j).

[0012] S25, extracting the overall texture morphology characteristic indexes of all groups, and establishing a preliminary classification method for surface texture morphology based on the overall texture morphology characteristic index extraction results. Specifically, extracting the overall texture morphology characteristic indexes of all groups, analyzing the index extraction results in combination with the quartile method, and obtaining a preliminary classification result for surface texture morphology.

[0013] S3. Establish a surface texture morphology subdivision method based on the Z value residual height matrix Hc(i,j) to obtain the surface texture morphology subdivision result. The specific method is as follows:

[0014] S31, taking the Z value residual height matrix Hc(i,j) data as a sample, and the Z value residual height matrix data corresponding to the digital model of the surface texture morphology of all groups in the orthogonal test table as a data set;

[0015] S32, reducing the dimension of the data set obtained in step S31 based on the principal component analysis method; based on the reduced-dimensional data set, determining the optimal number k of classifications of the data set based on the silhouette coefficient method.

[0016] S33. Classify the data set based on the Kmeans clustering algorithm to obtain the surface texture morphology segmentation results.

[0017] S4, establishing a method for determining the tool posture angle range corresponding to each type of surface texture morphology, including the following sub-steps:

[0018] S41. Set the classification accuracy a of the tilt angle and the rotation angle.

[0019] S42, using the data set established in step S31 as the initial data set, establishing a tool attitude angle combination table, and storing the tool attitude angle combinations in the orthogonal test table in the tool attitude angle combination table in sequence.

[0020] S43, starting from the first group of tool attitude angle combinations in the tool attitude angle combination table, determine each discrete attitude angle combination {φ i ,θ i}The range of θ i ∈(θ l ,θ r ), φ i ∈(φ l ,φ r ), where φ is the tilt angle, θ is the rotation angle, i is the number of the attitude angle combination in the tool attitude angle combination table, φ l is the left boundary point of the tilt angle, φ r is the right range boundary point of the tilt angle, θ l is the left boundary point of the rotation angle, θ r It is the boundary point of the right range of the rotation angle.

[0021] The specific method is as follows:

[0022] Step S431, set the right boundary point value θ r ∈(r l ,r r ), let (r l ,r r ) Upper boundary r l =θ i ; Based on the principle that all division ranges do not overlap, determine the upper boundary r r The initial value of

[0023] Step S432: Calculate the range (r l ,r r )Intermediate value r m :

[0024]

[0025] The attitude angle combination {φ i ,r m} add the tool posture angle combination table, and add its corresponding samples to the data set; according to step S3, the data set classification result is obtained. If the posture angle combination {φ i ,r m} samples and {φ i ,θ i} belongs to a class, then in the range (r l ,r m ) randomly select n attitude angle combinations {φi ,r x}, r x ∈(r l ,r m ), x = 1, 2 ... n, add n posture angle combinations to the tool posture angle combination table, and add the corresponding samples to the data set; according to step S1, n groups of surface texture morphology digital models are established based on the n posture angle combinations to obtain n groups of samples, where n satisfies the formula:

[0026]

[0027] If n groups of samples are all the same as the sample {φ i ,θ i} belongs to a class, then r l =r m ; If there is a sample r that does not belong to a class x , take r x Minimum value r xmin , r r =r xmin ;

[0028] If the attitude angle combination {φ i ,r m} corresponding to the sample and {φ i ,θ i}The corresponding sample does not belong to a class, then r r =r m ;

[0029] Repeat the above steps to continuously narrow the range (r l ,r r ) until it satisfies: r r -r l ≤a, determine the right boundary point θ r =r l ;

[0030] Step S433: Determine θ using the same method as step S432. l ,φ l ,φ r Specific value, get tool posture angle combination {φ i ,θ i}Corresponding attitude angle range θ i ∈(θ l ,θ r ), φ i ∈(φ l ,φ r );

[0031] Step S434, determining the range of the next set of tool attitude angle combinations in the tool attitude angle combination table according to steps S431-S433, until the ranges of all tool attitude angle combinations are determined.

[0032] S44. Based on the data set classification result obtained in step S3, the tool attitude angle ranges corresponding to the tool attitude angle combinations in each category are integrated to obtain the tool attitude angle ranges corresponding to each type of surface texture morphology.

[0033] Compared with the prior art, the present invention is beneficial in that:

[0034] The method of the present invention proposes an evaluation index that can evaluate the overall texture morphology characteristics and trends of the complex surface machining surface; combined with the evaluation index, a surface overall texture morphology classification method and a method for determining the tool posture angle range corresponding to each category are proposed, providing reasonable constraints for the adjustment of the tool posture in the accuracy maintenance strategy, achieving the goal of improving the quality of the machined surface morphology while ensuring the overall texture morphology quality and integrity; avoiding the problem of poor texture morphology due to improper tool posture adjustment.

[0035] Other advantages, objectives and features of the present invention will be embodied in part through the following description, and in part will be understood by those skilled in the art through study and practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 Schematic diagram of the Z value residual height matrix.

[0037] Figure 2 Schematic diagram of digital extraction of overall texture contour of the surface.

[0038] Figure 3 The second-order moment extraction result.

[0039] Figure 4 is the entropy extraction result.

[0040] Figure 5 Extract results for correlation.

[0041] Figure 6 Extract results for roughness.

[0042] Figure 7 This is the result of surface texture contour extraction under partial rotation angle φ=1°.

[0043] Figure 8 This is the result of surface texture contour extraction under partial rotation angle φ=30°.

[0044] Fig. 9 Determine the flow chart for the right boundary point value of the range.

[0045] Fig.10 Classification of surface texture morphology and corresponding tool attitude angle range.

[0046] Fig.11Area plot of results for tool attitude angle range. DETAILED DESCRIPTION

[0047] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0048] like Figure 1-11 As shown in FIG. 1 , a method for evaluating the overall texture morphology of a free-form surface during five-axis ball head milling is described. The specific steps are as follows:

[0049] Step 1: design an orthogonal experiment based on the tool attitude angle of five-axis ball head milling, and establish a digital model of the texture morphology of the free-form surface of five-axis ball head milling;

[0050] Step 1.1: Based on the tool inclination angle φ and rotation angle θ of the five-axis ball end milling, select the tool inclination angle range (φ l ,φ r ) and tool rotation angle range (θ l ,θ r ) An orthogonal experiment was designed, and a digital model H(i,j) of the surface texture morphology of the workpiece during five-axis ball head milling was established based on the orthogonal experiment, where H is the height value matrix of the digital model, i is the row number of the workpiece surface grid point, and j is the column number of the workpiece surface grid point.

[0051] Step 1.2: Project the height value matrix corresponding to each set of digital models to X W O W Y W Surface, get the Z value residual height matrix Hc(i,j), such as Figure 1 shown.

[0052] In this embodiment, the tool attitude angle inclination angle φ and the rotation angle θ are taken as the main factors, the tool inclination angle range (0°, 30°) and the tool rotation angle range (-90°, 90°) are selected, and the orthogonal experiment is designed, as shown in Table 1. According to Table 1 and the tool attitude angle combination φ = 0°, θ = 0°, 31 groups of surface texture morphology digital models are established.

[0053] Table 1. Orthogonal experimental scheme

[0054]

[0055] Step 2: Extract the overall texture morphology characteristic index of the digital model of the free-form surface texture morphology produced by five-axis ball head milling. Figure 2 As shown, the specific method is as follows:

[0056] Step 2.1, scale the element values ​​in the Z value residual height matrix Hc(i,j) to the grayscale interval [0,255], where 0 represents black and 255 represents white, to obtain the grayscale image matrix G(i,j) of the surface texture morphology. The calculation formula is as follows:

[0057]

[0058] Step 2.2: Calculate the grayscale co-occurrence matrix in four directions based on the grayscale image matrix, and respectively obtain the second-order statistical feature quantity of each co-occurrence matrix, and use the mean value of each feature quantity as the final evaluation index.

[0059] In this embodiment, the extracted gray level co-occurrence matrix features include second-order moment, entropy, and correlation, such as Figure 3-Figure 5 shown.

[0060] Step 2.3, calculate the Tamura texture morphology features based on the grayscale image matrix. In this embodiment, the extracted Tamura texture morphology features include roughness, such as Figure 6 shown.

[0061] Step 2.4: Extract contour lines based on the Z value residual height matrix. In this embodiment, Figure 7 It is the result of surface texture contour extraction under partial rotation angle φ=1°. Figure 8 It is the result of surface texture contour extraction under partial rotation angle of φ=30°.

[0062] Step 2.5: Extract the overall texture morphology feature indexes of all groups, analyze the index extraction results by combining the quartile method, and obtain the preliminary classification of the surface texture morphology by combining the index numerical analysis results. Figure 3-Figure 6 The middle gray line segment is the upper quartile or the lower quartile. The index extraction results are analyzed in combination with the quartiles, and the tool posture angle is divided into two categories: small inclination angle and large inclination angle.

[0063] Step 3: Establish a surface texture subdivision method based on the Z value residual height matrix of the surface texture digital model. The specific method is as follows:

[0064] Step 3.1, taking the Z value residual height matrix data of the surface texture morphology digital model as a sample, the Z value residual height matrix data corresponding to the surface texture morphology digital model of all groups in the orthogonal test table is a data set.

[0065] Step 3.2: Reduce the dimension of the height value data set based on the principal component analysis method. Based on the reduced dimension data set, determine the optimal number of classifications k for the data set based on the silhouette coefficient method.

[0066] Step 3.3: Classify the data set based on the Kmeans clustering algorithm to obtain the surface texture morphology segmentation results.

[0067] Step 4: Establish a method for determining the tool attitude angle range corresponding to each type of surface texture morphology; the details are as follows:

[0068] Step 4.1, set the classification accuracy a of the tilt angle and rotation angle. The higher the classification accuracy, the more accurate the classification result, but the lower the classification efficiency.

[0069] In this embodiment, the tilt angle classification accuracy a is assumed to be φ :a φ =2°, rotation angle classification accuracy a θ :a θ =10°.

[0070] Step 4.2: Use the data set established in step 3.1 as the initial data set to establish a tool attitude angle combination table, and store the tool attitude angle combinations in the orthogonal test table in the tool attitude angle combination table in sequence.

[0071] Step 4.3, starting from the first set of tool attitude angle combinations in the tool attitude angle combination table, determine each set of discrete tool attitude angle combinations {φ i ,θ i}The range of θ i ∈(θ l ,θ r ), φ i ∈(φ l ,φ r ), where φ is the tilt angle, θ is the rotation angle, i is the number of the attitude angle combination in the tool attitude angle combination table, φ l is the left boundary point of the tilt angle, φ r is the right range boundary point of the tilt angle, θ l is the left boundary point of the rotation angle, θ r It is the boundary point of the right range of the rotation angle.

[0072] The specific method is as follows:

[0073] Step 4.3.1, set the right boundary point value θ r ∈(r l ,r r ), the range right boundary point value determination process is as follows Fig. 9 As shown, let (r l ,r r ) Upper boundary r l =θ i Based on the principle that all the division ranges do not overlap, the classification results of the data set are obtained according to step 3, as shown in Table 2. Based on the classification results, the rotation angle set θ is obtained. j , where θj Two conditions must be met: (1) {φ i ,θ j}corresponding height value samples and {φ i ,θ i}The corresponding height value sample does not belong to one category; (2){φ i ,θ j} does not belong to the divided tool posture angle range.

[0074] If θ j If exists, then take all θ j The minimum value θ jmin , so that r =θ jmin If θ j If it does not exist, then r r＝ θ2.

[0075] Table 2. Classification results of the initial dataset

[0076]

[0077] Step 4.3.2, calculate the range (r l ,r r )Intermediate value r m , the formula is as follows:

[0078]

[0079] The attitude angle combination {φ i ,r m} is added to the tool posture angle combination table, and its corresponding samples are added to the data set. According to step 3, the classification result of the data set is obtained. If the posture angle combination {φ i ,r m} samples and {φ i ,θ i} belongs to a class, then in the range (r l ,r m ) randomly select n attitude angle combinations {φ i ,r x}, r x ∈(r l ,r m ), x = 1, 2…n, add n posture angle combinations to the tool posture angle combination table, and add the corresponding samples to the data set. According to step 1, n groups of surface texture morphology digital models are established based on n posture angle combinations to obtain n groups of samples, where n satisfies the following formula:

[0080]

[0081] According to step 3, the classification result of the data set is obtained. If the n groups of samples are all the same as the sample {φi ,θ i} belongs to a class, then r l =r m ; If there is a sample r that does not belong to a class x , take r x Minimum value r xmin , r r =r xmin .

[0082] If the attitude angle combination {φ i ,r m} corresponding to the sample and {φ i ,θ i}The corresponding sample does not belong to a class, then r r =r m .

[0083] Repeat the above steps to continuously narrow the range (r l ,r r ) until it satisfies: r r -r l ≤a (in this embodiment, a is taken here θ =10), determine the right boundary point θ r =r l .

[0084] Step 4.3.3: Determine θ according to the method in step 4.3.2 l ,φ l ,φ r Specific value, get tool posture angle combination {φ i ,θ i}Corresponding attitude angle range θ i ∈(θ l ,θ r ), φ i ∈(φ l ,φ r ). And remove the tool attitude angle combinations that belong to this range from the tool attitude angle combination table.

[0085] Step 4.3.4, according to steps 4.3.1 to 4.3.3, determine the range of the next set of tool attitude angle combinations in the tool attitude angle combination table, until the range of all tool attitude angle combinations is determined.

[0086] Step 4.4: According to step 3, the final data set classification result is obtained. Based on the classification result, the tool attitude angle range corresponding to the tool attitude angle combination in each category is integrated to obtain the tool attitude angle range corresponding to each type of surface texture morphology. The surface texture morphology classification results and the tool attitude angle range corresponding to each category are as follows: Fig.10 As shown, the area diagram of the tool attitude angle range division result is as follows Fig.11 shown.

[0087] The above description is only a preferred embodiment of the present invention and does not constitute any form of limitation to the present invention. Although the present invention has been disclosed as a preferred embodiment as above, it is not intended to limit the present invention. Any technician familiar with the profession can make some changes or modifications to equivalent embodiments of equivalent changes using the technical contents disclosed above without departing from the scope of the technical solution of the present invention. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention still falls within the scope of the technical solution of the present invention.

Claims

1. A method for evaluating the overall texture morphology of a free-form surface during five-axis ball end milling, characterized in that: The steps include: S1. Design an orthogonal experiment with tool attitude angle as a factor. The attitude angle includes the tilt angle and the rotation angle. Establish a digital model H(i, j) of free-form surface texture morphology of five-axis ball head milling. H is the height value matrix of the digital model, i is the row number of the workpiece surface grid point, and j is the column number of the workpiece surface grid point. Project the height value matrix corresponding to each group of digital models onto X W O W Y W Surface, get the Z value residual height matrix Hc(i,j); S2, extracting the overall texture morphology characteristic index of the digital model of the surface texture morphology, and establishing a preliminary classification method for the surface texture morphology; S3, taking the Z value residual height matrix Hc(i,j) data as samples, and the Z value residual height matrix data corresponding to the digital models of surface texture morphology of all groups in the orthogonal test table as data sets, classifying the data sets to obtain the surface texture morphology subdivision results; S4, establishing a method for determining the tool posture angle range corresponding to each type of surface texture morphology, including the following sub-steps: S41, setting the classification accuracy a of the tilt angle and the rotation angle; S42, using the data set established in step S3 as the initial data set, establishing a tool attitude angle combination table, and storing the tool attitude angle combinations in the orthogonal test table in the tool attitude angle combination table in sequence; S43, starting from the first group of tool attitude angle combinations in the tool attitude angle combination table, determine each discrete attitude angle combination {φ i ,θ i }The range of θ i ∈(θ l ,θ r ), φ i ∈(φ l ,φ r ), where φ is the tilt angle, θ is the rotation angle, i is the number of the attitude angle combination in the tool attitude angle combination table, φ l is the left boundary point of the tilt angle, φ r is the right range boundary point of the tilt angle, θ l is the left boundary point of the rotation angle, θ r is the boundary point of the right range of the rotation angle; S44. Based on the data set classification result obtained in step S3, the tool attitude angle ranges corresponding to the tool attitude angle combinations in each category are integrated to obtain the tool attitude angle ranges corresponding to each type of surface texture morphology.

2. The method for evaluating the overall texture morphology of a free-form surface by five-axis ball end milling according to claim 1, characterized in that: Step S43, the specific method is as follows: Step S431, set the right boundary point value θ r ∈(r l ,r r ), let (r l ,r r ) Upper boundary r l =θ i ; Based on the principle that all division ranges do not overlap, determine the upper boundary r r The initial value of Step S432: Calculate the range (r l ,r r )Intermediate value r m : The attitude angle combination {φ i ,r m } add the tool posture angle combination table, and add its corresponding samples to the data set; according to step S3, the data set classification result is obtained. If the posture angle combination {φ i ,r m } samples and {φ i ,θ i } belongs to a class, then in the range (r l ,r m ) randomly select n attitude angle combinations {φ i ,r x }, r x ∈(r l ,r m ), x = 1, 2 ... n, add n posture angle combinations to the tool posture angle combination table, and add the corresponding samples to the data set; according to step S1, n groups of surface texture morphology digital models are established based on the n posture angle combinations to obtain n groups of samples, where n satisfies the formula: If n groups of samples are all the same as the sample {φ i ,θ i } belongs to a class, then r l =r m ; If there is a sample r that does not belong to a class x , take r x Minimum value r xmin , r r =r xmin ; If the attitude angle combination {φ i ,r m } corresponding to the sample and {φ i ,θ i }The corresponding sample does not belong to a class, then r r =r m ; Repeat the above steps to continuously narrow the range (r l ,r r ) until it satisfies: r r -r l ≤a, determine the right boundary point θ r =r l ; Step S433: Determine θ using the same method as step S432. l ,φ l ,φ r Specific value, get tool posture angle combination {φ i ,θ i }Corresponding attitude angle range θ i ∈(θ l ,θ r ), φ i ∈(φ l ,φ r ); Step S434, determining the range of the next set of tool attitude angle combinations in the tool attitude angle combination table according to steps S431-S433, until the ranges of all tool attitude angle combinations are determined.

3. The method for evaluating the overall texture morphology of a free-form surface by five-axis ball end milling according to claim 1, characterized in that: In step S1, based on the inclination angle φ and rotation angle θ of the tool for five-axis ball end milling, the tool inclination angle range (φ l ,φ r ) and tool rotation angle range (θ l ,θ r ) Design an orthogonal experiment and establish a digital model H(i,j) of the surface texture morphology of the workpiece during five-axis ball head milling based on the orthogonal experiment.

4. The method for evaluating the overall texture morphology of a free-form surface by five-axis ball end milling according to claim 1, characterized in that: In step S2, the details are as follows: S21, scaling the element values ​​in the Z value residual height matrix Hc(i,j) to the grayscale interval [0,255], where 0 represents black and 255 represents white, to obtain the grayscale image matrix G(i,j) of the surface texture morphology; S22, based on the gray image matrix G(i,j), calculate the gray co-occurrence matrix in four directions, respectively obtain the second-order statistical feature quantity of each co-occurrence matrix, and use the mean value of each feature quantity as the final evaluation index; S23, calculating Tamura texture morphology features based on the grayscale image matrix G(i,j); S24, extracting contour lines based on the Z value residual height matrix Hc(i,j); S25. Extracting overall texture morphology feature indices of all groups, and establishing a preliminary surface texture morphology classification method based on the overall texture morphology feature indices extraction results.

5. The method for evaluating the overall texture morphology of a free-form surface by five-axis ball end milling according to claim 4, characterized in that: In step S25, the index extraction results are analyzed in combination with the quartile method to obtain a preliminary classification result of the surface texture morphology.

6. The method for evaluating the overall texture morphology of a free-form surface by five-axis ball end milling according to claim 1, characterized in that: The specific method of step S3 is as follows: S31, taking the Z value residual height matrix Hc(i,j) data as a sample, and the Z value residual height matrix data corresponding to the digital model of the surface texture morphology of all groups in the orthogonal test table as a data set; S32, reducing the dimension of the data set obtained in step S31 based on principal component analysis; S33. Classify the data set based on the Kmeans clustering algorithm to obtain the surface texture morphology segmentation results.

Images