Detection method of abnormal deformation degree of wear spot based on deformation area

Abstract

The present invention provides a method for detecting the abnormal deformation degree of wear spots based on deformation areas, which includes the following steps: step 1, segmenting the acquired four-ball friction wear spot image to obtain a wear spot segmentation map; step 2, obtaining an equivalent wear spot map of the wear spot segmentation map; step 3, extracting the deformation area map of the equivalent wear spot map; step 4, calculating the shape characteristics and geometric characteristics of the deformation area map; step 5, calculating the wear spot deformation degree corresponding to the deformation area map based on the obtained shape characteristics and geometric characteristics; step 6, detecting the abnormal morphology of the wear spots of the acquired four-ball friction wear spot image to be detected based on the obtained wear spot deformation degree; this method can realize the rapid comparison of the deformation degrees of multiple wear spots and the detection of the abnormal deformation degree of wear spots. The method has the advantages of being simple, efficient, and not requiring manual assistance throughout the process.

Description

Technical Field

[0001] The invention belongs to the technical field of image test data analysis and processing, in particular to a method for detecting the abnormal deformation degree of wear spots in a deformation area. Background Art

[0002] Research results from the Chinese Academy of Engineering consulting project, "Research on the Current Status and Development Strategy of Tribology Science and Engineering Applications," indicate that in 2006, losses in my country due to friction and wear reached approximately 950 billion yuan, accounting for 4.5% of the country's GDP. Furthermore, increased wear can lead to component failure and machine breakdown, even catastrophic consequences. A high-performance lubricant is a lubricating medium used to reduce frictional resistance and slow wear of friction pairs. It also provides cooling, cleaning, and contamination prevention. Therefore, timely and accurate lubricant performance measurement is crucial for protecting machinery and reducing energy consumption. Four-ball friction and wear testers are widely used in lubricant friction coefficient testing due to their ease of operation, simple structure, short test cycles, low oil consumption, and low cost.

[0003] my country's petrochemical industry standards (GB-T 12583-1998 and HT 0762-2005) define the testing process for lubricating oil friction coefficients and specify the method for observing wear spot morphology. Wear spots are generally circular or elliptical in shape. Severely deformed wear spots cannot be used for lubricating oil friction coefficient testing and require a new run-in test. Therefore, rapid and accurate determination of wear spot shape is crucial for determining wear spot morphology and test validity. However, no research has yet been conducted on automated detection of wear spot deformation in four-ball friction tests. Qualitative determinations are still made by testers based on their experience, inevitably leading to subjective errors. This is neither scientific nor objective, and offers limited practical guidance. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for detecting the abnormal deformation degree of wear spots in the deformation area, thereby solving the above-mentioned shortcomings in the prior art.

[0005] In order to achieve the above object, the technical solution adopted in the present invention is:

[0006] The present invention provides a method for detecting the abnormal deformation degree of wear spots in a deformation area, comprising the following steps:

[0007] Step 1: Segment the acquired four-ball friction wear spot image to obtain a wear spot segmentation map;

[0008] Step 2, obtaining an equivalent wear spot map of the wear spot segmentation map;

[0009] Step 3, extracting the deformation area map of the equivalent wear spot map;

[0010] Step 4, calculating the shape features and geometric features of the deformation area graph;

[0011] Step 5: Calculate the wear spot deformation degree corresponding to the deformation area map based on the obtained shape characteristics and geometric characteristics;

[0012] Step 6: detecting abnormal morphology of the wear spots in the acquired four-ball friction wear spot image according to the obtained wear spot deformation degree.

[0013] Preferably, in step 2, an equivalent wear spot map of the wear spot segmentation map is obtained by:

[0014] Calculate the position and size parameters of the wear spot segmentation map;

[0015] The equivalent wear spot map is constructed based on the obtained position and size parameters.

[0016] Preferably, in step 3, the deformation area map of the equivalent wear spot map is extracted according to the following formula:

[0017]

[0018] Where A is the deformation area; A(x,y)=1 indicates that the pixel (x,y) belongs to the deformation area; A(x,y)=0 indicates that the pixel (x,y) is in the non-deformation area; 1≤x≤M, 1≤y≤N.

[0019] Preferably, in step 4, the shape features of the deformation region map are calculated. Specifically, the shape features include the number of deformation region blocks, the total deformation rate, the maximum deformation rate, the average deformation rate, and the deformation rate variance. The deformation region map includes a plurality of deformation region blocks, wherein:

[0020] Denoising each deformation region block in the deformation region map to obtain a denoised deformation region map;

[0021] Calculate the centroid and area of ​​each deformed region block in the denoised deformed region map;

[0022] According to the centroid and area of ​​each deformed region block, the number of deformed region blocks, the total deformation rate, the maximum deformation rate, the average deformation rate and the deformation rate variance are calculated respectively.

[0023] Preferably, the number of deformed area blocks, the total deformation rate, the maximum deformation rate, the average deformation rate and the deformation rate variance are calculated according to the following formulas:

[0024]

[0025]

[0026]

[0027]

[0028] Among them, B1 is the total number of deformation area blocks in the denoised deformation area map; B2 is the total deformation rate; B3 is the maximum deformation rate; B4 is the average deformation rate; B5 is the deformation rate variance; E i represents the i-th deformation area block in the denoised deformation area E; S is the area of ​​the wear spot segmentation map.

[0029] Preferably, in step 4, the geometric features of the deformation region map are calculated, specifically, the geometric features include the maximum deviation, average deviation and deviation volatility of the deformation region block, wherein:

[0030] Calculate the deformation distance and deviation corresponding to each deformation area block in the denoised deformation area map;

[0031] According to the obtained deformation distance and deviation corresponding to each deformed area block, the maximum deviation, average deviation and deviation fluctuation of the deformed area block are calculated respectively.

[0032] Preferably, the maximum deviation, average deviation and deviation volatility of the deformed area block are calculated based on the obtained deformation distance and deviation corresponding to each deformed area block using the following formula:

[0033] B6=max(e i ) (11)

[0034]

[0035]

[0036] Among them, B6 is the maximum deviation; B7 is the average deviation; B8 is the deviation volatility; E i represents the i-th deformation area block in the denoised deformation area E; e i is the deviation of the i-th deformed area block.

[0037] Preferably, in step 5, the wear spot deformation degree corresponding to the deformation area map is calculated based on the obtained shape characteristics and geometric characteristics of each deformation area block in combination with the following formula:

[0038]

[0039] Among them, γ k is the weight coefficient of the kth feature parameter, B kis the value of the kth characteristic parameter; the characteristic parameters include the number of deformed area blocks, total deformation rate, maximum deformation rate, average deformation rate, deformation rate variance, maximum deviation of deformed area blocks, average deviation and deviation volatility.

[0040] Preferably, in step 6, the abnormal morphology of the wear spots in the acquired four-ball friction wear spot image to be detected is detected according to the obtained wear spot deformation degree and in combination with the following formula. The specific method is:

[0041]

[0042] Among them, L represents the mark of the abnormal deformation degree of the wear spot, L=0 represents normal wear spot morphology, L=0 represents slightly abnormal wear spot morphology, and L=0 represents abnormal wear spot morphology; T1 represents the low threshold for morphological judgment; T2 represents the high threshold for morphological judgment.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] The present invention provides a method for detecting the abnormal deformation degree of wear spots based on deformation areas. Starting from each deformation area block in the deformation area map, the wear spot deformation area is defined and the center distance of the deformation area is calculated, which lays the foundation for the subsequent representation of the distribution characteristics of the deformation area and can greatly reduce the amount of calculation; multiple image features are used to quantitatively describe the number and distribution characteristics of each deformation area, and the establishment and expression of the deformation characteristic parameters of the deformation area are realized; a method for determining the deformation degree of wear spots is established based on the multiple image features of the deformation area. This method can realize the rapid comparison of the deformation degrees of multiple wear spots and the detection of the abnormal deformation degree of wear spots. This method has the advantages of being simple, efficient, and not requiring manual assistance throughout the process. It is very suitable for development into automatic analysis and processing software for four-ball friction test data. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 is the wear spot image f;

[0046] Figure 2 is the wear spot segmentation map R;

[0047] Figure 3 The wear spot segmentation diagram D;

[0048] Figure 4 is the deformation area A;

[0049] Figure 5 is the denoised deformed area E. DETAILED DESCRIPTION

[0050] The present invention will be described in further detail below with reference to the accompanying drawings.

[0051] The specific implementation steps are as follows:

[0052] Step S0: Collection of wear spot images. After the four-ball friction test, wear spots will form on the surface of the steel ball. The surface wear spots (≤1mm) are not visible to the naked eye. It is necessary to use a high-power microscope or scanning electron microscope or other special collection equipment to collect the wear spot image f. Figure 1 shown.

[0053] Step S1: Segmentation of the wear spot image. The segmentation of the wear spot image includes manual segmentation and automatic segmentation methods, and its purpose is to distinguish the wear spot from the background. Compared with the manual segmentation method, the automatic segmentation method has the advantages of high segmentation accuracy and fast speed. The automatic segmentation method performs preliminary segmentation of the wear spot by establishing a dual-extreme filter difference measurement operator, and uses grayscale and distance dual-constrained boundary subdivision to accurately segment the wear spot area boundary. The method includes the following steps: grayscale, denoising, preliminary segmentation of the wear spot area, and dual-constrained boundary segmentation. The algorithm has a good boundary segmentation effect and runs fast.

[0054] In this embodiment, the automatic segmentation method is adopted, and the wear spot segmentation map R is obtained as follows: Figure 2 As shown, the white and black areas in the figure are wear spots and background, respectively.

[0055] Step S2: Size parameters of the wear spot segmentation image. The segmented wear spot (the white area in step S1) is equivalent to a circle. The size parameters (center position and area) of the equivalent circle are the same as those of the segmented wear spot. The size parameter expressions are shown in formula (1-4):

[0056]

[0057]

[0058] S=num(R) (3)

[0059]

[0060] Among them, x C y is the row coordinate of the center of the wear spot segmentation map; C is the column coordinate of the center of the wear spot segmentation map; (x, y) is the coordinate of any pixel, 1≤x≤M and 1≤y≤N; M and N are the maximum row number and the maximum column number of the wear spot segmentation map R, respectively; R(x, y) represents the value of the pixel (x, y) of the wear spot segmentation map R, R is a binary image, R(x, y) = 1 indicates that the pixel (x, y) belongs to the wear spot area, R(x, y) = 0 indicates that the pixel (x, y) is behind the wear spot area; S is the area of ​​the wear spot segmentation map, that is, the number of pixels in the wear spot area; num() is a statistical operator; num(R) is the total number of pixels in the wear spot area in the wear spot segmentation map R; r is the radius value of the wear spot segmentation map.

[0061] In this embodiment, the size parameter of the wear spot segmentation map is: C=365.2125,y C =465.3137, S=396979 and r=355.48.

[0062] Step S3: Construction of equivalent wear spot map. The equivalent wear spot map is constructed based on the size parameters of the wear spot segmentation map. The calculation formula is shown in (5):

[0063]

[0064] Where D is the equivalent wear spot image, which is a binary image. D(x,y)=1 indicates that the pixel (x,y) belongs to the equivalent wear spot area, and D(x,y)=0 indicates that the pixel (x,y) is in the non-equivalent wear spot area. 1≤x≤M, 1≤y≤N.

[0065] In this embodiment, the equivalent wear spot image D is constructed as follows: Figure 3 shown.

[0066] Step S4: Extract the deformation area of ​​the equivalent wear spot image. The present invention starts with the deformation area blocks of the wear spot, considers the number and distribution characteristics of the deformation area blocks, and determines the deformation degree of the wear spot. Taking the equivalent wear spot as a reference, the deformation of the wear spot refers to the area composed of pixels with inconsistent markings in the equivalent wear spot image and the wear spot segmentation image. Its calculation formula is shown in (6):

[0067]

[0068] Where A is the deformation area; A(x,y)=1 indicates that the pixel (x,y) belongs to the deformation area; A(x,y)=0 indicates that the pixel (x,y) is in the non-deformation area, 1≤x≤M, 1≤y≤N.

[0069] The deformation region map includes several deformation region blocks.

[0070] In this embodiment, the deformation area diagram A is as follows Figure 4 shown.

[0071] Step S5: performing denoising processing on each deformed region block in the deformed region map A.

[0072] Based on the deformation region extraction, denoising is performed to eliminate small noise regions that are incorrectly detected. The specific operations are as follows: first, the number of pixels of all deformation region blocks is counted; second, an adaptive denoising threshold T is determined based on expert knowledge; finally, deformation region blocks with an area not larger than T are removed from the deformation region to obtain the denoised deformation region map E:

[0073] T=min((0.005·num(A)),100) (7)

[0074] In this embodiment, T=min(71.04,100)=71, and the denoised deformation region graphs E after processing are respectively as follows: Figure 5 shown.

[0075] Step S6: Counting the characteristic parameters of each deformed region block in the denoised deformed region map.

[0076] The characteristic parameters include the centroid and area of ​​the deformed region block. The area of ​​the deformed region block can be used as one of the index parameters to measure the degree of deformation of the deformed region. Starting from the deformed region block, it is necessary to first perform statistical calculations on the area and position of each deformed region block. The calculation formula of the characteristic parameters of each deformed region block is shown in (8-10):

[0077]

[0078]

[0079] S Di =num(E i ) (10)

[0080] Where B1 is the total number of deformed region blocks in the denoised deformed region map; i represents the index of the deformed region block, i = 1, 2, ..., B1; x Di Indicates the row number of the centroid of the i-th deformed area block; y Di Indicates the column number of the centroid of the i-th deformed area block; E i Represents the i-th deformed area block in the denoising deformed area E. If the pixel (x, y) belongs to E i , then E i (x,y)=1, otherwise E i (x,y)=0;S Di represents the area of ​​the i-th deformed region block.

[0081] Step S7: Statistically analyze the shape characteristic parameters of the deformation area in the denoised deformation area map.

[0082] The overall shape characteristics of the deformation area of ​​the denoised deformation area map E are statistically analyzed, including parameters such as the number of deformation area blocks B1, the total deformation rate, the maximum deformation rate, the average deformation rate, and the deformation rate variance.

[0083]

[0084]

[0085]

[0086]

[0087] Among them, B2 is the total deformation rate; B3 is the maximum deformation rate; B4 is the average deformation rate; and B5 is the deformation rate variance.

[0088] In this embodiment, there are a total of B1 = 33 deformation area blocks; the total deformation rate B2 = 13208 / 396979 = 3.327%; the maximum deformation rate B3 = 2259 / 396979 = 0.569%; the average deformation rate B4 = 400.242 / 396979 = 0.1008%; and the deformation rate variance B5 = 507.642 / 396979 = 0.1279%.

[0089] Step S8: Calculate the deformation distance and deviation corresponding to each deformed area block.

[0090] The distance between the centroid of each deformed area block and the circle center C also reflects the degree of deformation of each deformed area block. The greater the distance, the greater the degree of deformation, and vice versa. The distance between the centroid of each deformed area block and the circle center C is expressed by the following formula:

[0091]

[0092]

[0093] Among them, x Di is the row number of the centroid of the i-th deformed area block; y Di is the column number of the centroid of the i-th deformed area block; d i is the deformation distance of the ith deformed region block, i.e., the Euclidean distance between the centroid of the ith deformed region block and the circle center C, i = 1, 2, ..., B1; e i is the deviation of the i-th deformation area block, reflecting the degree of deviation between the deformation area and the equivalent radius.

[0094] Step S9: Denoise the geometric feature parameters of the deformed area.

[0095] Based on the deformation distance of each deformation area block, the geometric characteristics of the deformation area of ​​the wear spot are statistically analyzed, including parameters such as the maximum deviation, average deviation and deviation fluctuation of the deformation area block.

[0096] B6=max(e i ) (11)

[0097]

[0098]

[0099] Among them, B6 is the maximum deviation; B7 is the average deviation; B8 is the deviation volatility.

[0100] In this embodiment, the maximum deviation B6 = 4.386%, the average deviation B7 = 1.745%, and the deviation volatility B8 = 1.182728%.

[0101] Step S10: quantitatively evaluating the wear spot deformation degree.

[0102] The wear spot deformation degree is determined by the shape characteristics and geometric characteristics of the deformation area block, mainly including eight characteristic variables: the number of deformation area blocks, total deformation rate, maximum deformation rate, average deformation rate, deformation rate variance, maximum deviation, average deviation, and deviation volatility. The number of deformation area blocks has a negative effect on the wear spot deformation degree, that is, the more deformation area blocks n, the smaller the wear spot deformation degree, and vice versa. The remaining seven characteristic variables have a positive effect on the wear spot deformation degree, that is, the larger the characteristic variable value, the greater the wear spot deformation degree, and vice versa. The wear spot deformation degree is represented by H, and the expression is shown in (14).

[0103]

[0104] Among them, γ k is the weight coefficient of the kth feature, and its value is mainly determined by the significance of the influence of each feature variable on . This scheme establishes 8 feature indices, k = 1, 2, 3, 4, 5, 6, 7, 8, which meet the following requirements: B k is the value of the kth feature parameter.

[0105] In this embodiment, considering the order of magnitude and data of the characteristic variables, γ k The values ​​of (k=1,2,3,4,5,6,7,8) are: and H = 1.05%.

[0106] Step S11: Detection of abnormal morphology of wear spots.

[0107] By establishing an abnormal wear spot library, the uniform characteristics of abnormal wear spot deformation are statistically analyzed as the basis for determining whether the wear spot morphology is abnormal. If the wear spot morphology is determined to be abnormal, the test process needs to be standardized and the test needs to be repeated. The established dual-threshold wear spot abnormal deformation determination method has a calculation process shown in Equation (15).

[0108]

[0109] Among them, L represents the mark of the abnormal deformation degree of the wear spot, L=0 represents normal wear spot morphology, L=0 represents slightly abnormal wear spot morphology, and L=0 represents abnormal wear spot morphology; T1 represents the low threshold for morphological judgment; T2 represents the high threshold for morphological judgment.

[0110] In this embodiment, the values ​​of T1 and T2 are 8% and 15% respectively, so L=0, thereby determining that the morphology of the wear spot is normal.

Claims

1. A method for detecting abnormal deformation degree of wear spots in deformation areas, characterized in that: The following steps are involved: Step 1: Segment the acquired four-ball friction wear spot image to obtain a wear spot segmentation map; Step 2, obtaining an equivalent wear spot map of the wear spot segmentation map; Step 3, extracting the deformation area map of the equivalent wear spot map; Step 4, calculating the shape features and geometric features of the deformation area graph; Step 5: Calculate the wear spot deformation degree corresponding to the deformation area map based on the obtained shape characteristics and geometric characteristics; Step 6: detecting abnormal wear spot morphology of the acquired four-ball friction wear spot image according to the obtained wear spot deformation degree; In step 3, the deformation area map of the equivalent wear spot map is extracted according to the following formula: (6) in, is the deformation area; Represents pixels Belongs to the deformation area; Represents pixels is the non-deformation area; , ; Wear spot segmentation diagram Pixels ( x , y ) value; Equivalent wear spot image Pixels ( x , y ) value; In step 4, the shape features of the deformation region map are calculated. Specifically, the shape features include the number of deformation region blocks, the total deformation rate, the maximum deformation rate, the average deformation rate, and the deformation rate variance. The deformation region map includes several deformation region blocks, where: Denoising each deformation region block in the deformation region map to obtain a denoised deformation region map; Calculate the centroid and area of ​​each deformed region block in the denoised deformed region map; According to the centroid and area of ​​each deformed region block, the number of deformed region blocks, the total deformation rate, the maximum deformation rate, the average deformation rate and the deformation rate variance are calculated respectively.

2. The method for detecting abnormal deformation degree of wear spots in deformation areas according to claim 1, characterized in that: In step 2, the equivalent wear spot map of the wear spot segmentation map is obtained by: Calculate the position and size parameters of the wear spot segmentation map; The equivalent wear spot map is constructed based on the obtained position and size parameters.

3. The method for detecting abnormal deformation degree of wear spots in deformation areas according to claim 1, characterized in that: The number of deformed area blocks, total deformation rate, maximum deformation rate, average deformation rate and deformation rate variance are calculated according to the following formulas: (11) (12) (13) (14) in, is the total number of deformation area blocks in the denoised deformation area map; is the total deformation rate; is the maximum deformation rate; is the average deformation rate; is the deformation rate variance; Represents the denoised deformation area Middle deformation area blocks; is the area of ​​the wear spot segmentation map.

4. The method for detecting abnormal deformation degree of wear spots in deformation areas according to claim 1, characterized in that: In step 4, the geometric features of the deformation region map are calculated. Specifically, the geometric features include the maximum deviation, average deviation, and deviation volatility of the deformation region block, where: Calculate the deformation distance and deviation corresponding to each deformation area block in the denoised deformation area map; According to the obtained deformation distance and deviation corresponding to each deformed area block, the maximum deviation, average deviation and deviation fluctuation of the deformed area block are calculated respectively.

5. The method for detecting abnormal deformation degree of wear spots based on deformation area according to claim 4, characterized in that: According to the deformation distance and deviation corresponding to each deformed area block, the maximum deviation, average deviation and deviation volatility of the deformed area block are calculated respectively by combining the following formulas: (11) (12) (13) in, is the maximum deviation; is the average deviation; is the volatility of deviation; Represents the denoised deformation area Middle deformation area blocks; For the The deviation of the deformation area block.

6. The method for detecting abnormal deformation degree of wear spots in deformation areas according to claim 1, characterized in that: In step 5, the wear spot deformation degree corresponding to the deformation area map is calculated based on the shape characteristics and geometric characteristics of each deformation area block obtained by combining the following formula: in, For the The weight coefficient of the feature parameter, ; For the The characteristic parameters include the number of deformation area blocks, total deformation rate, maximum deformation rate, average deformation rate, deformation rate variance, maximum deviation of deformation area blocks, average deviation and deviation volatility.

7. The method for detecting abnormal deformation degree of wear spots in deformation areas according to claim 1, characterized in that: In step 6, based on the obtained wear spot deformation degree, the abnormal wear spot morphology of the acquired four-ball friction wear spot image to be detected is detected in combination with the following formula. The specific method is: (15) in, A mark indicating the abnormal deformation degree of the wear spot, Indicates that the wear spot morphology is normal, Indicates that the wear spot morphology is slightly abnormal. Indicates abnormal wear spot morphology; Indicates the low threshold for morphological determination; Indicates the high threshold for morphological judgment.

Images