Automatic identification method and device for aging cracks of rubber material and medium

Abstract

The invention discloses an automatic identification method and device for rubber material aging cracks and a medium, and relates to the technical field of rubber product manufacturing. The method comprises the following steps: acquiring a rubber image; enhancing the rubber image to obtain an enhanced image; performing brightness correction on the enhanced image, and performing filtering processing to obtain a corrected image; performing filtering processing on the corrected image to obtain an RGB three-channel color image; converting the RGB three-channel color image into a binary black-and-white image; wherein the binarized black-and-white image is used for distinguishing different colors of a matrix and a crack of the sample; marking a crack region of the binarized black-and-white image; and based on the marked crack region, calculating the number of cracks and the proportion of the area of the cracks in the total sample so as to judge the degree of the crack. According to the method, high-resolution imaging and an advanced image processing method are utilized, the damage degree of aging to the rubber product is objectively quantified, and more accurate, efficient, objective and reliable rubber material aging crack recognition is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of rubber product manufacturing, and more specifically, to an automatic recognition method, device and medium for aging cracks of rubber materials. Background Art

[0002] Rubber materials are widely used in various industries due to their unique properties such as elasticity, durability and wear resistance. However, since they are often exposed to various environmental factors, it will cause surface cracking of rubber products and a significant shortening of service life, accelerating their aging. Aging is a common problem in the rubber industry. Aging factors include ozone aging, thermal-oxidative aging, photo-aging, etc. They will trigger a chain reaction in the rubber matrix, leading to the formation of cracks and the loss of mechanical properties. Accurately identifying and analyzing various rubber aging is crucial for quality control, life prediction, and developing more durable rubber materials. Although the national standards for ozone aging tests "Vulcanized rubber or thermoplastic rubber - Resistance to ozone cracking - Static tensile test" GB / T 7762-2014, the national standards for thermal-oxidative aging tests "Vulcanized rubber or thermoplastic rubber - Accelerated aging and heat resistance tests in hot air" GB / T 3512-2014, and the national standards for photo-aging tests "Laboratory light source exposure test methods - Part 2 Xenon arc lamp" GB / T 16422.2-2014 and other standards have requirements for the recording of experimental results and the evaluation of the degree of cracking, it is not easy to record the number of cracks per unit area and the average length of the 10 largest cracks. Traditional methods for evaluating aging cracking of rubber materials usually rely on manual inspection or qualitative evaluation, which will lead to inconsistencies in subjective judgment. In addition, some initial small cracks cannot be effectively observed and recorded in time, and early degradation cannot be detected, so preventive measures cannot be effectively taken. Summary of the Invention

[0003] To solve the above technical problems, the present invention provides an automatic recognition method, device and medium for aging cracks of rubber materials, which can automatically recognize aging cracks based on the captured rubber images, improve the accuracy of crack recognition, and can effectively observe and record initial small cracks, so as to detect early degradation.

[0004] In the first aspect, the present invention provides an automatic recognition method for aging cracks of rubber materials, the method comprising:

[0005] Obtain a rubber image;

[0006] Enhance the rubber image to obtain an enhanced image;

[0007] Perform brightness correction on the enhanced image and perform filtering processing to obtain a corrected image;

[0008] Filter the corrected image using a Gaussian low-pass operator and a boundary filling filter to obtain an RGB three-channel color image;

[0009] Convert the RGB three-channel color image into a binary black-and-white image; wherein, the binary black-and-white image is used to distinguish the different colors of the matrix and cracks of the specimen;

[0010] Label the crack area of the binary black-and-white image;

[0011] Based on the labeled crack area, calculate the number of cracks and the proportion of the area of the cracks in the total specimen, and judge the degree of the crack according to the proportion.

[0012] Further, the rubber image is a photo of the specimen taken after being subjected to set ozone aging conditions, thermal-oxidative aging conditions or ultraviolet light aging conditions.

[0013] Further, enhance the rubber image through the following formula to obtain an enhanced image:

[0014]

[0015] In the formula, I is the rubber image; low_in and high_in are respectively the low gray-scale threshold and the high gray-scale threshold of the input image, low_out and high_out are respectively the low gray-scale threshold and the high gray-scale threshold of the output image; I2 is the enhanced image.

[0016] Further, correct the brightness of the enhanced image through the following formula and perform filtering processing to obtain a corrected image:

[0017] I3 = imbothat(I2) = close(I2, SE) - I2

[0018] close(I2, SE) = erode(dilate(I2, SE), SE)

[0019]

[0020]

[0021] erode(I2, SE) = I2! SE = min(s,t)∈SE{I2(x + s, y + t) - SE(s,t)}

[0022] In the formula, I3 represents the corrected image, The symbol "⊕" represents the dilation operation, "!" represents the erosion operation, (x, y) are the pixel coordinates in the output image, (s, t) are the element coordinates in the structuring element SE, I2(x - s, y - t) is the pixel value of the enhanced image at the coordinate (x - s, y - t), I(x + s, y + t) is the pixel value of the rubber image I at the coordinate (x + s, y + t), SE(s, t) is the value of the structuring element SE at the coordinate (s, t), SE is a morphological structuring element, imbothat is to further decompose the bottom-hat filtered image into the difference between the original image and the result of its closing operation, close is to perform the closing operation on the enhanced image I2 through the structuring element SE, which is further decomposed into an operation of dilation first and then erosion, erode is the erosion operation, which checks whether the structuring element is completely contained within the foreground region of the image. If the structuring element is completely contained within the foreground region of the image, then the eroded image will maintain the foreground value at the corresponding position. If any part of the structuring element overlaps with the background of the image, then the eroded image will become the background value at the corresponding position. Dialate is the dilation operation. When the structuring element overlaps with the image, it checks whether any part of the foreground region of the image overlaps with the structuring element. If any part of the foreground region of the image overlaps with the structuring element, then the dilated image will become the foreground value at the corresponding position. max is the maximum value function, and min is the minimum value function.

[0023] Further, using a Gaussian low-pass operator and a boundary padding filter, the corrected image is filtered through the following formula to obtain an RGB three-channel color image:

[0024]

[0025] In the formula, I4 is the RGB three-channel color image, H is the filter kernel, M and N are the radii of the filter kernel in the row and column directions respectively, H(m, n) is the value of the filter kernel at the position (m, n), I3(x + m, y + n) is the pixel value of the corrected image at the position (x + m, y + n),'replicate' means that the pixel values outside the image boundary will be replicated, that is, the pixel values on the boundary will extend outside the boundary, and imfilter represents the operation of filtering the image.

[0026] Further, the RGB three-channel color image is converted into a binary black-and-white image through the following formula:

[0027] I5 = rgb2gray(I4) = 0.299×R + 0.587×G + 0.114×B

[0028]

[0029] Wherein, I5 is the grayscale image, I4 is the RGB three-channel color image, R is the red channel of the input color image, G is the green channel of the input color image, B is the blue channel of the input color image, level is a threshold value between 0 and 1 for determining the conversion boundary, I6 is the binary black-and-white image, rgb2gray is to convert the color image into a grayscale image, and im2bw is to convert the grayscale image into a binary image.

[0030] Further, the crack area of the binary black-and-white image is labeled by the following formula:

[0031]

[0032] Wherein, Area i is the area of the i-th connected region, (x, y) is the pixel coordinates in the i-th connected region, and the summation is performed on all pixels in the i-th connected region. The connected region is composed of pixel points with a value of 1.

[0033] Further, based on the labeled crack area, calculate the number of cracks and the proportion of the area of the cracks in the total specimen, including:

[0034] Based on the labeled crack area, determine the number of white connected pixel clusters, the number of white pixels, and the total number of pixels in the specimen area;

[0035] Divide the number of white connected pixel clusters and the number of white pixels by the total number of pixels in the specimen area as the number of cracks and the proportion of the area of the cracks in the total specimen.

[0036] In a second aspect, the present invention provides an automatic recognition device for aging cracks of rubber materials. The device includes:

[0037] A data acquisition unit configured to acquire a rubber image;

[0038] An image enhancement unit configured to enhance the rubber image to obtain an enhanced image;

[0039] An image correction unit configured to perform brightness correction on the enhanced image and perform filtering processing to obtain a corrected image;

[0040] An image filtering unit configured to perform filtering processing on the corrected image by using a Gaussian low-pass operator and a boundary filling filter to obtain an RGB three-channel color image;

[0041] An image conversion unit configured to convert the RGB three-channel color image into a binary black-and-white image; wherein, the binary black-and-white image is used to distinguish the different colors of the matrix and cracks of the specimen;

[0042] An image annotation unit, configured to annotate the crack regions of the binarized black-and-white image;

[0043] A crack judgment unit, configured to calculate the number and the proportion of the area of the cracks in the total specimen based on the annotated crack regions, and judge the degree of the crack according to the proportion.

[0044] In a third aspect, the present invention provides a readable storage medium storing one or more programs, which can be executed by one or more processors to implement the method as described above.

[0045] The present invention has at least the following beneficial effects:

[0046] The present invention can automatically identify and analyze ozone aging, thermal-oxidative aging or photo-aging cracks in rubber materials, that is, by using high-resolution imaging and advanced image processing methods, objectively quantify the damage degree caused by aging to rubber products, thereby changing the traditional identification method of rubber aging and realizing more accurate, efficient, objective and reliable identification of rubber material aging cracks. Description of the Drawings

[0047] Figure 1 Shows a flowchart of a method for automatically identifying rubber material aging cracks according to an embodiment of the present invention.

[0048] Figure 2 Shows a schematic diagram of rubber without cracks identification according to an embodiment of the present invention.

[0049] Figure 3 Shows a schematic diagram of rubber with a small number of cracks identification according to an embodiment of the present invention.

[0050] Figure 4 Shows a schematic diagram of rubber with more cracks identification according to an embodiment of the present invention.

[0051] Figure 5 Shows a schematic diagram of rubber with a large number of cracks identification according to an embodiment of the present invention.

[0052] Figure 6 Shows a schematic diagram of rubber photo-aging crack identification according to an embodiment of the present invention.

[0053] Figure 7 Shows a schematic diagram of rubber thermal-oxidative aging crack identification according to an embodiment of the present invention.

[0054] Figure 8 Shows a schematic diagram of rubber photo-aging and thermal-oxidative aging combined crack identification according to an embodiment of the present invention.

[0055] Figure 9Shows a schematic diagram for identifying rubber ozone aging cracks under a 2% tensile stress according to an embodiment of the present invention.

[0056] Figure 10 Shows a schematic diagram for identifying rubber ozone aging cracks under a 10% tensile stress according to an embodiment of the present invention.

[0057] Figure 11 Shows a schematic diagram for identifying rubber ozone aging cracks under a 20% tensile stress according to an embodiment of the present invention.

[0058] Figure 12 Shows a schematic diagram for identifying rubber ozone aging cracks under a 40% tensile stress according to an embodiment of the present invention.

[0059] Figure 13 Shows a schematic diagram for identifying rubber ozone aging cracks under an 80% tensile stress according to an embodiment of the present invention.

[0060] Figure 14 Shows a structural diagram of an automatic identification device for aging cracks of a rubber material according to an embodiment of the present invention. Detailed implementation manners

[0061] To enable those skilled in the art to better understand the technical solutions of the present invention, the present invention will be described in detail below in conjunction with the accompanying drawings and specific implementation manners. The embodiments of the present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments, but it is not a limitation to the present invention. For the various steps described herein, if there is no necessity for a sequential relationship between them, the order in which they are described as examples herein should not be regarded as a limitation. Those skilled in the art should know that they can be adjusted in order as long as the logical relationship between them is not destroyed and the entire process cannot be realized.

[0062] An embodiment of the present invention provides an automatic identification method for aging cracks of a rubber material, as Figure 1 shown, is a flowchart of the automatic identification method for aging cracks of a rubber material. The automatic identification method for aging cracks of a rubber material includes steps S10 to S70, which are introduced in detail as follows.

[0063] S10: Obtain a rubber image.

[0064] In this embodiment, the rubber image is obtained by shooting with a camera component. Exemplarily, the enhanced image is subjected to brightness correction and filtering to obtain a corrected image. The ozone aging test of the rubber specimen is carried out in accordance with the national standard of "Static Tensile Test for Resistance to Odor Cracking of Vulcanized Rubber or Thermoplastic Rubber" GB / T 7762-2014, or the thermo-oxidative aging test of the rubber specimen is carried out in accordance with "Accelerated Aging in Hot Air and Heat Resistance Test for Vulcanized Rubber or Thermoplastic Rubber" GB / T 3512-2014, or the light aging test is carried out in accordance with "Laboratory Light Source Exposure Test Method Part 2 Xenon Arc Lamp" GB / T 16422.2-2014. After the specified ozone aging conditions, thermo-oxidative aging conditions or ultraviolet light aging conditions, a photo of the specimen is taken. The photo of the specimen obtained by shooting is read into the computer to obtain the rubber image.

[0065] S20: Enhance the rubber image to obtain an enhanced image.

[0066] In this embodiment, considering that the shooting of the rubber surface may include a certain background area, it is quite important to eliminate the influence of the background. Therefore, image enhancement is used to suppress the background of the rubber image, enhance the visual perception of the specimen, present the specimen more clearly, and enhance the perception of the feature area for subsequent segmentation and calculation of the aging area.

[0067] In some embodiments, the rubber image is enhanced by the following formula to obtain an enhanced image:

[0068]

[0069] In the formula, I is the rubber image; low_in and high_in are respectively the low gray threshold and high gray threshold of the input image, low_out and high_out are respectively the low gray threshold and high gray threshold of the output image; I2 is the enhanced image.

[0070] In the above calculation formula, both the rubber image I and the enhanced image I2 can be characterized by their pixel values. low_in and high_in default to [0.10, 0.60], and low_out and high_out default to [0, 1], so as to highlight the pixel points between 0.1 and 0.6.

[0071] S30: Perform brightness correction on the enhanced image and perform filtering to obtain a corrected image.

[0072] In this embodiment, considering that the rubber surface is black and generally darker than the background color, the enhanced image is corrected for brightness. For example, the bottom-hat operation can be selected for brightness correction. The bottom-hat operation is also known as the black-hat operation and is a closing operation in image morphology that can highlight darker regions in the image. The enhanced image in step S20 is corrected for brightness to reduce the brightness difference at different shooting angles and is filtered to further eliminate the influence of noise.

[0073] In some embodiments, the enhanced image is corrected for brightness and filtered through the following formula to obtain a corrected image:

[0074] I3 = imbothat(I2) = close(I2, SE) - I2

[0075] close(I2, SE) = erode(dilate(I2, SE), SE)

[0076]

[0077]

[0078] erode(I2, SE) = I2! SE = min(s,t)∈SE{I2(x + s, y + t) - SE(s,t)}

[0079] In the formula, I3 represents the corrected image, ⊕ represents the dilation operation,! represents the erosion operation, (x, y) are the pixel coordinates in the output image, (s, t) are the element coordinates in the structuring element SE, I2(x - s, y - t) is the pixel value of the enhanced image at the coordinate (x - s, y - t), I(x + s, y + t) is the pixel value of the rubber image I at the coordinate (x + s, y + t), SE(s, t) is the value of the structuring element SE at the coordinate (s, t), SE is the morphological structuring element, imbothat is the bottom-hat filtering of the image further decomposed into the difference between the original image and the result of its closing operation, close is the closing operation on the enhanced image I2 through the structuring element SE, further decomposed into the operation of dilation first and then erosion, erode is the erosion operation that checks whether the structuring element is completely contained within the foreground region of the image. If the structuring element is completely contained within the foreground region of the image, then the eroded image will retain the foreground value at the corresponding position. If any part of the structuring element overlaps with the background of the image, then the eroded image will become the background value at the corresponding position. Dialate is the dilation operation that checks whether any part of the foreground region of the image overlaps with the structuring element when the structuring element overlaps with the image. If any part of the foreground region of the image overlaps with the structuring element, then the dilated image will become the foreground value at the corresponding position. max is the maximum value function, and min is the minimum value function.

[0080] Exemplarily, SE is a disk-shaped structuring element with a radius of 15 pixels.

[0081] S40: Filter the corrected image using a Gaussian low-pass operator and a boundary padding filter to obtain an RGB three-channel color image.

[0082] In this embodiment, to reduce image noise and highlight image edges, a 6*6 Gaussian low-pass operator and a replicate boundary padding filter are used for filtering to obtain an RGB three-channel color image.

[0083] In some embodiments, using a Gaussian low-pass operator and a boundary padding filter, the corrected image is filtered through the following formula to obtain an RGB three-channel color image:

[0084]

[0085] In the formula, I4 is the RGB three-channel color image, H is the filter kernel, M and N are the radii of the filter kernel in the row and column directions respectively, H(m,n) is the value of the filter kernel at position (m,n), I3(x+m,y+n) is the pixel value of the corrected image at position (x+m,y+n),'replicate' means that the pixel values outside the image boundary will be replicated, that is, the pixel values on the boundary will extend outside the boundary, and imfilter represents the operation of filtering the image.

[0086] For boundary processing, the'replicate' option means that the pixel values outside the image boundary will be replicated, that is, the pixel values on the boundary will extend outside the boundary. This can avoid discontinuities at the boundary. Specifically, if x+m or y+n exceeds the image boundary, then I3(x+m,y+n) will take the pixel value of the nearest pixel on the boundary. For example, if x+m is less than 1 (i.e., the top boundary of the image), then I3(x+m,y+n) will take the value of I3(1,y+n); if x+m is greater than the height of the image, then I3(x+m,y+n) will take the value of I3(height,y+n). The same rule applies when y+n exceeds the left and right boundaries. This boundary processing method ensures that the filter kernel can be fully applied at the image boundary, thus avoiding the loss of image information at the boundary.

[0087] S50: Convert the RGB three-channel color image into a binary black and white image; wherein, the binary black and white image is used to distinguish the different colors of the matrix and cracks of the specimen.

[0088] In this embodiment, the RGB three-channel color image obtained in step S40 is subjected to Binarized processing to be converted into a "binary black and white image", which can distinguish the different colors of the matrix and cracks of the specimen.

[0089] In some embodiments, the RGB three-channel color image is converted into a binary black and white image through the following formula:

[0090] I5 = rgb2gray(I4) = 0.299×R + 0.587×G + 0.114×B

[0091]

[0092] In the formula, I5 is the grayscale image, I4 is the RGB three-channel color image, R is the red channel of the input color image, G is the green channel of the input color image, B is the blue channel of the input color image, level is a threshold value between 0 and 1 for determining the conversion boundary, I6 is the binary black and white image, rgb2gray is to convert the color image into a grayscale image, and im2bw is to convert the grayscale image into a binary image.

[0093] Specifically, first, the RGB three-channel color image I4 obtained in step S40 is converted into a single-channel grayscale image. R is the red channel of the input color image. G is the green channel of the input color image. B is the blue channel of the input color image. These weights (0.299, 0.587, 0.114) are determined according to the sensitivity of the human eye to different colors. The weight of green is the largest because the human eye is most sensitive to green. The pixel range in I5 is [0, 255], where 0 represents pure black and 255 represents pure white. level is a threshold value between 0 and 1 for determining the conversion boundary. In the converted I6, the pixels are only 0 and 1, and the continuous 1 values are the connected white cracks.

[0094] S60: Label the crack area of the binary black and white image.

[0095] In some embodiments, the crack area of the binary black and white image is labeled through the following formula:

[0096]

[0097] In the formula, Area i is the area of the i-th connected region, (x, y) are the pixel coordinates in the i-th connected region, and the summation is performed for all pixels in the i-th connected region. The connected region consists of pixel points with a value of 1.

[0098] The binarized black-and-white image obtained in step S50 is displayed with white pixels. The connected regions are composed of pixel points with a value of 1, while the background consists of pixel points with a value of 0. Area i i is the area of the i-th connected region. Therefore, the process of calculating the area is actually counting all the 1s in the connected region. At the same time, it is marked with a rectangular box, which is the smallest rectangle defined by the pixels of the connected region and completely contains all the pixel points of the region.

[0099] S70: Based on the marked crack regions, calculate the number of cracks and the proportion of the crack area to the total specimen, and judge the degree of the crack according to the proportion.

[0100] In this embodiment, the purpose of step S70 is result recognition. The computer can automatically calculate the proportion of the number of cracks (the number of white connected pixel clusters) and the area (the number of white pixels in formula 10) to the total specimen (the total number of pixels in the specimen area), and then judge the degree of the crack.

[0101] Next, the embodiments of the present invention will further illustrate the feasibility and progressiveness of the present invention in combination with specific implementation cases.

[0102] Based on the above-described automatic recognition method for aging cracks in rubber materials, three different implementation cases are provided.

[0103] Implementation case 1: Identification of ozone aging cracks in rubber

[0104] Ozone-aged rubber strips follow the corresponding experimental standards. The method provided by the present invention can effectively identify the number of ozone aging cracks and the proportion of the crack area to the total rubber surface, and can mark the specific positions of the oxidation cracks. The length and maximum width of the cracks can be accurately calculated.

[0105] The method provided by the present invention can effectively identify ozone aging cracks in rubber. The experimental conditions for the ozone aging test are: ozone concentration, 50 ± 5 pphm; test temperature, 40 ± 2 °C; relative humidity, 60%; ozone flow rate, 500 ml / min, and ozone aging time 0 - 72 h. The obtained experimental pictures are images of specimens without cracks (see Figure 2 ), specimens with a small number of cracks (see Figure 3 ), specimens with more cracks (see Figure 4 ), and specimens with a large number of cracks (see Figure 5 ). Identify them, and the identification and identification results of ozone aging cracks in rubber specimens are shown in Figures 2 to 5 . The identification results are as follows.

[0106] Figure 2 The identification result of : No cracks in ozone aging of rubber, aging area 0%, number of oxidation cracks 0;

[0107] Figure 3 Recognition result: There are a small number of cracks in the ozone aging of rubber, the number of aging cracks is 37, and the area of aging cracks is 2.4864%;

[0108] Figure 4 Recognition result: There are more cracks in the ozone aging of rubber, the number of aging cracks is 501, and the area of aging cracks is 13.3219%;

[0109] Figure 5 Recognition result: There are a large number of cracks in the ozone aging of rubber, the number of aging cracks is 272, and the area of aging cracks is 52.0685%.

[0110] Implementation Case 2: Identification of Cracks in the Photoaging and Thermal-Oxidative Aging of Rubber

[0111] The method provided by the present invention can effectively identify the cracks in the photoaging and thermal-oxidative aging of rubber. In this real-time case, photoaging experiments, thermal-oxidative aging experiments, and experiments under the combined action of photoaging and thermal-oxidative aging are carried out. Figure 6 The photoaging conditions for the rubber specimen experiment are: ultraviolet intensity, 1.35 ± 0.02 W / m 2 / @340 nm, test temperature, 40 °C, experimental time, 48 h; Figure 7 The thermal-oxidative aging conditions for the rubber specimen experiment are: test temperature, 80 °C, experimental time, 48 h; Figure 8 The experimental conditions for the combined action of photoaging and thermal-oxidative aging of the rubber specimen are: ultraviolet intensity, 1.35 ± 0.02 W / m 2 / @340 nm, test temperature, 80 °C, experimental time, 48 h. Identifying it, the crack identification and recognition results of the photoaging and thermal-oxidative aging specimens of rubber are shown in Figures 6 to 8 , and the recognition results are as follows.

[0112] Figure 6 Recognition result: The cracks under the photoaging of rubber are irregular strip-shaped, the number of aging cracks is 26, and the area of aging cracks is 8.467%;

[0113] Figure 7 Recognition result: The cracks under the thermal-oxidative aging of rubber are small flocculent, the number of aging cracks is 194, and the aging cracks are 34.7347%;

[0114] Figure 8 Recognition result: The cracks under the combined action of photoaging and thermal-oxidative aging of rubber have both strip-shaped and flocculent shapes, and there are more aging cracks. Recognition result: The number of aging cracks is 141, and the aging cracks are 40.8123%.

[0115] Implementation Case 3: Identification of Ozone Aging Cracks in Different Tensile Strain States of Rubber

[0116] The method provided by the present invention can effectively identify the ozone aging cracks of rubber under tensile strain state. In this experiment, ozone aging experiments were carried out under different strain states of 6%, 10%, 20%, 40%, 80%, etc. Other ozone aging experimental conditions were 75±5pphm; test temperature, 40±2°C; relative humidity, 60%; ozone flow rate, 500ml / min, and ozone aging time was 111h. The identification of ozone aging cracks of rubber under different tensile stresses is shown in Figures 9 to 13 . Among them Figure 9 is the identification diagram of ozone aging cracks of rubber under 6% tensile stress, Figure 10 is the identification diagram of ozone aging cracks of rubber under 10% tensile stress, Figure 11 is the identification diagram of ozone aging cracks of rubber under 20% tensile stress, Figure 12 is the identification diagram of ozone aging cracks of rubber under 40% tensile stress, Figure 13 is the identification diagram of ozone aging cracks of rubber under 80% tensile stress. The identification results are as follows.

[0117] Figure 9 Identification result of : For ozone aging of rubber under 6% tensile stress, there are more long cracks, but the number of cracks is small, only 12, and the crack area is 4.3739%;

[0118] Figure 10 Identification result of : For ozone aging of rubber under 10% tensile stress, the number of cracks increases sharply, reaching 80, and the crack area is 7.7195%;

[0119] Figure 11 Identification result of : For ozone aging of rubber under 20% tensile stress, the number of cracks continues to increase to 119, and the crack area is 11.402%;

[0120] Figure 12 Identification result of : For ozone aging of rubber under 40% tensile stress, the number of cracks is 92, and the crack area is 14.7957%;

[0121] Figure 13 Identification result of : For ozone aging of rubber under 80% tensile stress, the number of aging cracks is 117, and the crack area is 14.9749%, which is equivalent to the crack area under 40% tensile stress, indicating that the crack area reaches saturation at about 40%. The number of cracks and the maximum crack size of each picture do not show an obvious increasing trend with the increase of strain level.

[0122] The embodiment of the present invention also provides an automatic identification device for aging cracks of rubber materials, as shown in Figure 14 . This device includes:

[0123] A data acquisition unit 141, configured to acquire rubber images;

[0124] An image enhancement unit 142, configured to enhance the rubber image to obtain an enhanced image;

[0125] An image correction unit 143, configured to perform brightness correction on the enhanced image and perform filtering processing to obtain a corrected image;

[0126] An image filtering unit 144, configured to filter the corrected image by using a Gaussian low-pass operator and a boundary filling filter to obtain an RGB three-channel color image;

[0127] An image conversion unit 145, configured to convert the RGB three-channel color image into a binary black-and-white image; wherein, the binary black-and-white image is used to distinguish the different colors of the matrix and cracks of the specimen;

[0128] An image annotation unit 146, configured to annotate the crack region of the binary black-and-white image;

[0129] A crack judgment unit 147, configured to calculate the number of cracks and the proportion of the area of the cracks in the total specimen based on the annotated crack region, and judge the degree of the crack according to the proportion.

[0130] In some embodiments, the rubber image is a photo of a specimen taken after being subjected to set ozone aging conditions, thermal-oxidative aging conditions, or ultraviolet light aging conditions.

[0131] In some embodiments, the image enhancement unit is further configured to enhance the rubber image by the following formula to obtain an enhanced image:

[0132]

[0133] In the formula, I is the rubber image; low_in and high_in are respectively the low gray threshold and the high gray threshold of the input image, low_out and high_out are respectively the low gray threshold and the high gray threshold of the output image; I2 is the enhanced image.

[0134] In some embodiments, the image correction unit is further configured to perform brightness correction on the enhanced image and perform filtering processing by the following formula to obtain a corrected image:

[0135] I3 = imbothat(I2) = close(I2, SE) - I2

[0136] close(I2, SE) = erode(dilate(I2, SE), SE)

[0137]

[0138]

[0139] erode(I2, SE) = I2! SE = min(s,t)∈SE{I2(x + s, y + t) - SE(s,t)}

[0140] In the formula, I3 represents the corrected image, ⊕ represents the dilation operation,! represents the erosion operation, (x, y) are the pixel coordinates in the output image, (s, t) are the element coordinates in the structuring element SE, I2(x - s, y - t) is the pixel value of the enhanced image at the coordinate (x - s, y - t), I(x + s, y + t) is the pixel value of the rubber image I at the coordinate (x + s, y + t), SE(s, t) is the value of the structuring element SE at the coordinate (s, t), SE is a morphological structuring element, imbothat is to further decompose the bottom-hat filtering of the image into the difference between the original image and the result of its closing operation, close is to perform the closing operation on the enhanced image I2 through the structuring element SE, which is further decomposed into an operation of dilation first and then erosion, erode is the erosion operation, which checks whether the structuring element is completely contained in the foreground area of the image. If the structuring element is completely contained in the foreground area of the image, then the eroded image will maintain the foreground value at the corresponding position. If any part of the structuring element overlaps with the background of the image, then the eroded image will become the background value at the corresponding position. Dialate is the dilation operation. When the structuring element overlaps with the image, it checks whether any part of the foreground area of the image overlaps with the structuring element. If any part of the foreground area of the image overlaps with the structuring element, then the dilated image will become the foreground value at the corresponding position. max is the maximum value function, and min is the minimum value function.

[0141] In some embodiments, the image filtering unit is further configured to use a Gaussian low-pass operator and a boundary padding filter to filter the corrected image through the following formula to obtain an RGB three-channel color image:

[0142]

[0143] In the formula, I4 is the RGB three-channel color image, H is the filter kernel, M and N are the radii of the filter kernel in the row and column directions respectively, H(m, n) is the value of the filter kernel at the position (m, n), I3(x + m, y + n) is the pixel value of the corrected image at the position (x + m, y + n),'replicate' means that the pixel values outside the image boundary will be replicated, that is, the pixel values on the boundary will extend outside the boundary, and imfilter represents the operation of filtering the image.

[0144] In some embodiments, the image conversion unit is further configured to convert the RGB three-channel color image into a binary black and white image through the following formula:

[0145] I5 = rgb2gray(I4) = 0.299×R + 0.587×G + 0.114×B

[0146]

[0147] Wherein, I5 is a grayscale image, I4 is an RGB three-channel color image, R is the red channel of the input color image, G is the green channel of the input color image, B is the blue channel of the input color image, level is a threshold value between 0 and 1 for determining the boundary of the conversion, I6 is a binary black and white image, rgb2gray is to convert a color image into a grayscale image, and im2bw is to convert a grayscale image into a binary image.

[0148] In some embodiments, the image annotation unit is further configured to annotate the crack region of the binary black and white image through the following formula:

[0149]

[0150] Wherein, Area i is the area of the i-th connected region, (x, y) are the pixel coordinates in the i-th connected region, the summation is performed on all pixels in the i-th connected region, and the connected region is composed of pixel points with a value of 1.

[0151] In some embodiments, the crack judgment unit is further configured to:

[0152] Based on the annotated crack region, determine the number of white connected pixel clusters, the number of white pixels, and the total number of pixels in the specimen region;

[0153] Divide the number of white connected pixel clusters and the number of white pixels by the total number of pixels in the specimen region as the proportion of the number and area of cracks in the total specimen.

[0154] It should be noted that the structure of each automatic recognition device for rubber material aging cracks described in this embodiment belongs to the same technical concept as the previously described automatic recognition method for rubber material aging cracks, and achieves the same beneficial effects through the same principle, which will not be elaborated here.

[0155] The embodiment of the present invention also provides a readable storage medium, and the readable storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the method described in any one of the above embodiments.

[0156] Moreover, although exemplary embodiments have been described herein, the scope includes any and all embodiments based on the present invention that have equivalent elements, modifications, omissions, combinations (e.g., schemes that cross various embodiments), adaptations, or alterations. The elements in the claims will be construed broadly based on the language employed in the claims and not limited to the examples described in the specification or during the implementation of the present application, and the examples will be construed as non-exclusive. Thus, the specification and examples are intended to be considered only as examples, and the true scope and spirit are indicated by the following claims and the full scope of their equivalents.

[0157] The above description is intended to be illustrative and not restrictive. For example, the above examples (or one or more aspects thereof) may be used in combination with each other. For example, other embodiments may be used by those of ordinary skill in the art upon reading the above description. Additionally, in the above detailed description, various features may be grouped together to simplify the present invention. This should not be construed as an intention that the features of an unclaimed invention are necessary for any claim. On the contrary, the subject matter of the present invention may be less than all of the features of a particular embodiment of the invention. Thus, the following claims are hereby incorporated into the detailed description as examples or embodiments, where each claim stands on its own as a separate embodiment, and these embodiments are considered to be combinable with each other in various combinations or permutations. The scope of the present invention should be determined with reference to the appended claims and the full scope of the equivalent forms empowered by these claims.

Claims

1. A method for automatically identifying aging cracks in rubber materials, characterized in that: The method comprises: Get the rubber image; enhancing the rubber image to obtain an enhanced image; Performing brightness correction on the enhanced image and performing filtering processing to obtain a corrected image; The rectified image is filtered using a Gaussian low-pass operator and a boundary filling filter to obtain an RGB three-channel color image; Converting the RGB three-channel color image into a binary black-and-white image; wherein the binary black-and-white image is used to distinguish the different colors of the matrix and cracks of the sample; marking the crack area of ​​the binary black-and-white image; Based on the marked crack area, the number of cracks and the proportion of their area to the total sample are calculated, and the extent of the cracks is determined based on the proportion.

2. The method for automatically identifying aging cracks of rubber materials according to claim 1, characterized in that: The rubber image is a photograph of a sample taken after being subjected to a set ozone aging condition, a thermal oxygen aging condition or an ultraviolet aging condition.

3. The automatic identification method of rubber material aging cracks according to claim 1 is characterized in that: The rubber image is enhanced by the following formula to obtain an enhanced image: Where I is the rubber image; low_in and high_in are the low grayscale threshold and high grayscale threshold of the input image, respectively; low_out and high_out are the low grayscale threshold and high grayscale threshold of the output image, respectively; I2 is the enhanced image.

4. The method for automatically identifying aging cracks of rubber materials according to claim 1, characterized in that: The enhanced image is brightness corrected and filtered using the following formula to obtain a corrected image: I3=imbothhat(I2)=close(I2,SE)-I2 close(I2,SE)=erode(dilate(I2,SE),SE) erode(I2,SE)=I2! SE=min(s,t)∈SE{I2(x+s,y+t)-SE(s,t)} Where I3 represents the corrected image, represents the expansion operation, ! represents the erosion operation, (x, y) is the pixel coordinate in the output image, (s, t) is the element coordinate in the structural element SE, I2(xs, yt) is the pixel value of the enhanced image at the coordinate (xs, yt), I(x+s, y+t) is the pixel value of the rubber image I at the coordinate (x+s, y+t), SE(s, t) is the value of the structural element SE at the coordinate (s, t), SE is the morphological structural element, imbothat is to perform bottom hat filtering on the image and further decompose it into the difference between the original image and its closed operation result, close is to perform a closed operation on the enhanced image I2 through the structural element SE, which is further decomposed into the operation of expansion followed by corrosion. Operator, erode is an erosion operation, which checks whether the structural element is completely contained in the foreground area of ​​the image. If the structural element is completely contained in the foreground area of ​​the image, the eroded image will maintain the foreground value at the corresponding position. If any part of the structural element overlaps with the background of the image, the eroded image will become the background value at the corresponding position. Dialate is an expansion operation. When the structural element overlaps with the image, check whether the foreground area of ​​the image overlaps with any part of the structural element. If the foreground area of ​​the image overlaps with any part of the structural element, the expanded image will become the foreground value at the corresponding position. Max is the maximum value function, and min is the minimum value function.

5. The automatic identification method of rubber material aging cracks according to claim 1 is characterized in that: The rectified image is filtered using the Gaussian low-pass operator and the boundary filling filter to obtain an RGB three-channel color image using the following formula: Where I4 is the RGB three-channel color image, H is the filter kernel, M and N are the radii of the filter kernel in the row and column directions respectively, H(m,n) is the value of the filter kernel at position (m,n), I3(x+m,y+n) is the pixel value of the corrected image at position (x+m,y+n), 'replicate' means that the pixel values ​​outside the image boundary will be copied, that is, the pixel values ​​on the boundary will extend outside the boundary, and imfilter means filtering the image.

6. The automatic identification method of rubber material aging cracks according to claim 1 is characterized in that: The RGB three-channel color image is converted into a binary black-and-white image using the following formula: I5=rgb2gray(I4)=0.299×R+0.587×G+0.114×B Wherein, I5 is a grayscale image, I4 is an RGB three-channel color image, R is the red channel of the input color image, G is the green channel of the input color image, B is the blue channel of the input color image, level is a threshold between 0 and 1, which is used to determine the boundary of the conversion, I6 is a binary black and white image, rgb2gray converts a color image into a grayscale image, and im2bw converts a grayscale image into a binary image.

7. The method for automatically identifying aging cracks of rubber materials according to claim 1, characterized in that: The crack area of ​​the binary black and white image is marked by the following formula: Where Area i is the area of ​​the ith connected region, (x, y) is the pixel coordinate in the ith connected region, and the sum is performed on all pixels in the ith connected region. The connected region consists of pixels with a value of 1.

8. The method for automatically identifying aging cracks of rubber materials according to claim 7, characterized in that: Based on the marked crack area, calculate the number of cracks and the proportion of their area to the total specimen, including: Based on the marked crack area, the number of white connected pixel clusters, the number of white pixels and the total pixels of the sample area are determined; The number of white connected pixel clusters and the number of white pixels divided by the total pixels in the sample area are taken as the number and area ratio of cracks to the total sample.

9. An automatic identification device for aging cracks of rubber materials, characterized in that: The device comprises: a data acquisition unit configured to acquire a rubber image; an image enhancement unit, configured to enhance the rubber image to obtain an enhanced image; An image correction unit is configured to perform brightness correction on the enhanced image and perform filtering processing to obtain a corrected image; An image filtering unit is configured to filter the corrected image using a Gaussian low-pass operator and a boundary filling filter to obtain an RGB three-channel color image; An image conversion unit is configured to convert the RGB three-channel color image into a binary black-and-white image; wherein the binary black-and-white image is used to distinguish different colors of the matrix and cracks of the sample; An image annotation unit, configured to annotate the crack region of the binary black-and-white image; The crack judgment unit is configured to calculate the number of cracks and the proportion of the area of ​​the cracks to the total sample based on the marked crack area, and judge the degree of the crack according to the proportion. 10 . A non-transitory computer-readable storage medium storing instructions, which, when executed by a processor, perform the method according to claim 1 .

Images