Machine vision detection system for texture uniformity of repairing anti-wrinkle face cream

Abstract

The invention discloses a repairing anti-wrinkle face cream texture uniformity machine vision detection system, and relates to the technical field of face cream texture uniformity detection.The repairing anti-wrinkle face cream texture uniformity machine vision detection system comprises the steps that a face cream image standing after stirring is obtained and marked as a real-time face cream image; performing graying processing on the real-time face cream image to obtain a real-time face cream grey-scale map; obtaining a first face cream threshold value and a second face cream threshold value based on the first number of face cream images with uniform texture; based on the real-time face cream grey-scale map, the first face cream threshold value and the second face cream threshold value, obtaining a face cream binary image; obtaining a real-time abnormal ratio based on the face cream binary image; obtaining a historical abnormal ratio based on the second number of face cream images with uniform texture; acquiring a normal ratio threshold based on the historical abnormal ratio; and judging whether the face cream is uniformly stirred based on the real-time abnormal ratio and the normal ratio threshold. The method and the device are used for solving the problem that the accuracy of identifying the uniformity of the face cream is relatively low due to the fact that an uneven area of the cream body is easily identified as non-uniform texture in the prior art.

Description

Technical Field

[0001] This invention relates to the field of cream texture uniformity detection technology, specifically a machine vision detection system for the texture uniformity of anti-wrinkle creams. Background Technology

[0002] In the production process of repair and anti-wrinkle face cream, stirring and emulsification are the core steps to form its final texture and stability. After this process, the cream needs to achieve the preset texture uniformity, especially the texture uniformity after adding anti-wrinkle materials to the repair and anti-wrinkle face cream. Therefore, it is necessary to test the texture uniformity of the face cream.

[0003] Traditional quality control relies on visual inspection at the end of the production line to determine if the texture is uniform after mixing. The results are heavily dependent on the inspector's experience, eyesight, and mental state, lacking objective and standardized criteria, leading to significant fluctuations in judgments between different personnel or batches. With the rapid development of machine vision technology and artificial intelligence algorithms, their application in industrial quality inspection has become an inevitable trend. However, directly applying machine vision technology to detect the uniformity of face cream texture is problematic. Because the surface of the cream is not perfectly smooth, uneven areas appear as shadows under light, which can be mistaken for uneven texture. This results in inaccurate identification of cream texture uniformity. In other words, current technology easily identifies uneven areas as uneven texture, leading to low accuracy in identifying the uniformity of face cream. Summary of the Invention

[0004] This invention aims to at least partially solve one of the technical problems in the prior art. It involves acquiring an image of a face cream after stirring and allowing it to stand, marking it as a real-time face cream image; performing grayscale processing on the real-time face cream image to obtain a real-time face cream grayscale image; obtaining a first face cream threshold and a second face cream threshold based on a first number of face cream images with uniform texture; obtaining a face cream binarized image based on the real-time face cream grayscale image, the first face cream threshold, and the second face cream threshold; obtaining a real-time anomaly ratio based on the face cream binarized image; obtaining a historical anomaly ratio based on a second number of face cream images with uniform texture; obtaining a normal ratio threshold based on the historical anomaly ratio; and determining whether the face cream is evenly stirred based on the real-time anomaly ratio and the normal ratio threshold. This addresses the problem in the prior art where uneven areas of the cream are easily identified as uneven texture, leading to low accuracy in identifying the uniformity of the face cream.

[0005] To achieve the above objectives, this application provides a machine vision inspection system for the texture uniformity of a repair and anti-wrinkle face cream, including: an image acquisition module, a grayscale module, a face cream threshold acquisition module, a binarization module, a real-time anomaly ratio acquisition module, a historical anomaly ratio acquisition module, a ratio threshold acquisition module, and a uniformity judgment module;

[0006] The image acquisition module is used to acquire images of the face cream after stirring and letting it stand still, and marks them as real-time face cream images;

[0007] The grayscale module is used to perform grayscale processing on the real-time face cream image to obtain a real-time face cream grayscale image.

[0008] The face cream threshold acquisition module is used to acquire a first face cream threshold and a second face cream threshold based on a first number of uniform face cream images;

[0009] The binarization module is used to obtain a binarized image of the face cream based on the real-time face cream grayscale image, the first face cream threshold, and the second face cream threshold.

[0010] The real-time anomaly ratio acquisition module is used to acquire the real-time anomaly ratio based on the face cream binarization image;

[0011] The historical anomaly ratio acquisition module is used to acquire the historical anomaly ratio based on a second number of uniformly textured face cream images;

[0012] The ratio threshold acquisition module is used to acquire the normal ratio threshold based on historical abnormal ratios;

[0013] The uniformity judgment module is used to determine whether the face cream is evenly stirred based on the real-time abnormal ratio and normal ratio threshold.

[0014] Furthermore, the face cream threshold acquisition module is configured with a first historical grayscale value acquisition strategy, which includes:

[0015] The first number of uniformly textured face cream images are marked as the first historical images;

[0016] The first historical image is converted to grayscale to obtain the first historical grayscale image;

[0017] Obtain the grayscale value of the pixel in the first historical grayscale image and mark it as the first historical grayscale value.

[0018] Furthermore, the face cream threshold acquisition module is configured with a first quantity threshold acquisition strategy, which includes:

[0019] A number axis is created using the data with the first historical grayscale value as the number axis, and it is marked as a grayscale number axis.

[0020] All the first historical grayscale values ​​are plotted on the grayscale axis to obtain data points, which are then marked as grayscale data points.

[0021] Create a movable line segment of length 1 on the grayscale axis and label it as the first line segment;

[0022] Obtain the length of all grayscale coordinate points on the grayscale axis and mark it as the first overall length;

[0023] The average number of grayscale data points on the first line segment within the first overall length range of the number axis is calculated as: G1 = (H1 ÷ H2) × Z1; where G1 is the average number of grayscale data points on the first line segment within the first overall length range of the number axis, H1 is the first length, H2 is the first overall length, and Z1 is the total number of grayscale data points.

[0024] The first quantity threshold is set as: Y1 = k1 × G1; where Y1 is the first quantity threshold and k1 is the first average proportion value.

[0025] Furthermore, the face cream threshold acquisition module is configured with a first face cream threshold and a second face cream threshold acquisition strategy, the first face cream threshold and the second face cream threshold acquisition strategy including:

[0026] The number of real-time grayscale coordinate points on the first line segment is marked as the real-time search count;

[0027] Mark the left and right endpoints of the first line segment as the first endpoint and the second endpoint, respectively;

[0028] Move the first line segment so that the first endpoint coincides with the smallest grayscale data point. Determine the size of the real-time search count and the first quantity threshold. If the real-time search count is less than the first quantity threshold, move the first line segment to the right so that the first endpoint coincides with the next grayscale data point, until the real-time search count is greater than or equal to the first quantity threshold and then stop. Obtain the first historical grayscale value corresponding to the first endpoint after stopping and mark it as the first face cream threshold.

[0029] Move the first line segment so that the second endpoint coincides with the largest grayscale data point. Determine the size of the real-time search count and the second quantity threshold. If the real-time search count is less than the first quantity threshold, move the first line segment to the left so that the first endpoint coincides with the next grayscale data point, until the real-time search count is greater than or equal to the first quantity threshold and then stop. Obtain the first historical grayscale value corresponding to the second endpoint after stopping and mark it as the second face cream threshold.

[0030] Furthermore, the binarization module is configured with a binarization strategy, which includes:

[0031] Mark the grayscale values ​​of the pixels in the real-time face cream grayscale image as real-time grayscale values;

[0032] Set the real-time grayscale value that is less than the first face cream threshold or greater than the second face cream threshold to 0, and set the real-time grayscale value that is greater than or equal to the first face cream threshold and less than or equal to the second face cream threshold to 255 to obtain the face cream binarized image.

[0033] Furthermore, the real-time anomaly ratio acquisition module is configured with a real-time anomaly ratio acquisition strategy, which includes:

[0034] Obtain the area of ​​the region composed of the grayscale values ​​of pixels in the binarized image of the face cream, and mark it as the real-time abnormal area;

[0035] Obtain the total area of ​​the face cream in the binary image and mark it as the real-time total area;

[0036] Obtain the ratio of the real-time abnormal area to the real-time overall area, and mark it as the real-time abnormal ratio.

[0037] Furthermore, the historical anomaly ratio acquisition module is configured with a historical anomaly ratio acquisition strategy, which includes:

[0038] The second number of uniformly textured face cream images are marked as the second historical images;

[0039] The second historical image is converted to grayscale to obtain the second historical grayscale image;

[0040] The second historical grayscale image is treated as a real-time face cream grayscale image to obtain the real-time anomaly ratio, which is then marked as the historical anomaly ratio.

[0041] Furthermore, the ratio threshold acquisition module is configured with a second quantity threshold acquisition strategy, the second quantity threshold acquisition strategy including:

[0042] A number line is created using historical outlier ratios as the number line, and it is labeled as the ratio number line.

[0043] Plot all historical outlier ratios on the ratio axis and mark the acquired data points as ratio data points;

[0044] Create a movable line segment of length 2 on the ratio axis and label it as the second line segment;

[0045] Obtain the lengths of all ratio coordinate points on the ratio axis and mark them as the second overall length;

[0046] The average number of ratio data points on the second line segment within the second overall length range is calculated as: G2 = (H3 ÷ H4) × Z2; where G2 is the average number of ratio data points on the second line segment within the second overall length range, H3 is the second length, H4 is the second overall length, and Z2 is the total number of ratio data points.

[0047] The second quantity threshold is set as: Y2 = k2 × G2; where Y2 is the second quantity threshold and k2 is the second average proportion value.

[0048] Furthermore, the ratio threshold acquisition module is configured with a normal ratio threshold acquisition strategy, which includes:

[0049] The number of real-time ratio coordinate points on the second line segment is marked as the second real-time count;

[0050] Mark the right endpoint of the second line segment as the third endpoint;

[0051] Move the second line segment so that the third endpoint coincides with the data point with the largest ratio. Determine the size of the second real-time number and the second quantity threshold. If the second real-time number is less than the second quantity threshold, move the second line segment to the left so that the third endpoint coincides with the next ratio data point, until the second real-time number is greater than or equal to the second quantity threshold. Obtain the historical abnormal ratio corresponding to the third endpoint after stopping and mark it as the normal ratio threshold.

[0052] Furthermore, the uniformity judgment module is configured with a uniformity judgment strategy, which includes:

[0053] Determine if the real-time abnormal ratio is greater than the normal ratio threshold. If so, generate a signal indicating uneven texture of the face cream after stirring. If not, generate a signal indicating uniform texture of the face cream after stirring.

[0054] The beneficial effects of this invention are as follows: This invention acquires an image of a face cream after stirring and allowing it to stand, and marks it as a real-time face cream image; it performs grayscale processing on the real-time face cream image to obtain a real-time face cream grayscale image; it obtains a first face cream threshold and a second face cream threshold based on a first number of face cream images with uniform texture; it obtains a face cream binarized image based on the real-time face cream grayscale image, the first face cream threshold, and the second face cream threshold; it obtains a real-time anomaly ratio based on the face cream binarized image; it obtains a historical anomaly ratio based on a second number of face cream images with uniform texture; it obtains a normal ratio threshold based on the historical anomaly ratio; and it determines whether the face cream is stirred evenly based on the real-time anomaly ratio and the normal ratio threshold. The advantage is that it can identify the uniformity of face cream stirring based on the image and eliminate interference caused by uneven face cream, thereby improving the accuracy of face cream uniformity.

[0055] This invention obtains the real-time anomaly ratio based on the binarized image of face cream. Its advantage lies in eliminating interference caused by uneven face cream by using the real-time anomaly ratio, thereby improving the accuracy of face cream uniformity. Attached Figure Description

[0056] Figure 1 This is a schematic diagram of the system of the present invention;

[0057] Figure 2 This is a schematic diagram of the grayscale axis of the present invention;

[0058] Figure 3 This is a schematic diagram of the real-time abnormal area and the real-time overall area of ​​the present invention. Detailed Implementation

[0059] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0060] Example 1, please refer to Figure 1 As shown, this application provides a machine vision inspection system for the texture uniformity of a repair and anti-wrinkle face cream, including an image acquisition module, a grayscale module, a face cream threshold acquisition module, a binarization module, a real-time anomaly ratio acquisition module, a historical anomaly ratio acquisition module, a ratio threshold acquisition module, and a uniformity judgment module.

[0061] The image acquisition module is used to acquire images of the face cream after stirring and then letting it stand, and marks them as real-time face cream images. Acquiring images of the face cream after stirring and letting it stand makes the face cream surface smoother, making it easier to judge whether the face cream is stirred evenly. If it is not evenly stirred, the face cream can be remedied in time. This is especially important for the texture uniformity of anti-wrinkle face creams after adding anti-wrinkle materials.

[0062] The grayscale conversion module is used to process real-time face cream images into grayscale images to obtain a grayscale image of the real-time face cream; the real-time face cream image is an RGB image, which is converted into a grayscale image for easier image analysis;

[0063] In practical applications, for example, the RGB value of a pixel in a real-time face cream image is (180, 181, 182), and the corresponding grayscale value of the pixel is (180+181+182) / 3=181.

[0064] The face cream threshold acquisition module is used to acquire a first face cream threshold and a second face cream threshold based on a first number of uniform face cream images;

[0065] The face cream threshold acquisition module is configured with a first historical grayscale value acquisition strategy, which includes:

[0066] A first number of uniformly textured face cream images are marked as first historical images; the first number is 10 to more accurately obtain the distribution range of the grayscale values ​​of the face cream.

[0067] The first historical image is converted to grayscale to obtain the first historical grayscale image;

[0068] Obtain the grayscale value of the pixel in the first historical grayscale image and mark it as the first historical grayscale value.

[0069] The face cream threshold acquisition module is configured with a first quantity threshold acquisition strategy, which includes:

[0070] A number axis is established using the first historical grayscale value as the axis, and it is marked as the grayscale number axis. The grayscale number axis is established to observe the distribution of the first historical grayscale value, to facilitate the filtering out of the first historical grayscale value that is too small or too large, and thus to obtain a more accurate range of the first historical grayscale value.

[0071] All the first historical grayscale values ​​are plotted on the grayscale axis to obtain data points, which are then marked as grayscale data points.

[0072] For practical applications, please refer to Figure 2 As shown, the grayscale axis is plotted.

[0073] Create a movable line segment with a length of a first length on the grayscale axis, and mark it as the first line segment. The first line segment can be moved on the grayscale axis to facilitate the analysis of sparsely distributed grayscale data points. Therefore, the first length should be less than the first overall length. For example, if the first overall length is 12, then the first length is 1.

[0074] Obtain the length of all grayscale coordinate points on the grayscale axis and mark it as the first overall length;

[0075] The average number of grayscale data points on the first line segment within the first overall length range of the number axis is calculated as: G1 = (H1 ÷ H2) × Z1; where G1 is the average number of grayscale data points on the first line segment within the first overall length range of the number axis, H1 is the first length, H2 is the first overall length, and Z1 is the total number of grayscale data points.

[0076] The first quantity threshold is set as: Y1 = k1 × G1; where Y1 is the first quantity threshold and k1 is the first average proportion value; the first quantity threshold is to obtain the number of sparse grayscale data points on the number axis of the first overall length, so the first quantity threshold must be less than G1, so k1 must be less than 1, for example, k1 is 0.25;

[0077] In practical applications, for example, if the total number of grayscale data points is 200, the average number of grayscale data points on the first line segment within the first overall length range is calculated as: G1 = (1 ÷ 12) × 200 = 16.67. Here, the calculation result is rounded to two decimal places. The first quantity threshold is set as: Y1 = 0.25 × 16.67 = 4.2. Here, the calculation result is rounded to one decimal place, so the first quantity threshold is 4.2.

[0078] The face cream threshold acquisition module is configured with a first face cream threshold and a second face cream threshold acquisition strategy. The first face cream threshold and the second face cream threshold acquisition strategy include:

[0079] The number of real-time grayscale coordinate points on the first line segment is marked as the real-time search count;

[0080] Mark the left and right endpoints of the first line segment as the first endpoint and the second endpoint, respectively;

[0081] Move the first line segment until its first endpoint coincides with the smallest grayscale data point. Determine the size of the real-time search count and the first quantity threshold. If the real-time search count is less than the first quantity threshold, move the first line segment to the right until its first endpoint coincides with the next grayscale data point, until the real-time search count is greater than or equal to the first quantity threshold. Obtain the first historical grayscale value corresponding to the first endpoint after stopping, and mark it as the first face cream threshold. The first face cream threshold is the minimum value of the first historical grayscale value after filtering out values ​​that are too small and far from other first historical grayscale values. This eliminates interference and makes the range of the first historical grayscale values ​​more accurate.

[0082] Move the first line segment until the second endpoint coincides with the largest grayscale data point. Determine the size of the real-time search count and the second quantity threshold. If the real-time search count is less than the first quantity threshold, move the first line segment to the left until the first endpoint coincides with the next grayscale data point, until the real-time search count is greater than or equal to the first quantity threshold. Obtain the first historical grayscale value corresponding to the second endpoint after stopping, and mark it as the second face cream threshold. The second face cream threshold is the maximum value of the first historical grayscale value after filtering out excessively large values ​​and those that are far away from other first historical grayscale values. This eliminates interference and makes the range of the first historical grayscale values ​​more accurate.

[0083] For practical applications, please refer to Figure 2 As shown, for example, moving the first line segment so that the first endpoint coincides with the smallest grayscale data point 174, the real-time search count is 4, which is less than the first quantity threshold 4.2. Moving the first line segment to the right so that the first endpoint coincides with the next grayscale data point 175, the real-time search count is 20, which is greater than the first number threshold 4.2, and then stops. The first historical grayscale value corresponding to the first endpoint after stopping is 175, so the first face cream threshold is 175. Moving the first line segment so that the second endpoint coincides with the largest grayscale data point 186, the real-time search count is 4, which is less than the first quantity threshold 4.2. Moving the first line segment to the right so that the second endpoint coincides with the next grayscale data point 185, the real-time search count is 21, which is greater than the first number threshold 4.2, and then stops. The first historical grayscale value corresponding to the second endpoint after stopping is 185, so the second face cream threshold is 185.

[0084] The binarization module is used to obtain a binarized image of the face cream based on the real-time face cream grayscale image, the first face cream threshold, and the second face cream threshold.

[0085] The binarization module is configured with binarization strategies, which include:

[0086] Mark the grayscale values ​​of the pixels in the real-time face cream grayscale image as real-time grayscale values;

[0087] Set the real-time grayscale value that is less than the first face cream threshold or greater than the second face cream threshold to 0, and set the real-time grayscale value that is greater than or equal to the first face cream threshold and less than or equal to the second face cream threshold to 255 to obtain a face cream binarized image; obtaining the face cream binarized image facilitates the acquisition of the real-time anomaly ratio.

[0088] The real-time anomaly ratio acquisition module is used to obtain the real-time anomaly ratio based on the face cream binarization image;

[0089] The real-time anomaly ratio acquisition module is configured with a real-time anomaly ratio acquisition strategy, which includes:

[0090] Obtain the area of ​​the region composed of the grayscale values ​​of pixels in the binarized image of the face cream, and mark it as the real-time abnormal area;

[0091] Obtain the total area of ​​the face cream in the binary image and mark it as the real-time total area;

[0092] The ratio of the real-time abnormal area to the real-time overall area is obtained and marked as the real-time abnormal ratio. The real-time abnormal ratio is the proportion of the abnormal area. If the texture is uniform, the real-time abnormal ratio will be larger. If the cream texture is uniform but a small part of the cream is uneven and forms a shadow, the real-time abnormal ratio will be smaller. Therefore, the real-time abnormal ratio can be obtained to reduce false judgments.

[0093] For practical applications, please refer to Figure 3 As shown, this represents the region where the real-time anomaly area and the real-time overall area are obtained.

[0094] The historical anomaly ratio acquisition module is used to acquire the historical anomaly ratio based on a second number of uniformly textured face cream images;

[0095] The historical anomaly ratio acquisition module is configured with a historical anomaly ratio acquisition strategy, which includes:

[0096] A second number of uniformly textured face cream images are labeled as second historical images; in order to obtain the range of historical anomaly ratios for uniformly textured face creams, for example, the second number is 100;

[0097] The second historical image is converted to grayscale to obtain the second historical grayscale image;

[0098] The second historical grayscale image is treated as a real-time face cream grayscale image to obtain the real-time anomaly ratio, which is then marked as the historical anomaly ratio; the method used here is the same as that used to obtain the real-time anomaly ratio, and will not be repeated here.

[0099] The ratio threshold acquisition module is used to obtain the normal ratio threshold based on historical abnormal ratios;

[0100] The ratio threshold acquisition configuration includes a second quantity threshold acquisition strategy, which includes:

[0101] A number line is created using historical outlier ratios as the axis, and this is labeled as the ratio axis; the ratio axis is used to observe the distribution of historical outlier ratios.

[0102] Plot all historical outlier ratios on the ratio axis and mark the acquired data points as ratio data points;

[0103] Create a movable line segment of length 2 on the ratio axis and label it as the second line segment. The movement of the second line segment on the ratio axis facilitates the analysis of sparsely distributed ratio data points. Therefore, the second length should be less than the second overall length. For example, if the first overall length is 10, then the first length is 1.

[0104] Obtain the lengths of all ratio coordinate points on the ratio axis and mark them as the second overall length;

[0105] The average number of ratio data points on the second line segment within the second overall length range is calculated as: G2 = (H3 ÷ H4) × Z2; where G2 is the average number of ratio data points on the second line segment within the second overall length range, H3 is the second length, H4 is the second overall length, and Z2 is the total number of ratio data points.

[0106] The second quantity threshold is set as: Y2 = k2 × G2; where Y2 is the second quantity threshold and k2 is the second average proportion value; the second quantity threshold is to obtain the number of smaller ratio data points on the number axis of the second overall length, so the second quantity threshold should be less than G2, and therefore k2 should be less than 1, for example, k2 is 0.2;

[0107] In practical applications, the average number of ratio data points on the first line segment within the second overall length range is calculated as: G2 = (1 ÷ 10) × 100 = 10; the second quantity threshold is set as: Y2 = 0.2 × 10 = 2, then the first quantity threshold is 2.

[0108] The ratio threshold acquisition module is configured with a normal ratio threshold acquisition strategy, which includes:

[0109] The number of real-time ratio coordinate points on the second line segment is marked as the second real-time count;

[0110] Mark the right endpoint of the second line segment as the third endpoint;

[0111] Move the second line segment until the third endpoint coincides with the largest ratio data point. Determine the size of the second real-time count and the second quantity threshold. If the second real-time count is less than the second quantity threshold, move the second line segment to the left until the third endpoint coincides with the next ratio data point, until the second real-time count is greater than or equal to the second quantity threshold. Obtain the historical abnormal ratio corresponding to the third endpoint after stopping, and mark it as the normal ratio threshold. The normal ratio threshold is used to delete excessively large historical abnormal ratios, thereby obtaining relatively accurate historical abnormal ratios. The normal ratio threshold is the maximum real-time abnormal ratio allowed for a face cream with a uniform texture.

[0112] In practical applications, for example, the threshold for the normal ratio is 0.052.

[0113] The uniformity judgment module is used to determine whether the face cream is evenly mixed based on the real-time abnormal ratio and normal ratio threshold.

[0114] The uniformity judgment module is configured with a uniformity judgment strategy, which includes:

[0115] Determine if the real-time anomaly ratio is greater than the normal ratio threshold. If so, generate a signal indicating uneven texture of the face cream after mixing. If not, generate a signal indicating uniform texture of the face cream after mixing. Exclude cases where the grayscale value of the uneven area of ​​the face cream exceeds the normal grayscale value range. Also, because the face cream temperature is higher than normal during production, the uneven area of ​​the face cream is smaller. If the real-time anomaly ratio is less than the normal ratio threshold, it may be caused by a small number of uneven areas of the face cream, and the actual texture of the face cream after mixing is uniform.

[0116] In practical applications, if the real-time anomaly ratio is 0.032, which is less than the normal ratio threshold of 0.052, a signal indicating that the face cream has a uniform texture after stirring is generated.

[0117] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0118] In the embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the displayed or discussed mutual couplings or direct couplings or communication connections may be through some communication interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms.

Claims

1. A machine vision inspection system for the texture uniformity of repair and anti-wrinkle face cream, characterized in that, include: The system includes an image acquisition module, a grayscale conversion module, a face cream threshold acquisition module, a binarization module, a real-time anomaly ratio acquisition module, a historical anomaly ratio acquisition module, a ratio threshold acquisition module, and a uniformity judgment module. The image acquisition module is used to acquire images of the face cream after stirring and letting it stand still, and marks them as real-time face cream images; The grayscale module is used to perform grayscale processing on the real-time face cream image to obtain a real-time face cream grayscale image. The face cream threshold acquisition module is used to acquire a first face cream threshold and a second face cream threshold based on a first number of uniform face cream images; The binarization module is used to obtain a binarized image of the face cream based on the real-time face cream grayscale image, the first face cream threshold, and the second face cream threshold. The real-time anomaly ratio acquisition module is used to acquire the real-time anomaly ratio based on the face cream binarization image; The historical anomaly ratio acquisition module is used to acquire the historical anomaly ratio based on a second number of uniformly textured face cream images; The ratio threshold acquisition module is used to acquire the normal ratio threshold based on historical abnormal ratios; The uniformity judgment module is used to determine whether the face cream is evenly stirred based on the real-time abnormal ratio and normal ratio threshold.

2. The machine vision inspection system for the texture uniformity of the repair and anti-wrinkle face cream according to claim 1, characterized in that, The face cream threshold acquisition module is configured with a first historical grayscale value acquisition strategy, which includes: The first number of uniformly textured face cream images are marked as the first historical images; The first historical image is converted to grayscale to obtain the first historical grayscale image; Obtain the grayscale value of the pixel in the first historical grayscale image and mark it as the first historical grayscale value.

3. The machine vision inspection system for the texture uniformity of the repair and anti-wrinkle face cream according to claim 2, characterized in that, The face cream threshold acquisition module is configured with a first quantity threshold acquisition strategy, which includes: A number axis is created using the data with the first historical grayscale value as the number axis, and it is marked as a grayscale number axis. All the first historical grayscale values ​​are plotted on the grayscale axis to obtain data points, which are then marked as grayscale data points. Create a movable line segment of length 1 on the grayscale axis and label it as the first line segment; Obtain the length of all grayscale coordinate points on the grayscale axis and mark it as the first overall length; The average number of grayscale data points on the first line segment within the first overall length range of the number axis is calculated as: G1 = (H1 ÷ H2) × Z1; where G1 is the average number of grayscale data points on the first line segment within the first overall length range of the number axis, H1 is the first length, H2 is the first overall length, and Z1 is the total number of grayscale data points. The first quantity threshold is set as: Y1 = k1 × G1; where Y1 is the first quantity threshold and k1 is the first average proportion value.

4. The machine vision inspection system for the texture uniformity of the repair and anti-wrinkle face cream according to claim 3, characterized in that, The face cream threshold acquisition module is configured with a first face cream threshold and a second face cream threshold acquisition strategy, the first face cream threshold and the second face cream threshold acquisition strategy including: The number of real-time grayscale coordinate points on the first line segment is marked as the real-time search count; Mark the left and right endpoints of the first line segment as the first endpoint and the second endpoint, respectively; Move the first line segment so that the first endpoint coincides with the smallest grayscale data point. Determine the size of the real-time search count and the first quantity threshold. If the real-time search count is less than the first quantity threshold, move the first line segment to the right so that the first endpoint coincides with the next grayscale data point, until the real-time search count is greater than or equal to the first quantity threshold and then stop. Obtain the first historical grayscale value corresponding to the first endpoint after stopping and mark it as the first face cream threshold. Move the first line segment so that the second endpoint coincides with the largest grayscale data point. Determine the size of the real-time search count and the second quantity threshold. If the real-time search count is less than the first quantity threshold, move the first line segment to the left so that the first endpoint coincides with the next grayscale data point, until the real-time search count is greater than or equal to the first quantity threshold and then stop. Obtain the first historical grayscale value corresponding to the second endpoint after stopping and mark it as the second face cream threshold.

5. The machine vision inspection system for the texture uniformity of the repair and anti-wrinkle face cream according to claim 4, characterized in that, The binarization module is configured with a binarization strategy, which includes: Mark the grayscale values ​​of the pixels in the real-time face cream grayscale image as real-time grayscale values; Set the real-time grayscale value that is less than the first face cream threshold or greater than the second face cream threshold to 0, and set the real-time grayscale value that is greater than or equal to the first face cream threshold and less than or equal to the second face cream threshold to 255 to obtain the face cream binarized image.

6. The machine vision inspection system for the texture uniformity of the repair and anti-wrinkle face cream according to claim 5, characterized in that, The real-time anomaly ratio acquisition module is configured with a real-time anomaly ratio acquisition strategy, which includes: Obtain the area of ​​the region composed of the grayscale values ​​of pixels in the binarized image of the face cream, and mark it as the real-time abnormal area; Obtain the total area of ​​the face cream in the binary image and mark it as the real-time total area; Obtain the ratio of the real-time abnormal area to the real-time overall area, and mark it as the real-time abnormal ratio.

7. The machine vision inspection system for the texture uniformity of the repair and anti-wrinkle face cream according to claim 6, characterized in that, The historical anomaly ratio acquisition module is configured with a historical anomaly ratio acquisition strategy, which includes: The second number of uniformly textured face cream images are marked as the second historical images; The second historical image is converted to grayscale to obtain the second historical grayscale image; The second historical grayscale image is treated as a real-time face cream grayscale image to obtain the real-time anomaly ratio, which is then marked as the historical anomaly ratio.

8. The machine vision inspection system for the texture uniformity of the repair and anti-wrinkle face cream according to claim 7, characterized in that, The ratio threshold acquisition module is configured with a second quantity threshold acquisition strategy, the second quantity threshold acquisition strategy including: A number line is created using historical outlier ratios as the number line, and it is labeled as the ratio number line. Plot all historical outlier ratios on the ratio axis and mark the acquired data points as ratio data points; Create a movable line segment of length 2 on the ratio axis and label it as the second line segment; Obtain the lengths of all ratio coordinate points on the ratio axis and mark them as the second overall length; The average number of ratio data points on the second line segment within the second overall length range is calculated as: G2 = (H3 ÷ H4) × Z2; where G2 is the average number of ratio data points on the second line segment within the second overall length range, H3 is the second length, H4 is the second overall length, and Z2 is the total number of ratio data points. The second quantity threshold is set as: Y2 = k2 × G2; where Y2 is the second quantity threshold and k2 is the second average proportion value.

9. The machine vision inspection system for the texture uniformity of the repair and anti-wrinkle face cream according to claim 8, characterized in that, The ratio threshold acquisition module is configured with a normal ratio threshold acquisition strategy, which includes: The number of real-time ratio coordinate points on the second line segment is marked as the second real-time count; Mark the right endpoint of the second line segment as the third endpoint; Move the second line segment so that the third endpoint coincides with the data point with the largest ratio. Determine the size of the second real-time number and the second quantity threshold. If the second real-time number is less than the second quantity threshold, move the second line segment to the left so that the third endpoint coincides with the next ratio data point, until the second real-time number is greater than or equal to the second quantity threshold. Obtain the historical abnormal ratio corresponding to the third endpoint after stopping and mark it as the normal ratio threshold.

10. The machine vision inspection system for the texture uniformity of the repair and anti-wrinkle face cream according to claim 9, characterized in that, The uniformity judgment module is configured with a uniformity judgment strategy, which includes: Determine if the real-time abnormal ratio is greater than the normal ratio threshold. If so, generate a signal indicating uneven texture of the face cream after stirring. If not, generate a signal indicating uniform texture of the face cream after stirring.

Images